rinting

9g Celanese

J ndey é
(rin has 3

ip kes ae Se

At Celanese, Safety
is a pre-condition,
to be in business and a requirement
remain competitive

‘9 Celanese
Ee {aslo yo Ok} AN
ly
4

aia

ystye “Cs A. vy [oe

—_ GC | S te (Re

473-245-720)
Every accident can be prevented

Dei |
 Digitized by the Internet Archive
in 2023 with funding from
Kahle/Austin Foundation

https://archive.org/details/printingOOOOoitt
Papermaking Science and Technology
a series of 19 books
covering the latest
technology and
future trends

Printing

Series editors
Johan Gullichsen, Helsinki University of Technology
Hannu Paulapuro, Helsinki University of Technology

Book editor
Pirkko Oittinen, Helsinki University of Technology
Hannu Saarelma, Helsinki University of Technology

Series reviewer
Brian Attwood, St. Anne’s Paper and Paperboard Co. Ltd

Book reviewer
M. Bruce Lyne, International Paper

Published in cooperation with the Finnish Paper Engineers’ Association and

TAPPI
 Cover photo by Martin MacLeod

ISBN 952-5216-00-4 (the series)

ISBN 952-5216-13-6 (book 13)

Published by Fapet Oy
(Fapet Oy, PO BOX 146, FIN-00171 HELSINKI, FINLAND)

Copyright © 1998 by Fapet Oy. All rights reserved.

Printed by Gummerus Oy, Jyvaskyla, Finland 1998

co) Printed on LumiSilk 115 g/m, Enso Fine Papers Oy, Oulu Mills
f Foreword
Johan Gullichsen and
Hannu Paulapuro

PAPERMAKING SCIENCE AND TECHNOLOGY

Papermaking is a vast, multidisciplinary technology that has expanded tremendously in

recent years. Significant advances have been made in all areas of papermaking,
including raw materials, production technology, process control and end products. The
complexity of the processes, the scale of operation and production speeds leave little
room for error or malfunction. Modern papermaking would not be possible without a
proper command of a great variety of technologies, in particular advanced process
control and diagnostic methods. Not only has the technology progressed and new
technology emerged, but our understanding of the fundamentals of unit processes, raw
materials and product properties has also deepened considerably. The variations in the
industry’s heterogeneous raw materials, and the sophistication of pulping and
papermaking processes require a profound understanding of the mechanisms involved.
Paper and board products are complex in structure and contain many different
components. The requirements placed on the way these products perform are wide,
varied and often conflicting. Those involved in product development will continue to
need a profound understanding of the chemistry and physics of both raw materials and
product structures.
Paper has played a vital role in the cultural development of mankind. It still has a
key role in communication and is needed in many other areas of our society. There is no
doubt that it will continue to have an important place in the future. Paper must, however,
maintain its competitiveness through continuous product development in order to meet
the ever-increasing demands on its performance. It must also be produced economi-
cally by environment-friendly processes with the minimum use of resources. To meet
these challenges, everyone working in this field must seek solutions by applying the
basic sciences of engineering and economics in an integrated, multidisciplinary way.
The Finnish Paper Engineers' Association has previously published textbooks
and handbooks on pulping and papermaking. The last edition appeared in the early
80's. There is now a clear need for a new series of books. It was felt that the new series
should provide more comprehensive coverage of all aspects of papermaking science
and technology. Also, that it should meet the need for an academic-level textbook and at
the same time serve as a handbook for production and management people working in
this field. The result is this series of 19 volumes, which is also available as a CD-ROM.
When the decision was made to publish the series in English, it was natural to
seek the assistance of an international organization in this field. TAPPI was the obvious
partner as it is very active in publishing books and other educational material on pulping
and papermaking. TAPPI immediately understood the significance of the suggested
new series, and readily agreed to assist. As most of the contributors to the series are
Finnish, TAPPI provided North American reviewers for each volume in the series. Mr.
Brian Attwood was appointed overall reviewer for the series as a whole. His input is
gratefully acknowledged. We thank TAPPI and its representatives for their valuable con-
tribution throughout the project. Thanks are also due to all TAPPl-appointed reviewers,
whose work has been invaluable in finalizing the text and in maintaining a high standard
throughout the series.
A project like this could never have succeeded without contributors of the very
highest standard. Their motivation, enthusiasm and the ability to produce the necessary
material in a reasonable time has made our work both easy and enjoyable. We have
also learnt a lot in our "own field" by reading the excellent manuscripts for these books.
We also wish to thank FAPET (Finnish American Paper Engineers’ Textbook),
which is handling the entire project. We are especially obliged to Ms. Mari Barck, the
project coordinator. Her devotion, patience and hard work have been instrumental in
getting the project completed on schedule.
Finally, we wish to thank the following companies for their financial support:

A. Ahlstrom Corporation
Enso Oyj
Kemira Oy
Metsa-Serla Corporation
Rauma Corporation
Raisio Chemicals Ltd
Tamfelt Corporation
UPM-Kymmene Corporation

We are confident that this series of books will find its way into the hands of
numerous students, paper engineers, production and mill managers and even profes-
sors. For those who prefer the use of electronic media, the CD-ROM form will provide all
that is contained in the printed version. We anticipate they will soon make paper copies

of most of the material.

g List of Contributors aa
University of Technology, Laboratory of
Oittinen Pirkko — Ph.D., Professor, Helsinki
Media Technology
University of Technology, Laboratory of
Saarelma Hannu — Ph.D., Professor, Helsinki
Media Technology
a PB[ e Td C e Pirkko Oittinen and Hannu Saarelma

Printed matter in its different manifestations is familiar to and used by

everybody; printed products truly represent the much sought-after every-
citizen interface to information. The time spent each day reading and
browsing printed matter for work, education, information and leisure is
considerable and the needs fulfilled are diverse, as are also the user
situations. In other words, printed matter is an essential part of the fabric of
our daily lives.

Print production technologies have changed immensely during the last

forty years, more so than during the preceding four hundred years. This is
mainly due to the digitalization of the basic raw material of prints, namely
information. Previous progress was made largely through mechanization.
However, during this time the appearance of printed matter has not
changed radically.

The contents of this book have been collected and assembled over a long
period of time from lecture and study material used for entry level courses
in graphic arts technology at Helsinki University of Technology. The choice
of the subject matter has been influenced by the authorsi experience of
what students with no background in the technology area of the book are
capable of learning and understanding during a one-semester course. It is
hoped that this will also apply to the readership of the book.

As already mentioned, the book is intended to be readable without any

previous practical experience. It strives to generate a self-contained
framework of its own. This framework should be applicable in practical
working life situations and also when the reader wants to learn more. The
focus is on new technical principles, their relations and causes. The
authors try to show that most of the technology is inter-related and that
development forms a continuum rather than being a set of discrete events.

The approach of the book is consistent with the image science view of
printing and its quantification in terms of the concepts of information
theory. In this respect the book has been influenced most by some seminal
writing in the field of image science. This includes the books on
photography by James1, on color reproduction by Yule2, on image science
by Dainty and Shaw3, and on electronic imaging by William Schreiber4.
The authors express their appreciation of these books. Formulations of
image science and information theory provide a basis for comparisons of
the performance of the different steps in the printing reproduction chain, of
different types of printing and paper, and also of different media.

The book does not discuss how the principles are applied in the
construction of machines or equipment or how these are utilized in print
production or business. Technologies which are bound to be replaced by
more advanced ones are dealt with in as far as they are expected to have
long-term industrial applications. In view of the focus on principles and the
fairly concise presentation, it is expected that the book will also find use as
a handbook.

Printing, in the narrowest sense, means applying a coloring material to

paper in a printing press or other printing device. Here the title of the book
has a wider meaning, referring toéprinting reproduction processi. This is
process which covers the steps from technical generation of information
content to visual perception by the end user. A still wider interpretation,
that of a printing communication process, is also briefly discussed. In this
widest context, comparisons of printing and other communication media
become relevant.

Over the years, Ms Tiina Hartikainen has painstakingly corrected,

recorrected and re-recorrected the manuscript, for which the authors would
like to express their warmest thanks.

!. James, TH, The Theory ofthe Photographic Process. Macmillian Co. New York 1977. 750 p

2. Yule, J.A.C., Color Reproduction.

John Wiley and Sons, New York 1967.41 1 p

3. Dainty, J.C., Shaw, R, Image Science.Academic Press. London | 974. 402 p

4, Schreiber, W. Fundamentals of Electronic Imaging Systems. Springer Verlag. New York | 986. 187 p
Blable of Contents
ii
1. TaDlCvanCOMONtS meter. syrer weeny escraes ca ccrescenee emery camne yg seen tenes eaneupee seen )

2. ADICOMNCUON Sercreseee erence reas Gs enccece ache hoe tne ne eeknaecinae Shear ce, tees caveats cae ees 12

3. [ACINOLSVSIGINSiteaeee soe ncs cece rae ttre ye eaereniast ck aecananene toncsa von coeemeeenercenaee Zi

4. GoloteangiGOl Olas wesc a cteee cscs nason cc coccanecueseaamtannscaraniacsnecanteoneecements 65

9. BUINCIPLESTOUPTIOUNG aneet ss. caves cacdceate cca coc cuusenaustuetoases beanie cecne tana meme 119

6. MIGCHANIEAN OTINUNG crea ce Seve ctecae mms dns hace suman wenn sen evslonn Cree sektcmaenas aucean 133

ths PIGGCUONIG PRONUNG geen ees cee etree en aaa cins cone reeahmna@ee ee mama ke eeheande reLaracies nae 173

8. OnucalamMaginain DUMUNG cece teres cneee cusdanceoy estes ueeese na cer ecm eae 193

9. Papert OMe ecee ee eetont cra fee ch cat cuuea ss caeecsacate sPintaivecd anne taend incemons 213

FOP UAIVUIN COINS Semen ce tece cnc ceccunevetrce ss iced vsec ans scrcapun ace snr ceeuscsp rattan easeuenmceccne 257

GCORVETSIONACIOIS Sy coseteccnes races ter tics cavcevan anemone ape cccieesuaee eu tanionaeeearoe 287
CHAT
Introduction

1 ALYTT RESP a Ree OE ee apie TOR ET 12

2 CommuniGation [email protected]..,cc. cece aeosee eres ace oes occa sveaee rage as cdasmeeuery mceteee 12
3 DevelonimentarnistOty cesr:< 13.0 eecscc seve s totescec ecostetys conse ccrcec aetcecs arose 14
4 PrintegnroduelSeec cr tccrr.ccces cee re swe rece seco ceten ar ec cadanereene meme tere aces 17
41 EILORMAUONPSICIUTCSetss cases eee ce es oo eee ete say Nee ces ak ae fe mee ene 18
ALIX” |MSAARES(OF SIU TeaUIacae-dooke a ee es Dae ee nn eT ep Pei Aiea ree) ee)iP 20
+) PHO GEOR re eet crak eae ees ite rea ae kas se ete tia ence 22
ROTORS CGS met cmmenria uri: be: kareena aie rete re retin ten eer 25
 CHAPTER 1

Mintroductiion
The objective of this textbook is to give a composite view of the principles of printing
processes and the use of materials in the processes.
The first section of this chapter clarifies the scope. The remaining sections place
the topic in perspective by defining a general communication process and examining
the history of development. A review of the structure of reading products and other
printed products and an introductory outline of the operations required in making them
conclude the chapter.

1 Scope
This book considers technologies that are specific or fundamental to the production of
printed products. In such products, paper is a carrier of information. Paper and some
other substrates may have other functions as in tissue and packaging products. These
functions and the processes to achieve them lie outside the scope of this work.
The subject matter covers printing technology from a generic viewpoint trying to
make it applicable to all aspects of printing. Printing finds use mainly in the graphic arts
industry, offices, the paper converting industry, and households.
Chapter 2 covers general principles of imaging systems. A printing process is an
imaging system that consists of digital, analog, electronic, and mechanical components.
The ultimate “component” in the system is the human viewer whose total response
covers a range of activities beginning with the optical reactions of the eye. Chapter 3
concerns color including the principles of color, processing of color in reproduction sys-
tems, and the perception of color.
Chapters 5 to 10 concern printing starting with principles that are relevant for all
types of printing. Chapters 6 and 7 explain the principles of mechanical and electronic
printing methods, and Chapters 9 and 10 focus on paper and ink in printing. Chapter 8
covers the principles by which physical prints are converted to optical images.

2 Communication process
Printing processes are communication processes’. Communication is the transmission
and reception of signals between persons or groups of people. Depending on the
number of people involved in sending and receiving messages, the following basic types
of communication are possible:

- private communication: from one individual to another individual

- group communication: from one group to another group

- mass communication: from one individual or a group to the public

- business communication: from one business to another business.

t2
 Introduction

The first three types are becoming increasingly intermingled, and the boundaries
are disappearing. The term interactive communication is today commonly understood to
be computer mediated communication between individuals, or an individual and a com-
puter program and digital data.
A communication process has the generic steps shown in Fig. 1°. Generic means
that the same steps occur regardless of the communication media.

Information
Packaging > Dissemination Retrieval
creation

SS ee
Publishing

Figure 1. Steps in communication processes.

In the information creation step in Fig. 1, the contents of the communication mes-
sage begin by writing of text, taking and synthesizing of pictures and video, and record-
ing or synthesizing audio. The packaging step assembles the contents into an informa-
tion product such as a daily newspaper or a television program. The combination of infor-
mation creation and packaging is publishing of printed and networked communication.
The dissemination step involves distribution and delivery of the information. For most
printed matter, printing and physical distribution allow access for end users. Electroni-
cally printed products are often disseminated as data for retrieval and printing on paper
by the end user as necessary. Television programs are disseminated by broadcasting or
by transmission over a cable network and received by TV devices. In computer communi-
cation or digital media, dissemination may involve storage in a local or distributed reposi-
tory or an off-line medium for end users. Dissemination may be by electronic mailing or
through a channel mechanism such as narrow casting of multicasting.
In the retrieval step, the end user accesses and uses — reads, browses,
searches, watches, interacts, etc. — the information. Printed products differ from the
other media because they do not require any technical equipment other than a light
source for use. All the other cases require some interactive device such as a television
receiver or a computer connected to a network.
Until recently, mass media communication processes were one-way. The flow of
information was unidirectional from the information provider to the end user. The end

13
CHAPTER 1

user interacted with the product or service only at the retrieval stage. Communication is
undergoing many changes today and is becoming two-way”, The end user may interact
with the process in different steps as Fig. 2 shows.

Contents Contents
creation packaging Dissemination Retrieval

Contents
Traditional naus tly
j E
neues.
mass
communication

Customized
products

Personalized products
Self publications
Group communication

Figure 2. Communication process showing interaction of end user with the process indicated by arrows.

Two-way communication blurs the boundaries of private, group, and mass com-
munication. Compared to today’s communication scene, the future changes require that
the end user becomes an actor in the communication chain rather than a passive recip-
ient. Dissemination of a product on request is on-demand communication. When the
end user influences the packaging of information, the assembly of a media product or
service, the result is customized communication. Mass customization is an area of con-
tinuing change. Tailoring the contents of information according to personal data of the
end user leads to personalized communication. Any individual can create, package, and
disseminate a “self publication.” End users also “create content’ in discussion groups.

3 Developmental history
The techniques and procedures used to produce printed products have evolved over a
period of several centuries. Table 1 summarizes the development of printing and other
visualization methods. The oldest known printing methods date from the seventh century.
New technology emerges continually. Some steps have been revolutionary such as the
invention of movable type by Gutenberg in the 1450s. This brought the art of book
printing into the western culture. Another revolutionizing invention was the cathode ray
tube in the 1920s. This allowed visualization of moving images. The use of computers in
text and image processing was revolutionary. In the 1990s, data sources such as the
Internet are phenomena with revolutionary features although some scholars have the
view that they are expressions from a continual development’. Printing on paper directly

14
 Introduction

from computers began to emerge approximately 15 years ago. Distribution of

publications to computers via networks and end user driven printing are the most recent
developments.

Table 1. Development of printing and other visualization methods.

Visualization method Year

Letterpress 700
Letterpress, metal type 1450
Lithography 1800
Photography 1820
Motion picture 1895
Gravure 1900
Flexography 1900
Television 1926
Photo copying 1940
Computer display 1973
Electronic printing 1980
Networked printing 1990
Electronic publishing on display 1990
Distributed color printing 1995

The Chinese invented the oldest known printing method, letterpress. In the four-
teenth century in Europe, religious pictures and playing cards were engraved on
wooden plates that were inked followed by printing of the images on paper. Later, the
religious pictures had written explanations printed with them. Eventually, engraving was
only for printed text. Gutenberg invented separate characters — first wooden and later
metal. These provided easy assembly to text pages for printing. Gutenberg, his associ-
ate J. Fustin, and their assistant P. Schoffer also developed printing inks for metal
plates. A fire in 1462 destroyed Gutenberg’s and Fustin-Schoffer’s printing plant. The
workers moved to several countries where they established several new printing plants.
The first book intended for distribution in Finland was “Missale Aboense” printed in 1488
in Lybeck. The first book printed in the Finnish language, Mikael Agricola’s “Abc-kiria,”
was printed in 1542 or 1543 in Stockholm. The earliest Finnish printing plants were in
Turku in 1642 and 1668 and in Viipuri in 1689.
In 1998, approximately 700 printing plants in Finland employ 27 000 people with
an annual turnover of approximately 3 billion ECU. The structure is similar in other de-
veloped countries. A general rule of thumb is that printing shares approximately 2%—3%
of gross national product. In a given year, an average Finnish adult expends 7.7 hours
per day to follow mass communication. Of this amount, approximately one-third is read-
ing, one-third is watching television, and one-third is listening to the radio. The cost of
printed products is more than twice the cost of the electronic media when calculated per
time unit of use. Advertisers heavily subsidize the cost of printed products that contain
advertising.

15
CHAPTER 1

Printing technology consists of image and text processing, printing, and finishing.
The big invention in image processing was photography. Halftone photography made it
possible to print photographs by printing presses. Video technology generated television
broadcasting and contributed markedly to the image processing technology in printing
plants. Computers process most pictures printed today. Table 2 shows that computerized
processing of pictures as a production technology has a history of less than 20 years.

Table 2. Development of image processing.

Technology
Manual methods 30 000 BC
Photography 1820
Halftone photography 1880
Video technology 1920
Digital image processing 1980
| Automatic image processing 1995

In text processing, the big invention in western countries was phonetic talk coding
with 20-30 characters — writing — approximately 4 000 years ago. The development of
text processing and printing technology in general in Asian countries using Kanji-writing
with several thousands of characters was very different from that in the western coun-
tries. After Gutenberg’s invention, text production used his technology for over 400 years.
The invention of the hot metal typesetting machine simultaneously with halftone photog-
raphy made setting up press operations an industrial process instead of a manual craft.
With the accompanying improvement in capacity of printing presses; the printing industry
as known today came into being. The most commonly printed products, such as the
newspaper found their present form with halftone pictures. Before then, newspapers had
existed for 300 years.
Table 3 shows that hot metal typesetting was commonly the only technology in
printing plants for almost 100 years. Computerized text processing and phototypesetting
rapidly replaced this technique in the 1960s.

Table 3. Development of text processing.

Technology Year
Talk coding — writing 2 000 BC
Manual typesetting 1450
Typesetting machine 1880
Phototypesetting 1920
Digital text processing 1960
Optical reading 1970
Speech recognition 1990
Intelligent text processing 1995

16
 Introduction

The development of the key technologies — image processing, text processing,

display technologies, printing, and finishing — have until recently occurred independently
of each other. Computers and computer linked peripherals, networks, and devices are
currently causing major changes. Besides computerized printing reproduction, the
changes involve integration of the key technologies into products and services°®.

4 Printed products
One classification of printed media can use their functions: advertising, news and
information, transactions, education, and entertainment. Other communication media,
such as radio and television broadcasting, cable television, and net media fulfill these
same functions. This generates a competitive situation. A more detailed analysis reveals
that the user situations are very different, although the functions are the same’. In many
cases, synergistic effects between different media may be evident. The level of media
use in western countries is already so high that time will limit the use of media.
Packaging falls in the category of printed media. Its printing has the functions of
advertising, delivering information, and showing data such as barcodes. Decorative
products such as printed tissue and wallpaper are products whose communication func-
tion is minor or nonexistant.
Each functional category of printed matter contains several types of products as
Table 4 shows. When considering revenue, the most important products in the graphic
arts industry are newspapers, magazines, sales catalogs, books, advertising products,
and business forms.

Table 4. Printed matter.

Function Reports and leaflets

Advertising Direct mail
Sales catalogs
Directories
Advertising leaflets
News and information Newspapers
News magazines
Directories
Books
Reports and leaflets
Catalogs
Transaction Securities
Business forms
Reports
Education Books
Directories
Manuals
Encyclopedias
Entertainment Periodicals
Comics
Books

17
CHAPTER 1

Everyone recognizes a newspaper, but it is less evident how to define newspaper

or another type of printed matter formally. The information structure and the physical
structure are the primary technical characteristics that differentiate the products. Fig. 3
shows a hierarchical representation of printed matter. For simplicity, the illustration only
shows the top three levels.

| al
Information structure Physical structure

Logical lL t
Contents aa Block Cover Finish
structure

Figure 3. Information and physical structure of printed matter.

The following sections discuss the main issues related to structure.

4.1 Information structure

The technical information structure first defines the type of contents, objects, or
information in the product. In printed matter, the alternatives are text, monochrome or
color graphics, and images. The information structure does not define the contents. The
contents could range from a description of floods in India to home gardening at the
Arctic Circle.
The second characteristic of information structure is the manner of organizing the
contents. This is the logical structure. Standardized techniques are available for defining
logical structure. These include the SGML “Standard Generalized Markup Language.”4
For instance, text, graphics, and images may be components of independent stories
such as news stories grouped under sections as in newspapers. A story may begin with
a heading and an opening with the text organized under subheadings.
The third concern of information structure is the presentation of the information
on the pages — the layout or spatial structure. The layout is a visual realization of the
logical structure and is the result of graphic design. To organize informational content on
pages of a product, the typography and the positions of the information require consid-
eration. These elements are the minimum design process. The use of graphical ele-
ments, image processing, and layered structures for design purposes transforms page
layout into art. Figure 4 shows that the contents may have a parallel, serial, layered, or
linked configuration.

18
 Introduction

Parallel positioning allows

the reader to extract information at Hypertext
a glance. Parallel page layout requires
a large page size such as used in
newspapers. In serial positioning,
the contents flow one after another.
A serial arrangement supports read-
ing in a linear fashion. A layered struc-
ture means visualization of two or Parallel
three types of information by present-
ing them on top of each other. Display
advertisements, sales catalogs, and
magazines are the main uses of lay-
ers. Linking is commonly hierarchical
so the end user may move from one Figure 4. Page makeup principles.
level of depth to another. Hypertext
and hypermedia refer to text and other information objects linked with one another. The
use of links in printed matter is limited compared with digital media. In printed matter,
tables of contents, indices, and internal page references are links. In display products,
the technical implementation of the link may be hidden from the end user. Hypertext and
hypermedia presentation of information support nonlinear uses of information such as
associative browsing.

7 Display

Factual and Contents

t—| > fe J ; P. t
creative data cox information pocure ES

> Print

Parameter Graphics [4 Structural Layout

data rules rules
S|

Design
concept

Ree Continuous |_]

objects and
ne images
scenes iG) g

Figure 5. Information structure.

Combining text, graphics, and images into an entity called a document requires
organization according to some predetermined rules of logical structure. The definition

19
CHAPTER 1

of a document is therefore a collection of data organized according to some logical

rules. To illustrate the meaning of logical structure further, Fig. 6 shows the highest hier-
archical levels of the logical structure of this book. The structural elements can subdi-
vide further into several hierarchical levels.

Book

Cover, title pages,

List of Table of Conversion
reverse title and Preface Contenis Chapters Index factors
contributors
foreword

Title page Body

name

Figure 6. Illustration of the term logical structure.

A document may contain structural layout without definition of the logical struc-
ture. Another possibility would be embedding it as conversion rules. The layout can also
be the result of an application program used in retrieving information. Accomplishing the
rules of layout is page makeup or pagination. The result is presentation of the informa-
tion as page data.
Consider as an example the previous illustration of Fig. 6. The rules of layout with
knowledge of the page size, the margins, and the one column layout determine the size
of the graphics, the space between the top of the chart and the previous paragraph, the
space allocated between the chart and the caption, and the typography of the caption.
Separation of the logical structure and layout is essential when using pieces of
text in different contexts because it preserves the integrity of units.

4.2 Physical structure

Considering physical structure, Fig. 3 shows that printed matter has three parts:

- the block comprising the pages that are the primary carrier of the informa-
tional contents

- acover that often carries printed identification data of the product

- finishing that binds the pages and the cover together.

Newspapers have no separate covers. In magazines, the materials of the first and
last pages may not necessarily differ from the other pages. In long-life products such as

20
 Introduction

books and directories, the cover has the functions of providing identification and protec-
tion. Durability and durability requirements are related to the techniques used to bind
the pages together and to the cover. Alternatives include stapling, adhesive binding,
and thread binding. The combination of hard covers and thread binding commonly pro-
vides the best guarantee of durability.
Publication frequency of a product with a specific name and the particular edition
are the main variables associated with publication. The following text reviews the char-
acteristics of the major groups of printed matter in these regards.
Newspapers are published regularly. According to a conventional definition,
a product published at least three times a week is a newspaper. Readers are more apt
to define a printed product as a newspaper by its appearance and contents. A newspa-
per may be printed as a single edition or as updated, customized, or segmented edi-
tions. Updated editions of newspapers revise the news content. Customization means
selecting some contents according to the geographical region of distribution. A seg-
mented newspaper contains entire sections that vary according to some predefined
variable such as geography or end user subscription. Newspaper production technology
is gradually evolving to make customized newspapers possible. Customization is espe-
cially easy when dissemination occurs electronically. End user driven customization
essentially converts the newspaper from a product to a service and changes its charac-
ter from an “agenda setting forum” to the “daily me.”
Magazines are more stationary products than newspapers. International maga-
zines can have different geographical editions. Mail order catalogs often have the same
publication features as magazines. Their publication is more apt to feature different lan-
guages.
Considering publication frequency, books are “single event” products with the
exception of serially published books. Because the demand of books is often difficult to
predict and may span several years, their printing commonly involves multiple unchanged
editions. Textbooks, encyclopedias, and similar books have printings as updated or revi-
sed editions. New printing technology facilitates printing books on-demand even including
single copies. Factors that contribute to on-demand production include the difficulty of pre-
dicting the demand and the needs of updating the contents regularly.
Business forms such as lottery tickets, lotto coupons, transportation tickets, etc.,
primarily used in business transactions and tax forms used in public administration com-
prise a diverse group of printed matter. Forms commonly have use for a long time. Print-
ing of new batches involves multiple unchanged editions or minor modifications as
updated editions according to the consumption. A form may also have use for a specific
occasion with printing as a single edition. An example would be a questionnaire for an
advertising campaign or survey. Printing the constant and the variable information
simultaneously in a form is occuring more often.
The comments about the diversity of business forms are also true for advertising
material. Examples of advertising material are direct mail products to consumers and
commercial business-to-business information such as client magazines published regu-
larly or irregularly. Publication and printing of brochures and outdoor advertising prod-
ucts usually occurs once.

21
CHAPTER 1

When the publication of printed matter is regular such as a daily newspaper or

the number of product names published on a single event basis such as books is large,
dedicated production capacity is necessary. This means the publisher is typically also
the “printer” who tends to the technical production. In the other cases, the functions of
publishing and printing are more distinct.

5 Production
This section gives an introduction to the main technical operations in the production of
printed matter. The purpose is to familiarize the reader with concepts discussed in more
detail in subsequent chapters.
Production consists of the following steps:
- information processing

- printing

- finishing.

Traditional terms for each step are “pre-press,” “press” and “post-press” respec-
7 66

tively. The names reflect the importance that printing has in the process. It also indicates
that printing capacity is capitally the most intensitive unit in the chain. The departmental
division in printing plants follows these same lines. With more and more integration of
the three types of operations into one computerized production line, the distinction will
gradually become obsolete. The generic name for the whole process is printing repro-
duction process. The term emphasizes the aspect of printed matter in production.
The information structure of printed matter comes in digital form. Text in comput-
ers has coded characters. Display devices — soft copy — or print — hard copy — feature
images of the codes, called fonts. Graphics are synthetic images such as bar charts.
Parametric data using computer programs that synthesize images generate the graph-
ics. Alternatively, graphics may be freely designed images. Continuous tone images are
representations of actual scenes and objects.
The technical measure in characterization of information objects is the amount of
data — bits, bits/character, bits/pix, or bits/mm?. A pixel is the smallest addressable
image element. The level of quality usually increases as the amount of data increases.
Data transmission and storage uses a compressed form to minimize the time and space
requirements. Sections 3 and 5 in Chapter 2 and Section 7 of Chapter 8 provide addi-
tional detail about data and information.
In the case of images representing natural scenes and objects, the first produc-
tion step involves capture by optoelectronic (commonly called digital or electronic) or
photographic cameras. In multi-color capture, registeration is by three sensors that are
sensitive to different colors. In film based photography, the sensors have light sensitive
particles in the film. Optoelectronic capture uses spectrally selective semiconductor
materials. Three signals to represent color occur in the first step. In display and printing
of images, color reconstruction uses three, four, or more primary signals. This means
that the signals from developed color photographs must be separated by color filters.
The operation is color separation.

22
 Introduction

In compression of image data to reduce the amount of data for transmission and
storage and in computerized processing, three or four color signals manipulate the color
images. The manipulations include adjustment of image darkness or tone, color, and
details. Most printing processes cannot render more than two tones, although photo-
graphs have a continuous tonal scale. Use of halftones creates the impression of tones
and colors by converting an image into a pattern of dots.
The page makeup step converts documents into page format. Following the page
makeup, the production of printed matter can use the alternative imaging steps indi-
cated in Fig. 7. Typical terms are the following:
- computer to film

- computer to plate

- computer to press

- computer to print

- computer to net to print.

Film exposure
and
development

Figure 7. Production steps for printed matter starting with page data.

In each case, the page data first undergoes decoding into a form understood by
the device on which the data will be converted in a two-dimensional form. This may be
an image setter, a plate maker, or a printer. Data decoding uses raster image proces-
sors (RIP). This term gave rise to the term ripping. In computer to fiim type production,
digitally stored pages — one for each primary color signal — undergo exposure on a light
sensitive material, film, or paper. A development processing step converts the exposure
to a visual image of the page data.
The next step involves transfer of the page data to a printing plate or cylinder in
another imaging step. The plate functions in the subsequent step as a stamping device
to transfer ink to the paper. In the oldest but still feasible printing methods, the latent
image on the plate is a geometrical relief. The image on the plate is not necessarily
visually discernible because the data are primarily present as another type of signal
than optical. Regardless of its type, it is a latent image. Making ready for printing
involves mounting the plates on a printing press. Printing with lignt absorbing printing

23
CHAPTER 1

inks converts the latent image into a visible image on the paper. The computer to film
route is still the most common procedure in industrial printing technology due to the
speed and quality of film making.
In computer to plate type imaging for printing, the film making step is absent.
Imaging of the page data from a computer memory occurs directly on a printing plate.
Computer to press imaging refers to a procedure where the printing plates undergo
imaging after mounting on the press.
Computer to film, computer to plate and computer to press techniques use con-
ventional, mechanical printing methods. As the name suggests, computer to print imag-
ing eliminates the film making and plate making steps. Computer printing methods
called electronic or digital printing facilitate this. In electronic printing, the address data
can be printed in the product with the content information. Electronic printing methods
fully integrate information processing and printing. They may even integrate finishing.
Computer to net to print means print production in a network environment such as the
Internet where page data to a printing device comes from a network linked computer
system or server.
As has become evident from the discussion above, the process tools in the pro-
duction of printed matter include computers, computer programs, and different kinds of
devices and machines. Until recently, the process equipment in printing production was
totally application specific and did not have use in other fields. It is evident that this
should be true of printing and finishing equipment because of the specific nature of the
respective operations. In the past, the large amounts of data in print production, the
requirements for high quality, and the frequent tight production schedules favored the
use of special processors and programs in the information processing steps. This is no
longer the case due to the performance development of computer systems with associ-
ated availability of general purpose programs. The development has brought more inte-
gration than before between printing companies and their customers who provide the
information for printing.
The information processing steps end with one digital copy of the product. The
production capacity requires scaling to deal with the information flow that arises from
the page data. Scaling of the printing and finishing capacity faces other problems than
the pre press steps. These relate to the physical structure of the product range and the
time schedules of production. In computer to print production, the amount of data per
page influences the speed of production because ripping is part of the printing opera-
tion. In this regard the issues of digital printing approach those of the information steps.
Printing processes are progressing towards reduced use of materials, as the
order of listing the above terms suggests. Film development and often plate making
require the use of chemicals. In terms of volume, paper is the major raw material in
printing. In comparison ink volumes are very small, and binding materials are smaller
yet. For turnover in the publishing and printing industry, the share of paper is approxi-
mately 10%, and the share of ink is 1%. Most materials are disposable meaning recy-
cling, down-cycling, or burning after use. Films are burned, plates are recycled or down-
cycled, and printed matter is recycled or burned.

24
 Introduction

ccc ==
1. Straubhaar, J. and LaRose, R., Communications media in the information society,
Wadsworth, Belmont, Ca, 1996, p. 490.
2. Anon., Electronic publishing: Strategic developments of the European Publishing
Industry toward the year 2000, European Commission DG XIII/E, Brussels,
Luxembourg, 1996, p. 444.
3. Structured Information/Standards for Document Architectures., et al., J. of the Am.
Soc. for Information Sci., Special Topic Issue, 48:7(1997).
4. Winston, B., Media Technology and Society. Routledge, London, 1998, p. 374.
5. Oittinen, PR, Saarelma, H., Huhtanen, H., Graphic Arts in Finland 26(2):3(1997).
3. Interactive Digital Media. The Impact of Technology to 2003. (Ed. jvan Oosterom).
EU Directorate General XIII/E 1997. p. 100.
6. Future of Paper. CAP Ventures, Marshfield Ma, 1997.

25
CHR
Imaging systems

1 Image Signals acwics.-ssaeeweage tee ccrstecore cece nae ccd he cement aen acer ei eeee enema eee 27
2 Image: metrics 22s: Bocce eee ras ercas cece cored pane beeeoes est aeer eeeeee cee eee 30
2ale + WN ISU MMAGE MCAS UES casscecyscacenutvs Got cre tartws ayes ce ee aoe at aoe e a ae, ee eee 35
3 Information: Capacity vec cc oe nk eeecetaee es coe ee seies eed ocean een aan eee 37
32lisey AFANSMISSION Sew ee ee ee ee ee eee 42
4 Digital:image: proGeSsSing =.ssives-sc-akascacchtessecscananaecuresd-naceses <uaenee eee eee 43
+) [MAGIAG SYSTEMS cccaetincnndiu cas -cxronetoieoes tremeerceancscnes ddzacsce se. cuyual atte case ernie 46
5. Wi MASVSTCIT TBODY 3cecgertes vA e vated wise Phe aden, cate na, a ead 46
BZ, — SYSTEMECNOIMGELING, cere-siecessensnsdivuecssacchasecsescs
teers eecnaan taeeeteat eae ee 50
RETCNGNICOS svecsssorcsasruseoevucsadioasit
innssenst stares: eatsucaaaeee turee gh setae eae ae 62

26
 Imaging systems

Iimaging systems
1 Image signals
The definition of an image used here is a two-dimensional intensity distribution of light.
According to this definition, a printed picture without external illumination is not an
image. The definition of an imaging system renders the visual sensation of original
scenes in a different place and time for communication purposes. Passive images
absorb existing white light to form desired colors. Figure1 shows that passive images
are prints or hard copies. In a light produced by a display system, active images may be
projections or displays. They are soft copies. For passive images, color formation is
subtractive. Color formation of an active image system is additive.

Display

|
Image intensity; Image intensity;
independent of |, constant percentage of |,
Figure 1. Contrast formation of an active image (display) and a passive image (print).

An image signal is a code or a character signal carrying image information. An

optical signal may be the image itself, i.e., light reflected from the actual object, a hard
copy, or a display or digital image code in the case of optical data storage or transmis-
sion. The human visual system can detect only optical signals. At the end of the imaging
process chain, the image signal must be optical. A coded optical signal such as an opti-
cal disc or optical fiber may be visible, but humans cannot understand it. The processing
of an optical signal uses optics and photography. Previously, all reproduction operations
for pictures involved manipulating the optical signal in several photographic steps.

27
CHAPTER 2

An analog or a video signal is an electronic signal in which the current or the volt-
age varies with the image illumination as Fig. 2 shows. Conventional televisions, video
recorders, still video cameras, older two-functional scanners (input and output simulta-
neously) use a video signal. Essentially, the processing of the video signal occurs “on
the fly,’ i.e., while transferring the signal at the speed of light.

= Ro]=
1st scanning line
One scanning line
2nd scanning line
3rd scanning line

Synchronizing pulses
AMPLIFIED
PHOTOCURRENT
Image plane

Figure 2. Principle of raster scan and corresponding video signal.

Digital television broadcasts digital signals. Similar to computer processing, the

video signal requires digitalizing. The digital image signal is primarily a two-dimensional
number matrix. The spatial division of a picture contains picture elements (pixels, also
called pels) with the darkness and color of each element expressed by a numerical
value. The division into pixels is sampling. The sampling frequency defines the image
quality level. Table 1 gives the pixel frequencies for some common imaging processes.
High quality output needs higher frequency to make the pixel structure vanish visually
because of higher constant level and lower noise level.

Table 1. Pixel frequencies for some imaging processes and necessary

frequency levels for visual quality.

Imaging process Pixels/mm

300 lines/in. laser (normal office quality) 12
600 lines/in. laser (high office quality) 24
Visual limit of newspaper printing | 33
Visual limit of high quality printing 50 ail
High quality scanner for printing 80 |
Top quality scanner for printing 160

Quantization is giving a numerical value to pixel luminance. In binary images

such as text and line drawings, a pixel has two values — white and black. This means
that one bit is necessary to quantize the darkness of the pixel. The length of the code
word for one pixel is one bit. For continuous tone pictures, pixels are often coded by one
byte or 8 bits. This provides 2° darkness level of i.e. 256 levels. Today 8 bits is insuffi-
cient for high quality printed pictures because of quantization noise — the error in round-

28
 Imaging systems

ing. Table 2 shows quantization noise for signal to noise ratio defined by Eq. 1 as a
function of the number of allocated bits. A noise level of 40 dB gives moderate image
quality, and 60 dB gives excellent quality. Human visual systems see the noise up to
approximately 50 dB’.

SNR,» = 20log (S/N) (1)

Table 2. Quantization noise.

Bits/pixel S/N ratio SNR, dB

14 23
28 29
55 35
890 59
3550 71
et
4200 83

Images are of the types shown in Fig. Shes

- continuous tone images or contones

- halftone images and

- line graphic images.

In acontone, darkness varies as a continuous function of the original scene. Pho-

tographs are typical contones. In black and white photographs, the amount of metallic
silver varies locally. Metallic silver as small grains is black and forms the image. Silver
grains are particles with a finite size providing a definite micro structure in a photograph.

rads @e-- | oe
See @@-- | | PN
“ Sea @®@ee B |

@@e-e i

Contone Halftone Digital image Line graphic

Formed of small Formed of dots Formed of Formed of

light absorbing with a varying picture elements continuous
particles relative area (pixels) 1-tone lines
Optical density The percentage of Each pixel has a
determined by dot area determines discrete density
grain density darkness or value
optical density

Figure 3. Different image types.

29
CHAPTER 2

The grain size varies slightly from one grain to another making the darkness vary con-
tinuously. When digitized, contones are coded on a pixel basis.
Most printing methods do not render continuous tones. They print a black or
colored surface or do not print. This is why continuous tone images must be halftoned —
screening is another name for the process — to obtain a halftone image. Halftone images
are formed of small halftone dots with a varying surface coverage. Dots cover locally
0%-100% of the surface controlled by dot size. This is called halftone percentage.
Graphics consist of two tone levels — white and black. The two main types of
graphics used in publications are line graphics or line pictures and font graphics or text
characters. Graphics may be coded on a pixel basis. Because of the two-tone structure,
only one bit/pixel is necessary. Font and pictorial graphics are increasingly coded as
vector data. Vector data models the image contours mathematically.

0 150. 200° 255 0 1 1 0

100 180 220 255 0 0 1 1
120 200 240 180 1 0 0 0
80 120 200 150 1 1 0 0

Digital contone Digital graphic of bit map

Figure 4. Digital images.

In a modern printing process, all images need digitizing. A digital image is a num-
ber matrix that contains the numerical luminance values of the pixels. This form is neces-
sary because normal computers can only process numerical data. When displayed, the
pixel structure and limited number of gray levels are often visible. This is why the term dig-
ital image is popular. Figure 4 shows digital images. Parameters of digital image are pixel
size and pixel depth. Figure 5 demonstrates their influence on image appearance.

2 Image metrics
Image metrics means representation of image properties as numerical values. Image
metrics quantifies product quality and controls automatic imaging processes. No
standard or specification fixes the quality of print as is the case with analog television
using PAL, NTSC, or SECAM standard. This is why the image quality level of print has
slowly increased during the years and has become an important competition factor
among different printing methods and among different printers.

30
 Imaging systems

NUMBER OF PICTURE ELEMENTS

200 x 200 100 x 100 50 x 50 25 x 25

LEVELS
GRAY
OF
NUMBER

Figure 5. Influence of detail and darkness resolution to image appearance.

31
CHAPTER 2

For images used for visual communication, the human eye and the human visual
system set the ultimate limit to the image quality. Improving the technical parameters of
the print increases the costs and improves its visual quality. Figure 6 shows that this is
only true to a certain point. After this point, the visual system can no longer see the
improvement due to reaching the resolution limit.
Measuring several physi-
cal image parameters is possi-
= a =F
ble. These can characterize a
single image or compare two
images. The former are univari-
ant measures, and the latter
are bivariant measures. Physi-
cal image measures quantify
the properties of the image sig-
nal, and visual measures quan-
o=
VISUAL
QUALITY
IMAGE
tify the image appearance.
Low High
Visual tests determine the
MEASURED IMAGE QUALITY
image appearance. For images
used in communication, the Figure 6. Relationship between visual and measured image
visual quality is very important. — quality in arbitrary units.
In an automated printing pro-
cess, the system operates according to physically measured image parameters. One
must be familiar with both and be able to define how a good visual quality forms and
how it relates to physical image measures.
‘ Local luminance and differences in contrast or darkness are the fundamental
measures of an image. Contrast formation is different for active and passive images.
Active images control display contrast with the illumination distribution of the display and
the surface reflection of the screen as Fig. 1 indicates. Due to the surface reflection, the
contrast of an active display depends on the light intensity in the environment. Passive
images or prints locally reflect a constant percentage of the incoming light. This means
that the contrast is constant in all illumination conditions.
As an example, assume that the “white” of an average print reflects 100% of the
incoming light and that the “black” of the print reflects 1% of the incoming light. Assume
further that the light intensity in sunshine is 10° arbitrary units. White parts of the image
then reflect 10° units and black parts reflect 10* units of the incoming light. The same pic-
ture will be visible in candle light where maximum intensity is 1 unit. White then reflects 1
unit and black reflects 10~* units. In both situations, “white” is visible as white and “black”
is visible as black although the actual black in sunshine is 10 000 times brighter than the
white in candle light. In both situations, determination of “black” and “white” uses contrast
not absolute intensity. This is due to the adaptation of the human visual system to the rel-
ative contrast. The darkness of a passive image relies on the relative intensity of light.
Equation 2 gives a definition for optical density abbreviated as density, D:

32
 Imaging systems

DENSITY .=_D.=,log 0
(2)
101

where I, is the incoming light intensity

I the reflected (or transmitted) light intensity.
With print, density is therefore constant in all illuminations. In display, density will
depend on illumination level as Fig. 1 indicates. This is the reason a computer display or
television screen is not visible in bright daylight.
The density value is 0 when reflecting 100% of the incoming light, 1 when reflecting
10%, and 2 when reflecting 1%. Measurement of reflection density usually uses an angle
of 45° relative to the normal surface in illuminution and zero angle indetection. Transmis-
sion density uses 0°. The measurement angles affect the density value. The intensity
value in the specular reflection angle relates to gloss. Both density and gloss values
depend on the measurement angles, the measurement aperture, and the size of the illu-
minated light spot in the measurement device. Density measurement does not have stan-
dard conditions. This means that the numerical values depend on the densitometer used.
Univariant density based measurements characterize individual images. Univari-
ant measures include the following:

- dynamic range

- detail reproduction

- noise.
Table 3. Dynamic ranges in density units for
some imaging processes.

Newspaper printing 0.8

Matt offset printing 1.0-1.4
Glossy offset printing 1.5-2.0
CRT display 0.5-2.0
Photograph on paper 1.5=2.0
Photograph on slide 3.0-4.0
Sunny day 6.0-7.0

The dynamic range indicates the density range of an image or an imaging sys-
tem. Table 3 gives typical dynamic ranges in density units for different imaging pro-
cesses.
The table clearly shows that technical imaging systems have very limited
dynamic ranges compared with natural scenes and that great differences exist in
dynamic ranges from one imaging system to another. When reducing the dynamic
range, an information loss in the image usually occurs. The loss depends primarily on
the imaging system. Improvements are available only by proper selection of devices and

33
CHAPTER 2

materials. An imaging system operator can nevertheless decide what information will be
lost and what will be visible in the final image. This process is image enhancement.
Detail reproduction is an expression for the resolution or resolving power of an
image’. By definition, modulation is the signal amplitude divided by the signal mean
value. Iffis a signal and M is the modulation, then

(Ger = Tite oP ie ae = ee (3)

ia Ceri: npStmine Tne F iby

In a typical image, the modulation decreases with increasing frequency (decreas-

ing detail size). At zero frequency, the modulation is one, and contrast at this frequency
is equal to dynamic range. The frequency at zero modulation indicates the ultimate limit
beyond which no detail reproduction in the image is possible as Fig. 7 shows.
By definition, the modulation transfer function (MTF) is the ratio of output image
modulation and input image modulation in an image transfer system

MTF = —24 (4)

Ve

M in

where M,,,iS output modulation and

M;, input modulation.

S
Z ig
High 5
Z High
Hi

:
=
:
=
a 5 Aliasing

= Low = Low
Low High Low Hig
FREQUENCY FREQUENCY
Analog process Digital process
Figure 7. Modulation as a function of spatial frequency.

MTF as an expression of detail reproduction can characterize all imaging pro-

cesses — analog and digital. The M7F of analog processes typically decreases as a
function of increasing frequency. In digital processes, aliasing or the increase of false
contrast will occur at critical frequencies. Aliasing, if it occurs, spoils the image quality
considerably.
Image noise means uncontrolled density variation. By definition, a signal to noise
ratio or S/N-ratio is signal amplitude divided by root mean square (rms) value of noise.
S/N-ratio is usually in decibels.

34
 Imaging systems

Equation 1 gave the expression for S/N ratio. If SVR = 100, then SNRyp = 40 dB.
SNR can be measured for the sub steps of an imaging process. The noise of the whole
process is then

Ne ANCLNS ENE (5)

MTF and SNR can be measured from test patterns and with proper accuracy
from real images also®.

2.1 Visual image measures

Although visual measures are subjective, numerical values can quantify them7—/2 37,
One can define and fix numerical values for an “average viewer’ and a “99% viewer
limit.” The first is a mean value obtained by normal people, and the second is a visibility
range for 99% of the population.
An image system designer needs knowledge about visual image measures to
answer questions such as the following:

- What is the largest pixel size that is not visible?

- What is the coarsest halftone that is not disturbing?

- What tone and color rendering give best visual image quality?

- What is the limiting rate for display refreshment to avoid flicker?

- How small are the colored details that can be distinguished from a map or a -
radar screen?
- How many gray tones can be visually distinguished?
- How small are the details (high frequencies) that can be visually
distinguished?
- How many different colors can be distinguished?
- How many different colors can be recognized, e.g., traffic lights?
The questions are numerous, and they depend on the application. Many have
contradictory answers. In color halftones, individual dots should be low-pass filtered by
the human visual system to even the tone. In a map that has high frequency color
coding, colors should be separately visible.
The following are divisions of visual measures of image:

- visual thresholds

- correlations

- optimum image quality.

35
CHAPTER 2

Visual thresholds are measured as just noticeable differences (JND). These are
the smallest visible density, color, or detail difference the human visual system can see.
Knowledge of the JND is necessary for determining the basic structural parameters of
an imaging system. These parameters are pixel size, screen ruling and halftone struc-
ture for printing, and refreshment frequency and scanning resolution for displays. The
JND depends on the test person and the conditions. The JND values for an average
viewer are available in the literature. Because JND values strongly depend on condi-
tions, generalized conclusions as engineering design models are not possible. A con-
trast sensitivity function usually represents the performance of the human visual
system. If C is contrast sensitivity, then C = 1/T where T is the JND. Contrast sensitivity
has a maximum as a function of frequency as Fig. 8 shows. The exact location of this
maximum depends on the illumination conditions as an example. At a normal reading
distance of 30 cm, it corresponds to a detail diameter of 1.8 mm. The human visual sys-
tem thus enhances details of this size. From a viewing distance of 3 m, the correspond-
ing detail size is 138 mm.
Correlations between
physical and visual image ee
measures allow construc- Ee Sa
tion of control systems. >
Some correlations are obvi- E 100
ous. Wider dynamic range, i 30
higher frequency response =
and lower noise level < 10
improves visual quality to a 5
certain point. The limits of S ©
the visual system occur at ‘
approximately 20-um detail 25 1 4 16
size and at approximately SPATIAL FREQUENCY, cycles/degree
50-dB signal to noise ratio. of perimetric angle
An increase of detail fre- Figure 8. Contrast sensitivity function of the human
quency in a given process visual system.
may deteriorate visual qual-
ity at much lower frequencies because of filling in. This is why data of image correlations
need measurement with visual tests from pictures of the process for the actual results.
Optimum visual quality means the best possible appearance of each picture as
defined by measurable image parameters. Operations for optimum quality are image
enhancement. Recent advances have facilitated finding optimum quality. Gray balance,
contrast, memory colors — colors that humans intuitively remember like human skin,
blue sky, green grass etc., sharpness, and noise level can be measured from actual pic-
tures and enhanced into the correct directions. Automatic image enhancement algo-
rithms exist.

36
 Imaging systems

3 Information capacity

Information processing is an important aspect of printing reproduction. Information is

not a material. This is why the technology has changed and will change more drastically
than the processing of something like wood or ore. Delivery of the information to the
recipients occurs in immaterial form using telephone, radio, television, computer net-
work, etc. In this context, the immaterial distribution methods are options for information
stored in a data base primarily prepared for use in a printed product.
Printing reproduction processes primarily handle information consistent with the
information and transfer system theoretical approaches. These are “engineering”
approaches — information is understood as data. The contents of the message have no
importance. Pattern recognition and image analysis based methods are for the moment
used to a limited degree. An example is optical text recognition.

Table 4. The information capacity of some imaging processes.

Imaging process Information capacity |

Newspaper printing, coarse halftone (20 I/cm), black and white 20-40 bit/mm?
Newspaper printing, fine halftone (40 I/cm), four color 600-700 bit/mm?
High quality printing, fine halftone (80 |/cm), four color 3 000 bit/mm?
Television, 40x50 cm? screen, black and white 16 bit/mm?
Movie film, slow, silver halide based 3 000 bit/mm?
Micro film, silver halide or diazo base 20 000 bit/mm?
Spectroscopy film, silver halide based 2 000 000 bit/mm?

A reproduction process from original material to print is analogous to a pipe. The

exact amount of data requires pumping into the pipe. The ability of different imaging
methods to carry information varies considerably. Table 4 gives some examples. Infor-
mation capacity gives an expression to the information that can transfer from an original
scene to the image.
The estimates in Table 4 are ultimate values for the silver halide based processes
and technical values for printing and television. An ultimate value means the value that
is limited limited by the imaging process itself and determined by the signal to noise
ratio of the imaging system. Equation 6 allows calculation of the information capacity in
bits/area for such a noise limited imaging system:

Joc Plog, +>) (6)

where ff is the spatial frequency

S signal amplitude
N noise r.m.s. valve

< = SNR is signal to noise ratio.

37
CHAPTER 2

I can be calculated at any frequency, but it reaches its maximum value at the fre-
quency where SNR is 1 as Fig. 9 shows”. This value represents the ultimate informa-
tion capacity of the imaging process.
The technical information capacity of an imaging process results when one
accounts for the influences of technical parameters on the information of the output
image. These parameters include the following (Fig. 9):

- pixel size of a digital image

- number of gray levels of a digital image
- screen frequency in a halftone image

- scanning and modulation frequencies in a display image.

The data, B, in bits/area for a digital image that has a pixel frequency of frix and
a number of gray levels per pixel of m is
ep
B = fyi, log, m (7)
Equation 7 shows that the sampling frequency, tiv strongly influences the data
flow of a digital image, and the influence of the number of gray levels is minor. Storing
images digitally often uses 1 bit/pixel (2' = 2 gray tones, line graphics) or one 1 byte/pixel
(2 = 256 gray levels, contones). For moderate quality contones, fewer gray levels are
sufficient, but top quality requires more gray levels of 10 bit/pixel or 12 bit/pixel.
As practical examples of data flow, consider the calculation of the data flow in
three A4-format (210 x 297 mm?) documents. The documents are a text page, a con-
tone page, and a line graphic page coded as pixel graphics.
Assume that an A4 text page contains 40 lines and 60 characters/line. If the cod-
ing is ASCII, each character is coded by 7 bits, and the 8th bit of a byte is a parity-bit.
The information content is then A4 text pages 2 400 bytes = 19 200 bits = 19.2 kbits.
For the A4 contone, a pixel size suitable for normal quality printing is chosen.
Assume that the sampling pixel is 0.2 mm x 0.2 mm = 200 um x 200 um, and 8 bits per
each pixel to store the density information. In the area of an A4 sheet of 210 mm x 290
mm, the number of pixels is (210 x 290 mm?)/(0.2 x 0.2 mm*) = 1 522 500.
The data flow is A4 contone = 1 522 500 bytes = 1.5 Mbyte = 12 Mbits.
A rule of thumb is that the sampling frequency of line graphics must be ten times
as high as the one used for contones. A more precise definition for the sampling fre-
quency comes from the sampling theorem. This states that the sampling must be twice
the highest output frequency to avoid aliasing. This is also called the Nyqvist theorem.
Figure 7 shows that aliasing is uncontrolled contrast behavior of high frequencies.
Although line graphics have a hundred times as many pixels per area than con-
tones, only one bit/pixel is necessary to store the information. Pixel diameter in this
example is 1/10 of that for contones, i.e., 0.02 mm x 0.02 mm = 20 um x 20 um. The
number of pixels is (210 x 290 mm®)/(0.02 x 0.02 mm?) = 152 250 000. The data flow for
an Aé4 pixel graphic is 150 Mbits.

38
 Imaging systems

9
ete a7 eS leo io
“a BlcGolkaaleomlealies
= | | Wesel
ep)
Se A ey ap | 2 ze
6 < te
a S || ¢ | IS] WS] &

3 N97 "643829
i 7
PRN CRM CE | ees

16 grey leveis
(S/N +1) = number of B(bits) = 6 x 5 x log, 16 = 120
correct grey levels
D (bits/area) = {7
Figure 9. Formation of information carrying capacity in an analog and a digital image AI0.

The examples show that the different types of document information need differ-
ent reservations of memory and transfer capacity. These are only examples, and the
chosen parameters and coding affect the data flow markedly. Image data can be com-
pressed. Contones are usually coded and compressed on a pixel basis while line graph-
ics can be coded in vector form. Vector coded line graphics need much less memory
space when the figures are simple. Very complicated vector coded graphics may reduce
the system capacity drastically. Contones can be compressed by a factor of 2—10 with-
out losing information. The exact compression ratio depends on the compression algo-
rithm used and on the density structure of the image to be compressed.
The raster scan principle is the usual method for output of images. This means all
the information must be in bit map format. Pixel based coded line graphics is a such
a bit map. Contones are halftoned and presented as bit maps. The text code or ASCII-
code is changed to bit map code by decoding the digital images of each text character
from memory. Text font or collection of text characters are stored in the output device as
pixel based digital images or as vector codes.
The transformation of different codes into a bit map uses a raster image proces-
sor (RIP). RIP understands different codes and page description languages. It functions
as an image system decoder. A RIP may be programmed — a software RIP, or it may be
a special image processing computer — a hardware RIP. Table 5 gives data flow over
area after ripping for some typical raster scan imaging processes. Data flow after ripping
gives expression to the data flow needed to control a raster scan output device such as
a laser printer, phototypesetter, or electronic printing press.

39
CHAPTER 2

Table 5. Data flow after ripping at different pixel resolutions.

Pixel resolution Quality level Pixel diameter Data flow

300 I/in. Normal office quality 85 um 140 bit/mm?
600 |/in. Good office quality 42 um [560 bit/mm?
1 200 I/in. Newspaper quality 21 um 2 200 bit/mm?
2 000 |/in. High quality print 12.7 um 6 200 bit/mm?
4 000 |/in. Top quality print 6.3 ym 25 000 bit/mm? =

Comparing Table 5 to data flow of coded information (Table 4) shows that amount of
ripped data is essentially higher. Ripping consequently causes the data to explode to a flow
determined by the output resolution. This is why ripping occurs at the end of the process.
The magnitude of data flow correlates with the processing time in pre-press but in
particular with the data transmission time. Data transmission in the production and dis-
tribution chain of a printed product uses three steps:

- Collection of the information to be published for the editors: The data flow is
normally 10-100 times the information that is ultimately published.

- Data transmission in the process: A typical example is transmission of pages

as a data base in newspaper or magazine production. The data flow is 1—-1.5
times the information to be published. The excess comes from structure,
layout, and transmission coding.
- Data distribution: This occurs as off-line transmission of papers and books.
On-line transfer as data is becoming increasingly more interesting. The data
flow equals the distribution times the information to be published.

The first step in designing a data transmission process for manufacturing and dis-
tributing a newspaper is to calculate the data flow of the paper. The structure of newspa-
pers is very stable. Assume that the different components in the newspaper used as the
example are the following:

- 50 pages, broadsheet format (total area 50 x 40 cm x 50 cm = 10 m?)

- contones 15% of the total area, halftone frequency 34 I/cm

- line graphics 15% of the total area

- total text 750 000 characters.

The sampling frequency of the contones is 0.2 x 0.2 mm? (= 200 um x 200 um)
and 256 gray levels/pixel or 8 bit/pixel coding as Fig. 10 shows. The data flow of the con-
tones before ripping is as follows:

1.5x10°x 10° = 37.5 Mbyte

NQEO?

= uncompressed data flow of contones before ripping.

40
 Imaging systems

: = 34mm!; a = 295 um

b = a/V2=200um
: = halftone frequency
= grid frequency

Figure 10. The calculation of the grid frequency from the halftone frequency for a 34 I/cm halftone.

For line graphics, the sampling frequency is 0.02 x 0.02 mm* (= 20 um x 20 um)
requiring use of two gray levels or one bit/pixel. The data flow for pixel coded, uncom-
pressed graphics is the same before and after ripping:

Sic 08 S10" = 470 Mbyte (9)

0.02 x 0102 ~8
= uncompressed data flow of line graphics.

The text will be ASCll-coded before ripping. Then 1 byte/character is necessary.

The text needs 0.75 Mbytes. The structure information and layout information require
approximately 0.25 Mbytes for a total of 1 Mbyte. The sum of the three components will
be the data flow in the newspaper used as an example

500 MByte = data flow of a 50 page newspaper before ripping. (10)

After ripping, all material is in bit map format. If the pixel size of the output device
is that of line graphics, the data flow is the following:

10x10" x10) _ 3195 Mbytes.

10 x 10° ye 10°
0.02 x0.02 x8 ee ay)

Table 6 summarizes the data flow. Compression should be “nonlossy,’ i.e., infor-
mation retaining for contones with a compression ratio of 5. Line graphics should be
pixel base coded and compressed with a ratio of 20.

41
CHAPTER 2

Table 6. The data flow of a 50 page broadsheet format newspaper before and after
ripping in uncompressed and compressed form.

Information type Code type Data flow ea Data flow 4

uncompressed compressed
Contones PCM/8bit/pix 40 Mbyte 8 Mbyte
Line graphics Bitmap 500 Mbyte 25 Mbyte
Text ASCII 1 Mbyte 1 Mbyte
Whole paper before ripping Mixed > 500 Mbyte < 40 Mbyte
Whole paper after ripping Bitmap 3 000 Mbyte 150 Mbyte |

In different steps of the production chain, data processing and transmission occur
within the pre-press system and between the pre-press system and other computer sys-
tems. The most critical points inside the process are the following:

- ripping

- capacity of local networks.

3.1 Transmission

The capacity of local networks is typically 10-100 Mbits/second (1.25—12.5 Mbytes/second).

The practical efficiency of a loaded network is one-half to one-fourth the nominal capacity.
Table 7 gives the capacities of two networks. Ethernet is a coaxial cable network commonly
used in the printing industry. FDDI is an optical fiber network.

Table 7. Transmission times for the newspaper data in a local area network (LAN).

Network type Ethernet FDDI

Nominal capacity 10 Mbit/s | on
Loaded capacity 0.5 Mbyte/s 5 Mbyte/s |
Transmission time for newspaper text (1 Mbyte) 2S 0.25
Transmission time for compressed newspaper contones (8 Mbyte) 1 16s 1.6s
Transmission time for whole paper (compressed, before ripping, 40 Mbyte) 80 s | 8s

Wide area networks (WAN) connect geographically separated computer systems

using public networks. The public telephone network serves as a modem link. The
capacity can be 50 000 bauds or bits/second. In digital telephone networks, the transfer
capacity (digitalized speech or certain other data) is 64 000 bauds. Hiring permanently
connected telephone lines between two computers is another possibility. Their capacity
is 2 000 000 bauds. These are wired connections between two computers meaning that
the actual data transmission capacity equals the nominal data transmission capacity.

42
 Imaging systems

Applications of WAN in newspaper production include the following:

- Sending information to an editorial department by using the nearest modem

links of low capacity.

- Transmitting pages from an editorial department to the printers using perma-

nently connected lines or specially made links using transfer by electromag-
netic radiation. The entire paper is usually bit map coded and compressed.

- Transmitting the paper to the readers in an electronic form. Readers can

make their own copy using a fax device or a laser printer or examine the paper
using a display. This process uses modern connections. Table 8 presents
transmission times for the newspaper used as an example.
Table 8. Transmission times for newspaper components in wide area networks.
1 baud = 1 bit/second.

Connection type Modem Digital phone Permanent

line
Capacity 33 600 bauds 64 000 bauds | 2 000 000 bauds
Transmission time for newspaper text (1 Mbyte) 4 min 145) ¢ 4s
Transmission time for newspaper contones (8 Mbyte) 112 min 17 min 32S
Transmission time for entire paper before ripping
(40 Mbyte) = 85 min 160s
Transmission time for whole paper after ripping
(150 Mbyte) = = 10 min

A primary task for a process designer is to produce the information flow charts for
the process. As the previous discussion indicates, the chosen resolution strongly affects
the data flow. Resolution determines the quality level of the output. Doubling the resolu-
tion increases the transmission and processing capacity requirements by a factor of four.
In an office document production system, the pixel resolution is increases from 300 //in.
to 600 I/in. For halftone frequency in offset printing, it increases from 34 I/cm to 70 I/cm.

4 Digital image processing

The term image operations refers to the image processing tools used to manipulate a
two-dimensional image signal or its coordinates. Image operations can apply to optical,
video, and digital image signals. Because most images today use digital processing,
this discussion only considers that alternative. Several textbooks have discussed digital
image processing 3/7.
A digital image is a number matrix in which the value of each pixel is 0-63 (6 bit
coding), 0-255 (8 bit coding), 0-1023 (10 bit coding), etc. In digital image processing,

43
CHAPTER 2

new pixel values are computed. The computational operations or image operations have
classification according to their mathematical functions to do the following:

- replacement operations

- look-up table operations (LUT)

- logical operations

- arithmetic operations

- special function operations.

A classification is also possible according to the image area needed for computa-
tion to do the following:
- point operations or mapping

- filter operations

- global operations

- geometric operations.

In replacement, the corresponding pixel of image B replaces a pixel of image A

(original). No computation is necessary. A LUT stores new pixel values to a table. The
pixel values of the input image are addresses to the table from which the values of the
corresponding output pixels are read as Fig. 11 shows.

0 0) 0 0 20 20 20 20 20

2 2 1 1 40 100 100 40 40

2 1 1 1 100 100 40 40 40

2 2 3 3 fj 100 Gi 100 100 100 #4100

Ses ea
Original LUT Processed image

Figure 11. The principle of the look-up table.

The LUT can adjust tone and color rendering of pictures for calibration or
enhancement purposes. Statistical process models often model printing processes
today. They store the input-output relationship as a LUT. The procedure is fast to com-
pute, but implementation of changes in the process such as new ink, paper, or press is
not simple. New experimental data is necessary.
Logical operations include and, or, nor, xor, and not. Binary images — two tone
images — need only logical image operations. Arithmetic operations include sum, differ-
ence, product, and quotient. Arithmetic operations can correct distortions due to input

44
 Imaging systems

device shortcomings and noise reduction. Special functions like trigonometric functions
are calculated as a series. Special function operations find use for such operations as
image rotation and halftoning.
Point operations use only one pixel of the input image to calculate the corre-
sponding pixel of the output image. The algorithm is

Spas ee 12)
where g_is_ pixel of the output image
ia pixel of the input image
a constant for tone rendering
b constant for the average darkness.

Point operations compute rapidly. They find use for the adjustment of tone render-
ing and for color correction.
In filter operations, an area of n x m pixels of the input image calculates one pixel
of the output image. If the filter is

hy hy h,

hy hs he (13)
hy hg ho

the output pixel, g45, corresponding to the input pixel, /45, is

845 = Ny f3q t+ Aofgs + ha fx6 + Mafiag + MsSas + Not46 (14)

thy fs4 t+ hefss + hots6

Figure 12 gives exam-

ples of a sharpening filter and
a noise removing filter. Con- oar | 11 1]
volutions in practical compu- =e | Tey aoe
tation are often solved as —1 -1 -1 ib bed
filter operation using point
spread function as the filter. Sharpening Noise removing
Bigger filters give more accu- filter filter
rate results. Computing time
increases rapidly with Figure 12. Filter types for sharpening and noise emoving.
increasing filter size. A rule of
thumb is that the calculation time is n x m times more than that of point operation, when n
x mis the number of elements in the filter.

45
CHAPTER 2

Global operations calculate one pixel of the output image from all pixels of the
input image. Image transforms like Fourier transform are global image operations. Solv-
ing convolutions in the Fourier domain takes transforms from the input image and the
point spread function by computing the product and taking an inverse transform to obtain
the output using Equation. Because the transforms are slow, not all imaging computers
can do the procedure in a reasonable time.
Geometric operations consist of two algorithms. One computes new locations for
the pixels, and the other interpolates the missing pixels to give a nondiscrete image.
Geometric operations provide image sizing and rotation and correction of geometrical
distortions caused by factors such as the lenses of the camera. Most geometric opera-
tions compute slowly.

5 Imaging systems
5.1 System theory
In a system analysis approach, the printing process is a black box with a given input
and a given output as Fig. 13 shows.
Such a system may make photographs, newspapers, or other documents. The
editor chooses or creates the input and wants certain output. The imaging engineer
designs a system between input and output and the user interfaces for the editor.
The modelling of
image systems is necessary
to predict the output when
Imaging
the input is known and to system
calculate the input when the
output is known. The former
task is typical ina reproduc- _ Figure 13. Printing process.
tion process — the appear-
ance of print is calculated. The latter task is typical in image analysis for medicine and
astronomy as examples.
The model may be any of the following:

- logical

- Statistical

- mathematical

- fundamental.

Logical models use obvious relations of input and output. In normal photography,
darkness increases with increasing exposure. In printing, darkness increases with
increasing inking level. Logical imaging system models are very crude and inaccurate.
They can be components of control systems, but they usually do not predict the appear-
ance of the output or input image.

46
 Imaging systems

Statistical models use measurements of the input-output relationships in the

imaging process. Tables of corresponding input and output darknesses and colors are
measured from test printed samples with interpolation or extrapolation of the input-out-
put relationship from these tables. Interpolation and especially extrapolation may give
improper results. If the process parameters such as paper, ink, or press change, the
measured tables are no longer valid. For printing, this type of experimental calibration
should be separate for each halftone dot and screen ruling, printing press, paper quality,
and set of inks. The result could still be completely different in different inking conditions.
Statistical models find wide use in practical calibration of output and printing processes
such as the popular (ICC) standard profiles for printing.
The approach of mathematical models assumes solution of the input-output rela-
tionship of the imaging process in a closed form by a mathematical equation®®%”. This
means it is possible to generate a system model. A system function characterizes
changes to the input in the process to obtain a given output. The following equation
shows the general form of the mathematical imaging system model:

g(x.y) = SIf(xy)] (15)

where _g(x,y) is output image
S(%y) input image
S() a general system operator.

Without knowing more specific information for the system operator S(), the gen-
eral model is not very useful in practical engineering work. The imaging system is linear,
if inputsf; andfo give the outputs g; and go, respectively, and inputf; + fo gives the out-
put g4+2. In a linear imaging process, the expression for the input-output relationship is
a convolution integral:
+00

ony a= i) ff(%1.y¥,)h(%,,x,y,,y) dx, dy, (16)

co

where (x,y) is the system function called point spread function, PSF.

The PSF expresses how much the image spreads in the process. It is an output
for an input with infinite intensity but no spatial dimensions. Equation 6 is commonly has
an abbreviated from:

g(x,y) = f(x, y)*h(x,y) (17)

Convolution can apply for printing with certain limitations . The convolution inte-
gral can be solved in a discrete form or through an image transform. Image transform
involves the representation of an image in a frequency coordinate system not in a spa-
tial coordinate system. Fourier-transform is a widely used image transform. A Fourier

47
CHAPTER 2

series representation of the image means that the expression for the image uses sinu-
soidial components. The one-dimensional Fourier-transform is as follows:

en = I ee ek: (18)

where -F() is a signal in the Fourier domain

HO) an image in the spatial domain
u frequency
oe the spatial coordinate
l2 —1.

The inverse Fourier transform is the following:

f(x) = [Roe Ph (19)

Each image has its own Fourier transform, and each Fourier transform can trans-
form back to image /Fig.14/. Image processing in Fourier domain is possible. In princi-
ple, one operation then concerns all equal frequencies or detail sizes. When the basic
input-output relationship in a linear imaging process uses a convolution integral such as
Eq. 1 as the model, the relationship in Fourier domain is a simple product:

GUY p= Uv) Cy) (20)

where G/(u,v) is the Fourier transform of g(x,y).

The system function is usually expressed as modulation transfer function (MTF).

MTF is the following:

+oo

pide = ||\ilGaye ee (21)

—co

where = [(x) is_ the line spread function of the imaging process.

The line spread function is a response of a one-dimensional input — a line with

infinite intensity but without width — in an imaging process. The line spread function and
the point spread function h(x,y) have the following relationship:

+co

I(x) = | h(x.y,)dy, (22)

—co

48
 Imaging systems

HALFTONE IMAGE AS ORIGINAL FOURIER TRANSFORM OF THE IMAGE

MASKING OF HALFTONE INVERSE TRANSFORM PRODUCES

FREQUENCIES CONTONE

Figure 14. Fourier transform of a haftone image, an inverse transform which produces a contone.

49
CHAPTER 2

The relationship of the usually measured system function, MTF, and the basic
function needed for computations of input-output relationship or point spread function
h(x,y) is therefore:

—2TiuUx
MTF = | [h(x,y,)e dxdy (23)
—co

The computation of imaging systems as a convolution gives correct results in

cases when the systems are linear. Some imaging systems such as optical systems are
linear. Some like the human visual system and many printing processes are nonlinear.
In the case of nonlinear systems, different system functions for different frequency
bands apply. The convolution approach does not consider the influence of noise on the
detail formation. This is a definite disadvantage.
The benefit of the convolution approach is flexibility to variation in process condi-
tions. If we have an imaging process with input f(x) and output g(x), the input-output
relationship for the process will be the following:

G(u) = (MTF)F(u) (24)

where G() and F() are the Fourier transforms for the output and input
MTF is_ the system function of the process.

When process conditions such as devices, materials, or adjustments change, the

system function will change. One can easily divide MTF into subcomponents for the
optics, the scanning input, the scanning output, the printing, and the paper. The input-
output relationship then becomes:

G(u)=(MTF, X MTF,...MTF,)F(u) (25)

where k_ is the number of steps in the imaging process.

When the process conditions change, measurement and changing of one MTF
component requires a new process model.

5.2 System engineering

In a properly designed process, the different functions balance without bottlenecks.
Table 9 gives a block diagram for an image system in general terms’. Real systems
such as a printing reproduction process, a television system, or a computer based
publishing system will divide into corresponding variables. The intent here is to list the
system variables to separate and specify them when analyzing or designing a system.
In this approach, processing means the processing of symbolic images.

50
 Imaging systems

Table 9. General structure of an imaging system.

Source ie Actual scene or an existing image illuminated by light

Optical system Projects a 2-dimensional image of the scene on image plane
Camera Changes an optical signal to an electronic signal
Channel coder Prepares the signal for the channel, storage, and processing
Source/sink coder Image restoration, enhancement, and merging
Channel/memory Transfers/stores the image signal
Channel decoder Undoes channel coding
Display/printing Display/prints the image

A source or an object can be a collection of physical objects such as an actual

scene or an image produced by another imaging system. It may be self-luminous,
reflect light from the surroundings, or obtain its light from the camera or the scanner.
The amount and dispersion of the light falling on the object have a profound effect on
the quality of the image to be produced. The best result usually occurs with diffuse — not
directed — illumination of sufficient intensity.
Light from the object
reaches the light sensitive
surface of the camera either
directly or through an optical
system!!97° The basic
function of optics in an imag-
ing system is often to project Object Lens Image plane
the three-dimensional (3-D) Figure 15. Definition of a basic optical system for imaging.
scene to a two-dimensional
dimensional (2-D) illumination distribution on the image plane. An optical system con-
sists of a lens that collects the light from the object and focuses it on the image plane as
Fig. 15 shows.
If the illumination of the object is L, light intensity, 7, on the image plane using
geometrical optics is the following:
vas ji
4 f2(1 +m)? (26)

where m_ is magnification (the relative change of line length in the system)

a the “f-number” indicating the aperture.
ie F/d
F focal length
d lens diameter.

oll
CHAPTER 2

For geometrical reasons, f > 1. Larger f-number means smaller lens diameter. In
most optical systems, lens diameter is adjustable for exposure control. Exposure, E, is
the product of exposure time, ¢ , and intensity, 7 , on the image plane. Then, E = /t. The
intensity distribution of the original image locally controls the exposure. Adjusting expo-
sure time and the f-number of the objective also control the exposure. If the input of an
imaging system is a photograph or similar picture, the relation between the input and
the output images in terms of optical variables is the following:
ee Cate E
E, = E (=) (7) 1 eee (27)
as |

where’ index 2 is_ the new exposure

index 1 the previous exposure
E the exposure
a the minimum density of image
b the maximum density of the image
y the gradation of the imaging system.

Camera here is a device that transforms an optical signal into an electronic sig-
nal. According to the second law of photochemistry, one photon absorption triggers one
molecule reaction in a light sensitive system. In an electronic camera, one photon
absorption generates a photocurrent equivalent to one electron. One lumen of “white”
light corresponds to a photon flux of approximately 1.36 x 10'° photons/s. Photocurrent,
I, in amps is then:

= 36 10Le (ge) 1 602105 (28)

where L_ is the luminance on the image plane, lumens

ge the quantum efficiency (ge < 1).

The charge on one electron is 1 602 x 10~'? coulombs. In a camera, a raster

scan method normally scans the image plane as Fig. 2 shows. The scanning produces
a continuous image signal whose photocurrent varies with the input image darkness.
When the scanning speed is constant and the signal is a function of time, it is called a
video signal.
Characterization of cameras uses their sensitivity, spatial resolution (the number
of raster scan lines and modulation frequency in analog cameras and the number of
light sensitive elements in digital cameras), signal to noise ratio (SNR), and tone and
color reproduction. Camera tube-type cameras have been used for decades. Digital
semiconductor matrix cameras are currently finding increasing use instead of tube cam-
eras. For scanning photographs, mechanical scanners have been used for decades.
A channel coder prepares the information for transmission and storage’2 1727,
A channel may be analog or digital. Originally, telephone, television, telefoto telephoto,
and scanner operated in an analog mode. Now, these may all operate digitally. A user

ay
 Imaging systems

does not necessarily see the difference in normal use, but the digital mode offers sev-
eral new features. Digitalization separates the video signal into signal elements with
each element is represented by a numerical value. The numerical values then receive
digital code words. If there are eight possible values for the signal element, the binary
codes can be 000, 001, 011, 111, 100, 110, 101, and 011.
A three bit long code word is necessary to express eight possibilities. The length
of the code words, p, in bits is p = log»m, where m is the number of possible signal ele-
ment values. An equally long code word for each possible signal value is the “basic”
method of coding. Coding of pictures often uses one bit/pixel (two tone levels for binary
images) or 8 bits/pixel = 1 byte/pixel (256 tone levels for continuous tone pictures). Cod-
ing of text usually uses a standardized 7-bits/character ASCll-code with the 8th bit of a
byte being a parity bit. The text code only consists of addresses to the text font. A text
font is a collection of the characters as digital images.
Data compression often saves transfer time and storage capacity. Data compres-
sion means coding digital information more effectively than using equally long code
words for every possible signal value. There are “lossy” and “nonlossy” coding methods.
The former allows loss of information, and the latter codes the information in a way that
saves bits. Compression coding methods are the following:

- character probability coding

- area probability coding

- image transform coding

- fractal coding.
Character probability coding uses the fact that different characters or darkness
levels of an image have a different probability of occurence. The characters occuring
often have short code words, and the seldom occurring longer words have a code word
of constant length for each character. The coding efficiency of character probability cod-
ing can be computed from the signal information. According to Shannon's information
theory, self information, /(x), for incident, x, is the following:
I(x) = —log,p(x) os)
I(x) = -Inp(x) (30)

where p(x) is probability of incident x.

Shannon defined /(x) as a 2-based logarithm (bit) and as a natural logarithm

(nat). The latter has not found wide use. Self information is a random variable. Its mean
value, H(x), is entropy:

H(x) = -Zp(x,)log,p(x;) (31)

53
CHAPTER 2

Entropy (bits/character or bits/pixel) is image dependent. Different original image

signals have different entropy. It expresses the maximum efficiency of coding using
character probability.
The first widely used character probability code was the Morse code for text
transmission. In this code, common characters are short combinations of the two signal
elements used, and seldom occurring ones are long combinations. For pictures, charac-
ter probability coding is not very effective. Character probability coding is a nonlossy
coding method.
Area probability coding uses the positive darkness correlation between adjacent
image pixels. The most often used area probability coding method is run length coding.
Instead of coding each pixel value, the length of the row having the same pixel values is
coded. This method is effective for binary images that have only two darkness levels.
The method is natural for pictures scanned into the system by means of the raster scan
method. Each scan line can be compressed on the fly if there is a buffer memory in the
system for few scan lines.
The positive darkness correlation of adjacent pixels is two-dimensional. This is
why two-dimensional area probability codes or block codes are more effective than one-
dimensional. Computing these requires storing the entire picture. Block codes are also
slower to code and decode than run length codes. Area probability coding is a nonlossy
compression method.
Image transform coding applies image darkness presentation in a mathematical
form instead of a darkness value for each pixel. An example of such presentation is an
image transform. Image transform coding involves dividing the image into subimages of
k x k pixels. Image transform — usually cosine transform — is taken from each block. If
the darkness does not vary inside the block, only the first term is applied. For more com-
plicated blocks, more terms are applied. Compression comes from minimizing the num-
ber of terms. The original image is regenerated by means of the inverse transform.
Image transfer coding is a lossy compression method. High compression ratios result,
but the method is considerably image dependent and provides less image quality.
Fractal coding means that other spatial components than pixels are the basis for
coding. The compression needs a code book for fractals or image elements. Each frac-
tal has a code word. In input, the fractals are measured, and only their codes are sent.
The receiver seeks the corresponding fractals from a similar code book and recon-
structs the image. When the fractals are sought with a “close enough” principle, fractal
coding is a lossy compression method. The result approaches computer graphics if the
fractals are large and few.
In vector coding, contours are coded as mathematical models instead of pixel
darkness. Vector coding is finding increasing use for text fonts and binary images. For
uncomplicated images, it is an effective compression code compared with pixel based
coding. With complicated images, vector coding may slow down the system capacity
because vector codes are slow to decode.
A coder does the coding, and a decoder does the uncoding. A source coder or a
sink coder is usually a device that receives an unprocessed signal from the camera and

54
 Imaging systems

processes it into a form suitable for further use. In analog systems such as an analog
color separation scanner, this step may be simple like adjusting the linear amplification
and band limiting the signal. In digital systems, source coding includes an analog/digital
transformation and sophisticated digital processing of the image signal.
A source coder is usually a programmable computer — usually the image processing
computer of the system. This is because most external — not system dependent — image
operations occur in this step. Even interactive operations are often possible in this step.
The channel transmits information. A channel is a wire path, a coaxial cable, or
an optical fiber system. It may also be a nonwire system using electromagnetic wave
propagation. Channels are “analog” or “digital.” This refers to the character of input and
output signals. Bandwidth in Hz, signal-to-noise ratio in dB, and attenuation in dB/km
describe the performance of an analog channel. Performance of a digital channel is
given as channel capacity in bits/s and error rate in errors/10® bits.
Memory stores information for further use. Memory can store analog signal as in
videotape or digital signal as with semiconductor memory, magnetic disc, optical disc,
and magnetic tape. The specifications of memory are capacity (bytes), writing speed,
and reading speed.
A channel decoder recovers the channel output. A channel decoder is channel
dependent but is becoming more standardized due to the open computer system philos-
ophy. Channel decoding is often used by means of the same computer as source/sink
decoding. Source/sink decoding essentially makes inverse operations to the coding. It
also does special operations to prepare the data for output. A raster image processor
(RIP) — a computer that converts the page description code to a bit map for the output
device — is a typical channel/source/sink decoder.
The final step in the imaging process is the display or preparation of hard
copy '* 142122 For display, a cathode ray tube (CRT) is common, although flat matrix
displays are finding more use. The main properties of displays are the following:

- nominal resolution (a x b pixels)

- number of bits/color
- refreshment frequency, Hz — how many times/s an image is shown.

A hard copy uses computer output with silver halide photography for high quality
prints and plain paper output, electrophotography, or ink jet for moderate quality output.
Several factors determine the quality of hard copy output.
The general image processing scheme of Table 9 applies to several different
imaging systems including analog and digital television, digital image processing, copy-
ing, electronic printing, and conventional printing. For image technology, conventional
printing is not a single imaging technology but a collection of several different technolo-
gies. For example, newspaper manufacturing typically may contain a large variety of
imaging technologies like manual operations, conventional photography, analog and
digital image transfer, video and digital image processing, photomechanical printing,

55
CHAPTER 2

electrophotography, and ink-jet and conventional printing. A modern electronic printing

system may include only digital image processing and electrophotography.
A light sensitive system is a system in which a physical or chemical change takes
place due to light exposure. The physical systems used in imaging processes are cam-
era tubes and semiconductor camera chips. These convert photon absorption into an
electronic current as Eq. 28 indicates. Amplification of the current forms a video signal.
Figure 16 shows schematically the construction of the camera tube. A semiconductor
camera such as a charge coupled device (CCD) is an image sensor matrix that has as
many image sensors as the image has pixels. The exposure modifies the charging of
these detectors for reading by a clock to form a digital or video signal. With video cam-
eras, both tubes and semiconductors are the light sensitive devices. Scanners primarily
use semiconductors for matrices and lines.
The variety of chemically oper-
ating light sensitive systems is Intensity
large? *3-49.32-33 They are silver halo- Deflection coils 42:
genide systems or organic photosen- (ices ena | on
sitive systems. Silver halogenide a ; Photo
systems use the light sensitivity of trae ale)
impure silver halogenide (AgBr, AgCl) beam
charge
grains in gelatin. This suspension is a
light sensitive emulsion. Light sensi- <<
tive emulsions are coated on a base of
paper or film. When exposed to light,
each silver halide grain forms some
molecules of metallic silver. The result
is a latent image. In the development Figure 16. Principle of tube camera
process, the grains forming a latent
image reduce to metallic silver. Metallic silver in small grains is black. This forms the
contrast into a black and white photograph. After development, the remaining unex-
posed silver halide dissolves as necessary to provide image durability.
The basic reaction in latent image formation is the following:

AgX + hy > Ag? + 5X, (32)

where AgX is. silver halide

hy a light quantum or photon
Ag® metallic silver
Xo a halogen molecule.

The ultimate sensitivity of silver halides is four photons/grain. This means four or
more photons are necessary to expose one grain. This correspondingly generates four
atoms of metallic silver in the grain. There are 10°-10'° molecules in one grain. Devel-
opment consequently causes the number of image forming atoms to increase by a fac-

56
 Imaging systems

tor of 10°. This is amplification of the photoreaction. The development reaction is a

normal redox reaction:

Aig act red > Ag° + ox (33)

For high gradation development, the only active development agent is hydrochi-
non. Continuous tone developers include besides hydrochinon, also metol or phenidon
as active ingredients. The development reaction of hydrochinon is shown in Eq. 34.

OH O
I
+ 2AgX OC +2Ag + 2X + pie) (34)

II
OH O
hydrochinon chinon

The fixing solution removes unexposed and undeveloped silver halide from the
photograph. The active solvent is thiosulphate. The fixing reaction is as follows:

AgBr+S,03 > complex — Ag* +S,0, + Br (35)

Stabilization can be a substitute for fixing. Stabilization destroys the activity of the
remaining silver halide. There are several stabilization processes intended especially for
paper prints. A stabilized photograph is good for temporary use but not for archiving.
After fixing, the photographs undergo rinsing to remove the remaining thiosul-
phate (S5027-). Rinsing is important for photographs being archived, because thiosul-
phate reacts slowly with silver and destroys the image.
Normally silver halide photography operates negatively. More emulsion exposure
forms more metallic silver for a darker image. Several direct positive methods are avail-
able:
- use of solarization area as Fig. 17 shows
- diffusion transfer development

- reversal development.

Solarization uses the property of the emulsion wherein a very heavily exposed
area starts to bleach due to additional exposure. Emulsions are consequently pre-
exposed, and the additional exposure and normal development give a positive image.
Another term for the diffusion transfer principle is the Polaroid-principle. In this pro-
cess, a negative and a positive image result simultaneously. The emulsion that is to be
negative is exposed normally and fed into development in contact with the emulsion that
will be positive. The exposed silver is reduced in the negative forming of the image. The
unexposed silver halide diffuses into the positive emulsion. This contains small grains of
metallic silver that initiate the reduction of the halide. A positive image then forms.

5/7
CHAPTER 2

Reverse development is a multi step development process for normal negative

material. The emulsion is first exposed and developed normally. Instead of fixing, the
metallic silver is dissolved. Then the emulsion is exposed evenly — the unexposed silver
halide is exposed — and developed. A positive image forms. Direct positive methods are
common for slides and in certain reproduction processes to avoid an extra imaging step.
Color photography also uses the light sensitivity of silver halide grains. Color
emulsions have three layers. One is red sensitive, one is green sensitive, and one is
blue sensitive. In development, dye replaces metallic silver. Each layer has its own dye.
Conventional silver halide photography still gives the best sensitivity and quality
results of the known electronic and chemical imaging methods. Its use is common for
the preparation of black and white and color originals and documents and the produc-
tion of text and pictures from the pre-press process for printing when optimum quality is
necessary.
In the group of organic photosensitive systems, photopolymers are the most com-
mon. Photopolymerization includes the following steps:

- A UV-quantum absorbs in an initiator molecule causing radical formation

called initialition

R+hv oR (36)

- The radicalized initiator reacts with the monomer donating an electron

R +M—>RM (37)

- The polymerization occurs

RM +M-—RM, (38)
RM,+M—>RM,,, (39)
- Polymerization terminates spontaneously when two radicals react

RM,, + RM, > RyM on « m) (40)

The amplification of photopolymer reactions is 10°-10°. Several image forming

molecules form corresponding to each quantum reaction. Common uses are for printing
plates and as a curing mechanism in printing ink (ultraviolet and electron beam curing
inks). The benefit compared with the silver halide reactions is good mechanical strength
(important in printing plates where the hardened photopolymer resin transfers ink). Its

58
 Imaging systems

disadvantage is slower sensitivity. Photopolymers always operate negatively — exposure

hardens the layer.
Silver halides and photopolymers have a high amplification. A wide variety of
chemical compounds are light sensitive but have no amplification. This means that the
whole reaction occurs when exposed without development or spontaneous polymeriza-
tion. Typical of these reactions is very slow light sensitivity and very high resolving power.
Sensitometry con-
cerns the measurement of
the properties of light sensi- Saturation ak,
tive systems. The basic con- Solarization;
density decreases
cept of sensitometry is the as a function of
characteristic curve of den- exposure
Linear tone
sity vs. logarithm of expo- rendering
DENSITY
sure for a light sensitive
system that Fig. 17 shows.
Sensitometric proper-
ties primarily provide an
LOG EXPOSURE
expression for the effect light
causes in a material. This Figure 17. A characteristic curve of a light sensitive system (silver
may be a change of dark- halide photography).
ness, a change of color, ora
change of physical or chemical properties of the material. The procedure to determine
sensitometric properties includes exposure of the light sensitive material, its develop-
ment, and possibly other postprocedural operations like fixing and measurement. Sensi-
tometric properties are not physical quantities like length and weight but results of the
testing procedure in given conditions.
Sensitometric properties were first understood as the parameters of characteris-
tic curves. Parameters of the curve are now macroscopic sensitometric properties. Mac-
roscopic sensitometric properties can compare different light sensitive materials, but
they have no use for predicting the darkness structure of real pictures. This is because
darkness formation depends on the local exposure and the exposure in the neighbor-
hood — the spatial structure of the image. This is why microscopic sensitometric proper-
ties or rendition of small details also require measurement. Modulation transfer function
is a microscopic sensitometric property.
Speed or sensitivity of a light sensitive system expresses the amount of light
needed to obtain a defined density level. Numerically, speed is the inverse of the expo-
sure (ASA-standard) or its logarithm (DIN-standard) required at the given density level.
The measurement procedure is very time consuming and complex. Photographers
specify sensitivity by ASA or DIN speed and use the numerical value for predictions of
the needed exposure. In the case of professional films and light sensitive papers, expo-
sure calculations use the test exposures indicated in Eq. 26 and 27. Speed is usually
not a specification.

59
CHAPTER 2

Gradation means the steepness of the characteristic curve. Table 10 shows the
grades of professional films and papers.

Table 10. Grades of some light sensitive materials.

Light sensitive material = Gradation

Tone film OSs:
Tone paper 0.6-1.4 ~)
Phototypesetting film 1-2
Phototypesetting paper 18
Page film 9)
Paper for halftoning 5
Film for halftoning 10-20

Spectral sensitivity expresses the wavelength range that exposes the material.
Table 11 provides classifications according to the spectral sensitivity.

Table 11. The groups for spectral sensitivity.

Sensitivity in wavelength of light

U\V-sensitive < 400 nm |
Blue sensitive < 500 nm
Orthochromatic < 600 nm
Panchromatic 400-700 nm
IR-sensitive max. 800-1300 nm

Photography of an actual scene uses panchromatic materials. The sensitivity dis-

tribution of panchromatic materials corresponds to that of the human visual system. In
technical image exposure devices, the spectral sensitivity of the material should match
the light quality of the device. Most image exposure devices today use lasers as light
sources. Table 12 gives the peak wavelengths of some commonly used lasers. Section
1 in Chapter 3 defines the color temperature area.

Table 12. The energy of some commonly used light sources.

Wavelength range E: Color temperature

Tungsten lamp 400 nm to IR 2800-3200 K
Xenon lamp [ UV to 700 nm 5000 K
HeNe-laser 632.8, 1152, 3391 nm -
Art and Kr* -lasers Several peaks 333.6-799.3 nm | -
Semiconductor laser Several peaks -

60
 Imaging systems

10%
N
=
Spl ee
=

€ 10 Ag-emulsions
>
O resolution Ss
c 4 Ag-paper = Ag-emulsions D
o
i Electro- 2 S
Lu 2 photography =
ao 10 a
=) 1S DB

a 10 £e
it Photopolymers om
ets
6 a2
10 Cr+ and diazo-compounds es
jm
10 10° 10
RESOLVING POWER, mm_
Figure 18. Photographic speed as necded energy and corresponding resoluing power of some commonly
used light sensitive materials
22.

The resolving power expresses the ability of a light sensitive system to render
small details. In the case of silver halogenide materials, the granularity of the image lim-
its the resolving power, because the image is formed of small grains of metallic silver.
Particle size correlates with the sensitivity of the material — bigger particles give better
sensitivity or higher speed but the resolving power remains poor. Figure 18 gives the
ultimate performance of light sensitive systems and the correlation between resolving
power and sensitivity.

61
 CHAPTER 2

MiReferencecs
1. Saarelma, H., Oittinen, P, Kuvatekniikan perusteet (Basics of Image Science)
Otakustantamo 842b. Otaniemi 1986. p. 101. (textbook, in Finnish)
2. Schreiber, W., Fundamentals of Electronic Imaging Systems. Springer Verlag. New
York 1986. p. 187.
3. Saarelma, H., Perila, O., Virtanen, J., KEMIA-KEMI 8(3):92(1981).
4. Oittinen, PR, Saarelma, H., Kuvatekninen laatu (Image Quality), Otatieto 882.
Otaniemi 1992. p. 141. (textbook, in Finnish)
5. Dainty, J. C., Shaw, R., Image Science. Academic Press. London 1974. p. 402.
6. Katajamaki, J., Saarelma, H., Measurement of Print Quality Potential from Color
Images. NIP 12: International Conference on Digital Printing Technologies. San
Antonio, Texas 1996.
7. Anastasi, A., Psychological Testing. The Macmillan Company. New York 1962.
p. 657.
8. Guilford, J.P, Psychomatric Methods. Mc Graw Hill. New York 1954.
9. Marr, D., Vision. Freeman. San Francisco 1982.
10. Pinker, S., Visual Cognition. The MIT Press. Cambridge 1985.
11. Oittinen, P, Saarelma, H., (8):1987(1991).
12. Saarelma, H., Kuvan ja tekstinkasittelytekniikka (Picture and text processing).
Otatieto 578. Otaniemi 1997. p. 172. (textbook, in Finnish)
13. Andrews, H.C., Hunt, B.R., Digital mage Restoration. Prentice Hall Inc.
New Jersey 1977. p. 238.
14, Budford, J.F., Multimedia Systems. ACM Press. New York 1994. p. 451.
15. Conzales, R.C., Wintz, P, Digital Image Processing. Addison Wesley Publishing.
Reading, Mass. 1987. p. 503.
16. Dougherty, E.R. (ed), Digital Image Processing Methods. Marcel Dekker, Inc.
New York 1994. p. 472.
17. Pratt, W.K., Digital Image Processing. John Wiley&Sons. New York 1978. p. 750.
18. Oittinen, PR, Saarelma, H., TAPP! 65(2):47(1982).
19. Anagnostopoulos, C.N., Lesser, M.(eds), Cameras and Systems for Electronic
Photography and Scientific Imaging. SPIE Proceedings 2416. California 1995.
20. Dean, PJ., Optoelectronics. Prentice Hall Inc. New Jersey 1983. p. 445.

62
 Imaging systems

21. Saarelma, H., Oittinen, P, Basics of Printing Technology. Otatieto 563. Otaniemi
1995, p. 235.
22. Anon., Proceedings and Seminars of SID 28th Annual Conference. May 1997.
23. Becker, H.G.O., Einfuhrung in die Photochemie. VEB Deutscher Verlag de
Wissenschaften. Berlin 1983. p. 511.
24. Bottcher, H., Epperlein, J., Moderne Photographisce Systeme. VEB Deutscher
Verlag fur Grundstoffindustrie. Leibzig 1983. p. 330.
25. Diamond A.S., (ed), Handbook of Imaging Materials. Marcel Dekker. New York
TIDIA D625,
26. James, T.H., The Theory of the Photographic Process. The Macmillan Company.
New York 1977. p. 750.
27. Katz, J., Fofel, S., Photographic Analysis. Morgan&Morgan Inc. New York 1971.
p. 658.
28. Kosar, J., Light-Sensitive Systems. John Wiley&Sons, Inc. New Yorks 1965. p. 473.
29. Shannon, C.E., Weaver, W., The Mathematical Theory of Communication. Univ. of
Illinois Press, Urbana, 111. 1949.
30. anon., Proceedings of [S&T 50th Annual Conference. May 1997.
31. Nasanen, R., Visibility of Halftone Dot Tectures. IEEE Transactions on Systems,
Man and Cypernetics SMC-14(1984), pp. 920-924.
32. Oittinen, P, Saarelma, H., Graafinen Materiaalitekniikka (Material Technology in
Graphic Arts). Otakustantamo 493. Otaniemi 1987. p. 215. (textbook, in Finnish)
33. Saarelma, H., Advances in Printing Science and Technology, ed. W.H. Banks.
Pentech Press, London. 15:116(1980).
34. Shannon, C.E., The Bell System Technical Journal XXVIII(3):379(1948).

63
BOAT
Color and color images

1 Lightand Golotincsis cia a cicecncen tie tuatpnewecateeereameass teed ia rs ere ae eee 65

2 Human Visual SYSt@M .i...cc0cecsacccerescoascund-ces0cuneencrs-decasacccssncseaneeeccmerseeea 68
3 Measurementiof:colors 22sec eee aoe ted nc occ oeeca ee see een at ener tenes 73
4 Reproduction of CONT 20 22erencdecaeenaacesnoe tnenaetecpasecaee dea. acs. cated. need 79
AD se BaSICsDUINGIDICSEc at exsatt om netenniyaa evesvanrals baste atee ace anual ek ee eee 79
G2, GOIOMSED Ala tlOU dca: craven tseicusdes so tcacntenacbtsucatscactcceaes
dct taedace aesace icra steak eee eee ee 86
ABP WCOOTCOLECTONM tect cetotecevecetecvaeraa eet dnc tate ce eer ea aac ieee ee 88
4A SeScreening: (halftoning) tek ceo...ttendc rca estede octets: ¥ tectreeane Se Akcreas esaeeeee Se ee eee en 91
Ab Pye Restorationandemharie Mest <vcc.s-vcczoeccacesane cesssnseeeet scteehee eccaueececeaee omen a me 96
5 Color tormation ee ieosso.-cacoceucaccotncascceacactacascnccusssuersseu@ecter ssae eee 105
BA COlOm Cal Raton rterecccrt cert cate be carseat cette nate see ee 105
52” seulOVICCIMECDET COME COLON Sete: oe. cti cc fod “ascites eden al decease 109
6 Text processing and page make up ..................ccccecceececceceeceeseeseeceeeseusenees 111
ROLE FENCES caries tort Nias caG we iam oases ean da etait nis a Pac eh 116

64
CHAPTER 3

EColor and color images

1 Light and color
Light and certain wavelengths of electromagnetic radiation cause a response in the
human visual system. We can see these wavelengths in different colors’. Discussion
of light and color can have different aspects:

- The psychological nature of color or how the human visual system operates
from a psychological point of view

- The physiological nature of color or how the human eye and brain operate on
the mechanical and fundamental level for seeing and understanding color

- The physical nature of color or how the physical properties of electromagnetic

radiation correlate with the sensation of color

- The reproduction of color or how to generate an illumination distribution tech-

nically that gives the same color sensation as the original object.

Visual color raises two basic questions:

What is light?
What is color?

Physically, light is
electromagnetic radiation in
a wavelength range of app-
roximately 400—700 nm. One IR radiation Visible red
nm is 10-2 m = 10°° mm = radiation

0.000001 mm. A body usu-

ally emits light because it is Room
Approximately 800°C
hot. In room temperature, femoperaiure
the heat radiation is not visi-
Yellow light 5000°C
ble. It becomes visible at a (tungsten lamp) white light
temperature of approxi- (gas discharge lamp)

mately 800°C as Fig. 1

shows. When body tempera-
imately 2500°C 18000°C
blue light
ture rises, the wavelength of — “PProximately 2
the emitted light becomes
shorter, and the energy of Figure 1. Light generation as a function of body temperature.
single photons increases.

65
CHAPTER 3

Most light on earth comes from the sun. The surface temperature of the sun is
approximately 6 000°C, and the light emitted by the sun is neutral white for the human
visual system. The relation between body temperature and the visual color of light is a
measure for light quality.
Color temperature is by defi-
nition the real temperature of
an object emitting light of a
3 000K (yellow) known color.
Color temperature uses
degrees Kelvin (K). The con-
version is that O°C = 273.16
K, and K = —273.16°C. Fig-
5 500K (white)
EMISSION ure 2 shows the spectra for
18 000K (blue) some color temperatures.
Color temperature is a com-
mon measure for light qual-
400 500 600 700
(blue) (green) (red) ity. It operates very well
where the light actually origi-
WAVELENGTH OF LIGHT, nm
nates from a hot body such
Figure 2. Color temperatures and the corresponding energy as the sun or a tungsten
spectra. lamp. No solid materials can
survive heating to tempera-
tures higher than approximately 3 000°C without melting them. This is why white light
with a color temperature of approximately 5 500K requires artificial generation with gas
discharge lamps. These generate an uneven spectra with peaks as Fig. 3 indicates.

=So

LW
cm
nm
50
at
cm
2
SPECTRAL
IRRADIANCE

WAVELENGTH (nm)

Figure 3. The actual spectra of some light sources.

66
 Color and color images

Characterization of spectra by color temperature may lead to wrong conclusions.

Figures 4 and 5 illustrate the spectrum of electromagnetic radiation including visible light.

RELATIVE
SENSITIVITY
OR EYE

UV Radio waves
X-rays

400 S155 700

WAVELENGTH OF ELECTROMAGNETIC RADIATION, nm

Figure 4. The sensitivity of the human visual system to electromagnetic radiation with the highest sensi-
tivity obtained at the wavelength of 555 nm (yellowgreen light).

400 500 600 700

Figure 5. Visual color or hue as a function of the wavelength of light.

The quantum and wave approach

are physical ways to deal with light. The Light source
emits radiation
Brain
quantum approach assumes that light interprets
consists of small particles — photons or <a)signal
light quanta. A photon is the smallest unit
of light. At photon level, light behaves receives
discretely. The photon approach analyzes signal
the behavior of interactions between light
and material in light sensitive systems
and in image formation. The wave Surface
modifies
approach assumes that light propagates light
as waves. This is the basis for geometric
Figure 6. Visual color formation.
optics.

67
CHAPTER 3

Short visible wavelengths of light of approximately 400-500 nm are visually blue,

medium wavelenghts of approximately 500-600 nm are visually green, and long wave-
lengths of approximately 600-700 nm are visually red. No physical explanation exists
why blue looks like blue and red looks like red. Figure 6 shows that visual color is not a
physical quantity but a visual sensation.
Color therefore has the following components:
- The color of a light source means that it radiates light of some color.
- The color of an object means that any object absorbs a part of the incoming
light and reflects the rest. (An object’s selective light absorption therefore
determines the color of an object or the spectral constitution of the reflected
light.)
- Visual color means that the human visual system understands different wave-
lengths of light as different colors.

- Visual color is not a physical property. (It exists only as the sensory perception
of an observer.)

2 Human visual system

This chapter will introduce
the reader to those aspects Light sensitive ,
of human vision that are retina, 1.3 x 10
essential for understanding
and designing man-made
image processing systems
10° connections
and the visual quality of to brain
images intended for
communication@/?. Figure 7. The construction of the human eye.
Studying human vision can
follow two approaches — psychophysical and physiological. The former uses a visual
test. The visual system is a black box, and the test person is asked to tell what is seen.
The limitation of visual testing is that the test person usually answers “yes” or “no.” This
is because the human visual system is accurate for comparison but not for
quantification of differences or levels. The psychological data obtained from visual tests
depends on the test person and conditions of the test. This is why the results require
Statistical analysis. Visual testing is time consuming, and exact numerical values usually
only apply in conditions identical to the test conditions.
Physiology studies the structure of the visual apparatus. Physiological data tries
to explain at a basic level the mechanism of seeing in the eye and the brain. A man-
made camera is a copy of sorts of the human eye. The same terminology can character-
ize the eye to some degree. The ultimate performance of the human eye sets limits on
the necessary image quality in visual communication. Unfortunately, psychological and
physiological data do not always match probably due to the adaptation of the human

68
 Color and color images

visual system. A short explanation for the adaptation is that “average seems to be nor-
mal” or humans accept average conditions as neutral. The highest illuminance therefore
looks like white in all conditions, and tinted color such as a yellow tungsten lamp seems
white without a reference. In addition, the mechanisms of color vision and noise vision
are not completely understood today.
For an imaging system engineer, a basic knowledge of the ultimate performance
of the visual system is important for planning the structural parameters of the techical
systems. The methods for visual testing must also be known.

-BLIND SPOT
Rods (number)
----- Cones (number)
—-— Relative visual 1.0
acuity in degrees

TA from the fovea

N us °
v1
-

acuity
visual
Relative

So:¢
- Sos,

rods
of
mm
Number
sq.
area
an
in
cones
or

Degrees from the fovea

Figure 8. Distribution of light receptors on the retina of a human eye.

The human visual system consists of the eye and a part of the brain. Figure 7
shows the structure of the human eye.
The retina is the light sensitive surface of the human eye. It consists of 1.3 x 108
light receptors. The two types of light receptors are cones and rods. Because their use
is for night vision, the sensitivity of rods is high. Their maximum sensitivity is Ama, = 510
nm. When the illumination level is low, only the rods contribute to vision. Then the envi-
ronment appears black and white because rods do not perceive colors. Moviemakers
are aware of this phenomenon. In night scenes, the picture of a color movie turns to
black and white. The viewing conditions would do the same at a proper luminance level
but probably not at the correct point. In normal television or computer display viewing
conditions, the luminance level is always sufficiently high to keep the colors visible.
Cones have low sensitivity because their use is for day vision. The maximum sensitivity
iS Amax = 556 nm corresponding to a yellowgreen light. Figure 8 shows the distribution
of cones and rods on the retina.

69
CHAPTER 3

The density of cones is maximum near the optical axis of a human eye or a ret-
ina. Vision is sharpest in this region. The proper design of imaging systems considers
this maximum performance. The visual angle in a normal person is 180°. Due to the low
density of receptors, images are unclear at high angles. Humans are aware of move-
ment but do not see details. Movie technology sometimes uses this characteristic. The
maximum density of rods occurs some degrees from the fovea. They obviously are for
detecting movement rather than sharp details.
Cones see colors. The three different types of cones have their maximum sensi-
tivities at the blue, green, and red areas of the spectrum. Any visual color will result by
mixing three colored stimuli. These may be colored lights such as in television or col-
ored dye stuffs such as in photography and printing.
In the case of active images, mixing of colored lights provides the desired visual
color. Figure 9 shows the subcolors are red, green, and blue — abbreviated as RGB. In
the visible wavelength spectrum, blue corresponds to short waves (A = 0-500 nm),
green corresponds to medium waves (A = 500-600 nm), and red corresponds to long
waves (A = 600-700 nm). Color formation in an active image is additive. With no light,
the visual result is black. A red, green, or blue light source alone gives a sensation of
the corresponding color. With simultaneous red and green light sources, the observed
color will be yellow. Blue and green produce visual bluegreen or cyan. Blue and red pro-
duce violet or magenta. Figure 10 shows the combination of colors abbreviated as CMY
(called CMYK when black is added). All three light sources simultaneously give an
emission of the whole visible spectrum, i.e., all wavelengths. This is visible as white.

507

REFLECTION,
0 ] ]
400 500 600 700
WAVELENGTH, nm

= ro)ro) 1
1007 el
% %o
Zz Za
ie) ie)
5 50+ 5 50+
cr Ww

ui Wi a
Ww Ww eee
0 | | ) |
400 500 600 700 400. 500 600 700
WAVELENGTH, nm WAVELENGTH, nm

Figure 9. Visual colors of RGB and corresponding emission spectra using RBG for display.

70
 Color and color images

91007
3s
mz ee
ie)
©& 50+
nm
—
LL
Wi
ae
0)
400 500 600 709
WAVELENGTH, nm

o 100 7 1007
o- x
x4 ee Zz eae oe

o
Ke
fe
—

© 507 fF 50
a mf
= =|
TT iL
= Ww
as ia
Ort + + { (0) + + {
400 500 600 700 400 500 600 700
WAVELENGTH, nm WAVELENGTH, nm

Figure 10. Visual colors of CMY and corresponding emission spectra using CMY for printing.

The possibility to produce any visual color by mixing three different stimuli is a
result of the three spectrally different cones in the human eye. Technical color reproduc-
tion is consequently not an exact reproduction of the coloring of the original object. It is
a fake for the human eye causing the eye to produce the same visual sensation as the
original scene.
The property of the human eye to obtain the same visual sensation from spec-
trally different stimuli is metamery. A big advance from point of image technology is
the reproduction of color images with reasonable costs. In real objects, there are hun-
dreds of dyestuffs generating thousands of different spectra. If the human eye could
selectively see different wavelengths, color reproduction would be practically impossi-
ble. With sound reproduction, reproduction of the pressure wave exactly matches the
original. That is an easier task than reproducing an exact color spectrum. Taste and
smell are molecule selective. Their reproduction in a technically simple way is not cur-
rently possible.
The intuitive definitions of color are red, green, yellow, etc.. Why the visual sensa-
tion of different colors is so different is not known. The only physical difference between
colors is the different distribution of wavelengths. No methods are available to measure
whether different people see colors similarly. As children, we learn that red is red, but
actually we do not know whether some people see red like we see green. One thing we
know is that some people do not distinguish between colors such as red and green.
These people are color blind. Color blindness of some degree is more common for the
male population at approximately 8% than for the female population at approximately

71
CHAPTER 3

1%. The reason for color blindness is that there are less than three different types of
cones in the individual's eye or that performance of some cones are at very low level.
Features of human color vision are the following:
In comparison, the human eye can distinguish hundreds of thousands of col-
ors. Individual differences are large — artists with trained eyes may distinguish
millions of colors.
In the technical reproduction of a colored picture, many colors are therefore
necessary to obtain high quality. In a three-color image stored with 8
bits/pixel/subcolor, there are approximately 16.7 million color combinations.
Due to the noise, all these are not necessarily different. In top quality digital
reproduction, 10-12 bit/pixel is necessary for each subcolor.

Humans can “remember” only 5-10 colors. Without a comparative reference,

the human eye is therefore a very poor device to measure whether a color is
correct. With five different colors, the error for the normal population would be
approximately 20%. This means that if traffic lights had five different color sig-
nals instead of three, every fifth driver would make a mistake. In technical
reproduction, typical memory colors are the classic “green grass,” “blue sky,”
and “skin color.’ These must correspond to the viewer's understanding of the
real color.

Humans cannot find the same darkness levels if colors of the object are differ-
ent. A person is unable to compare whether a red surface reflects more inten-
sity than a blue surface. In actual objects, there are differences in darkness. At
present, these are not factors of technical color reproduction.

Humans see 20-200 different darkness levels for one color. The exact number
depends on the quality level of the general image and the detail structure of
the picture. Poor general quality and many high frequency details allow a
lower number of gray levels.

Due to the three types of color receptors in the human eye, humans can sepa-
rate three components of a color. These components do not correspond to the
spectral sensitivities of the receptors — red, green, and blue. The visual color
components are the following:

- Hue or dominant wavelength seen as red, green, blue, etc.

- Saturation or the amount of white light in the color seen as an axis of pure
red, pale red, gray, etc., as Fig. 11 shows.

- Value or luminance or the total light reflection of the colored surface. (For
nonchromatic colors, this refers to axis white, graygray, or black. For chro-
matic colors, this means intensity variations in which dominant wavelength
and greyness are on constant levels.)

ve
 Color and color images

Figure illustrates the components of visual color.

”
Yo

507

no gray
REFLECTION component
+ t |
400 500 600 700
WAVELENGTH, nm

100 100%
% %
= Zz
fe) Se
5 50 5 507
Ww WwW gray
— gray +
iL rm component
Ww component im
(ae: a
0 + t { Ot + + {
400 500 600 700 400 500 600 700
WAVELENGTH, nm WAVELENGTH, nm

Figure 11. Visual influence of saturation on red and corresponding emission spectra.

EERE
y a Ngee

a = = = a e is ea ae

Figure 12. Visual scales for hue, value, and saturation.

3 Measurement of color
Color measurement of an original scene, an object, or an already existing color image is
necessary to reproduce it?> ’°-'3, Color measurement is also necessary for control,
specification, and standardization purposes of imaging processes.
Quantification of a colored surface or light source uses the following:
- measurement of the spectrum of light

- visual comparison

- three filter measurement.

73
CHAPTER 3

The spectrum of light

measures energy emitted by
the colored surface as a
function of wavelength.
Spectrum measurement is
physically the only exact way
to measure a colored object.
It does not directly give the
visual components of a
color, although calculation of
these is possible from the
spectra. The same spectra
always give the same visual
sensation, but the same
visual sensation can result
with several spectra
because of metamery. This
0:27 10:3" 0:40:57) Ooms Ose
makes spectrum measure- &
ment less useful.
In visual comparison, Figure 13. Basic ClE-coordinate system as x-y-cut of the three
the test person compares dimensional color range
the sample to be measured
to a set of reference samples. The measurement value is the numerical specification of
the reference sample giving an exact match. The method is the only exact way to char-
acterize visual color numerically. It is slow and impossible to use for the technical mea-
surement of color for reproduction purposes.
In a three filter measurement, measurement of the colored surface uses three
colored filters. Local color then has three numerical values. The filters used are color
separation filters or tristimulus filters.
Color separation filters are red, green, and blue. Each filter transmits one-third of
the visible spectrum with no overlap in filter transmission. Transmissions of tristimulus
filters correspond to the spectral sensitivities of the three different types of cones in the
human eye. The transmission spectra overlap considerably.
Results of three filter measurement have technical coordinates as RGB (red,
green, and blue), CMY (cyan, magenta, and yellow), or colorimetric coordinates as HSV
(hue, saturation, and value). RGB values are direct local image intensities or densities.
In an ideal system, CMY is obtained as

C= i 7
Wes oe (1)
Ve eB

74
 Color and color images

SPECTRAL ENERGY LOCUS

(WAVELENGTH, NANOMETERS)

580

GOLD !
WARM WHITE ,

COOL WHITE yt
DAYLIGHT mi

EQUAL ENERGY

Figure 14. Visual colors in the CIE triangle.

Conventional colorimetric expression uses the CIE standardization proposition

introduced in 1931. This involves first measuring color through three tristimulus filters:
FMX (average wavelength 450 nm = blue), FMY (555 nm = green), and FMZ (610 nm =
red). This provides three values designated X, Y, and Z. Calculation of color coordinates
x and y uses Eq. 2.
Figure 13 shows the basic CIE coordinate system. This commonly used repre-
sentation presents color but not the different levels of gray. Pure colors are on the curve
of the horseshoe while purple colors are on the connection line. Neutral gray is in the
middle of the horseshoe. Figure 14 shows the visual colors in the CIE triangle.

The)
CHAPTER 3

XG
= Fe
— X .

oe
—_ Y
(2)
The original idea of the CIE color representation system was to describe the colors
as the human eye sees them. The system did not fulfill these expectations. The human
eye should see color differences equally at different areas of the coordinate system. In
green areas, the system is very insensitive. In different positions of the CIE-color range,
the just noticeable color difference or JND is therefore different as Fig.15 indicates. In the
visual scale, the numerical coordination of the basic CIE is therefore not linear.

0.1 0.20.3 0.4 05 0607 08 01 O02 0:3 64°05 06 (07

x vm

Figure 15. Visually equal chromaticity steps at constant luminance level in the CIE xy and CIE uv coordi-
nate system.

Because the of nonideality of the basic CIE system, several modifications exist
whose basic intent was to force the numerical coordinate system to be uniform with the
visual sensation. These are “rubber sheet’ operations for the basic CIE-coordinate system
to correct the uniformity geometricaly. A theoretical basis for the corrections does not
exist. The most widely used CIE modifications are the CIE Lab or CIE Luv coordinates
1
a Y™\3
a 116 =) —16
n

h =atan(b
/a )

76
 Color and color images

where X,, Y,, and Z,, are stimuli for reference white
h_ is hue angle,
C saturation or chroma.

For CIE Luv, the formulas are the following:

ui = 4X(X + 1SY
+ 3Z)
Via OY
(CX ta1SY +43Z)

Le =eyyel 16 (4)
meer (fl=us)
vy =13L (v'-v)
While the basic CIE representation is two-dimensional (the luminance component
is neglected), CIE Lab and CIE Luv contain the all three visual components. Figure 16
shows the schematic shape of a three dimensional color coordinate system.

Figure 16. The three-dimensional representation of Luv-color coordination system where the L-axis is
equivalent to the luminance — different levels of achromatic colors i.e., levels of gray from white to black.
The range of reproducible colors is the color gamut.

Besides CIE modifications, several other color representation systems are in use.
The Munsel representation of color uses the visual comparison of the color to be mea-
sured with a reference color of known coordinates. The method is exact but subjective
and slow. Measured data can also calculate Munsel values.

Lh
CHAPTER 3

The GATF color circle uses a densitometer measurement with color separation
filters. The measurement gives three density values for one color measured through
red, green, and blue filters. RGB density values provide calculations of greyness as fol-
lows:
D,
Greyness
~/
G = 100—
D,

(Di?)
Hue error E = 100
(D,,— D))
where Dy, is highest density value
De middle density value
D, lowest density value.

Figure 17 shows the presentation of colors graphicly in a GATF color circle. The
GATF system is suitable for process control, but at certain areas it is very insensitive for
indication of visual changes.

ll \S

Figure 17. GATF color circle.

78
 Color and color images

4 Reproduction of color
4.1 Basic principles
The reproduction of a colored scene in a different place and time uses visual
trichromatic theory *3!21518.17 The basis of this theory is that a chromatic stimulus
can be matched as a mixture of three suitable selected sources of optical radiation as
Fig. 18 shows. Metamery is the phenomenon of using different spectra to produce the
same visual match. It has application in color reproduction by matching any visual color
from three independent stimuli.
The principles of color reproduction are the following:

Selection of three colored inks: The colors of the inks must be different, but
there is no “correct” set of inks but several sets can be applied.

Predefine the color matching functions: Color matching functions tell how
much of each ink is necessary to form the desired color. Color matching func-
tions are different for different sets of ink.
Separate the colors of the original picture into three components: The separa-
tion involves measuring picture colors point by point through three colored fil-
ters (color separation filters or tristimulus filters).

Multiply the separation signais by color matching values: This gives the
amount of each ink or other dyestuff necessary to produce the desired color.

Mix the colored inks or colored lights in a controlled way on the image plane
as instructed by the matching to obtain a multicolor reproduction of the original
picture.

> kK YAN
ZZ Za
O Visual match ©
¢ D
a ii \

WAVELENGTH WAVELENGTH
OF LIGHT OF LIGHT

Figure 18. Different spectra produce the same visual match in a phenomenon called metamery.

Technical color reproduction uses the absorption of the existing light or the emis-
sion of the desired spectrum. Absorption means that a part of the spectrum of light is
cut off. Printing ink on paper absorbs controlled spectral parts of the existing white light.

79
CHAPTER 3

Three colored inks corresponding to the thrichromatic theory are necessary. Emission
gives the energy distribution of a light source or any light reflecting surface. When the
printing ink layer cuts off part of the existing spectrum, the remaining part is the emis-
sion of the surface. Three colored light sources generate the emission in displays such
as color television or computer monitor.
Figure 19 shows that
subtractive color formation is Cyan Magenta Yellow
the spectrally selective White "97" white "ht white light
absorption of existing (white) Hit
light. In an ideal case, each
ink absorbs one-third of the
visible spectrum, and the
absorption spectra of the Cyan
inks do not overlap. The inks absorbs
red
are cyan (absorbs red),
magenta (absorbs green), Paper reflects diffuse light
and yellow (absorbs blue).
Figure 19. The principle of print color formation.
Each ink absorbs one-
third of the visible spectrum.
Table 1 shows the possible combinations and the corresponding color in subtrac-
tive reproduction corresponding to Fig. 20.

Table 1. Inks and their corresponding colors.

Inks Color Inks Color

No inks White M+Y Red
Yellow(Y) Yellow C+Y Green
Magenta(M) Magenta C+M Blue
Cyan(C) Cyan C+M+Y Black

Additive color formation Subtractive color formation

Figure 20. Additive (colored lights are mixed) and subtractive (colored inks are overprinted) color formation.

80
 Color and color images

In additive color formation, the subcolors for the light sources to be mixed are
red, green, and blue.
In additive color for-
mation, the spectra of the Raster
light sources need not Blue 4) —T scan
exactly fill the third of the
spectrum. Even narrow peak
spectra such as lasers pro- Modulation
and deflection Image
duce satisfactory reproduc- Light
plane
tion. A narrow absorption SOUrCeS
band in subtractive color for- Figure 21. The principle of display color formation.
mation would spoil the
result. This is the main reason why additive color imaging technologies generally give
wider color ranges than subtractive technologies. The basic technology in additive color
formation is focusing the color dot by dot on the screen operating as an image surface
as Fig. 21 shows.
Table 2 shows the possible combinations and the corresponding colors in additive
systems.

Table 2. Colors resulting from lights on various colors.

Lights on Color Lights on Color

No light Black B+G Cyan
Blue(B) Blue B+R Magenta
Green(G) Green G+R Yellow
Red(R) Red B+G+R White

Display technology uses three colored lights: red, green, and blue (RGB). Figure
22 shows that the resulting color range is usually very wide. In color printing, black ink is
commonly the fourth ink. The corresponding abbreviation for the image signal is then
CMYK (cyan, magenta, yellow, key = black). The black printer is used for the following
reasons:
- total density range of the print will increase

- detail rendering will improve

- black ink can compensate the gray component of the three colored inks

- black ink is cheaper than colored inks and its demand on gray areas of images
is only one-third of the total demand of colored inks

- text printing already uses black ink.

81
CHAPTER 3

Color space Tone space

, Visible colors as CIE coordinates
Color television _ Color television
0.8 Color film a
~ Color film

. Color printing ei Color printing

0.8 0 1 2 3
X DENSITY
Figure 22. Color ranges obtained in display and printing.

Image operations are those used to manipulate the picture signal for the desired
quality consistent with the method used. Image operations can be any of the following:

- manually

- optical technology

- analog technology

- digitally.
Manual operation such as the retouching of photographs, printing plates, and cyl-
inders was a common technology up to the 1980s. The use of conventional manual
methods has now decreased significantly. Interactive manual methods in the computer-
ized processing of image data have increased correspondingly. The major drawbacks of
manual methods are the high skill demands required and the high quality variations of
the result. Reproduction of the highest quality still requires manual adjustements.

Display technology
Display devices are light emitting active displays or light modulating passive
displays ‘+1819. Active displays generate radiation. The method applied to generate
radiation can be the following:

- photoluminecence in which photon absorption triggers the radiation

- cathode luminecence in which an electron beam gun generates radiation

- electroluminecence in which an electronic current generates radiation.

Regardless of the form of the input energy, the final step in the display process is
the transfer of an electron from one energy level to another. If E,; and E> (E> > E;) are
the energy levels for a molecule corresponding to light emission for wavelength A, then

82
 Color and color images

he
ai E,-E, (6)

where c_ is speed of light

= 2.9978 +10°m/s
h the Planck constant = 6.62 - 1074 Nms

Instead of one wavelength, the emission is usually distributed to a wider wave-

length range. After the feed of the input energy has stopped, the radiation will still con-
tinue for the time necessary for the energy transmission in the molecule. For most
molecules, the time is 10-9-10°° s. It may also be several seconds. This phenomen is
phosphorescence, and it finds wide use in display technology.
The cathode ray tube (CRT) is the most common display device. It generates
electron beam radiation using cathode luminecence and transforms it to visible light
using phosphorence. The properties of the phosphorus used determine the colors of the
display. Several sets of red, green, and blue phosphorus are available.

1)cE lectron gun

2) ;
3) }Deflection coils
4)P hosphorus
5) Display plane

Figure 23. The operating principle of the cathode ray tube.

Figure 23 shows the operating principle of a catode ray tube. The electron beam
generated by the electron gun deflects in the x- and y-directions by deflection coils. The
image surface that operates as an anode has a phosphorus coating that transforms the
electron beam to visible light. Thermal emission from heating of the cathode generates
electrons. Metal electrodes focus the electron beam on the display surface. The system
usually contains a grid to control the electron beam. The entire system is an electron gun.
The electron beam is raster scanned over the image plane. The scanning direc-
tion is usually horizontal, and the entire picture contains k scanning lines. Scanning is
electrostatic or electromagnetic. Electrostatic scanning gives a higher speed, and elec-
tromagnetic scanning gives a better beam acceleration. The result of the latter method
is a smaller scanning spot — better detail accuracy in the picture — and higher screen
luminance.

83
CHAPTER 3

When the electron

beam hits the image surface,
it generates visible light Electron
through phosphorescence.
Figure 24 shows that the g Emission
size of the phosphorus parti- ont
cles is approximately 5 um, 06
and the thickness of the alu- ever Xo)
minum surface that is the in
return path for the electrons Phosphorus ~ Dark
is approximately 0.1 um. particles glass
Both dimensions are critical.
Thick aluminum surfaces Figure 24. A cut of the CRT display surface.
can absorb the electron
beam energy. A thin layer does not reflect light sufficiently. An excessively thick phos-
phorus layer causes light scattering and excessive point spreading. A thin phosphorous
layer leads to an uneven illumination on the image surface.
In the figure, electrons
penetrate the aluminum
layer and trigger the cathode
luminesence of phosphorus
particles. Aluminum trans-
fers the charge and operates
as a “mirror” for the light gen-
erated by the phosphorus.
The glass surface has tinting
to avoid the reflection of the
external light back from the Shadow mask Phosphorus
aluminum surface. Figure 25. Shadow mask of a CRT color display.
Color formation ina
CRT display uses a shadow mask. The CRT has three electron guns. Figure 25 shows
that the focus of each will hit a certain position on the shadow mask. The shadow mask
directs the electron beams of the three guns to the correct positions at the phosphorus
surface. The mask is a plate with holes. The three different types of phosphorus for the
red, green, and blue lights are separate from each other.
Most television and computer displays are CRT tubes. An analog or a digital sig-
nal can control such a display. The display itself operates analogically. The signal in the
display can be a television signal or an RGB signal. A television signal consists of three
visual color components, one luminance component (determining the luminance level —
the only component in black and white television), and two chrominance components
(determining hue and saturation). A television receiver transforms this visual signal into
an RGB signal to control the three electron guns. Computers usually generate RGB sig-
nals directly.

84
 Color and color images

In a CRT for a television or a computer display, an electron beam scans the sur-
face line by line. This is raster scanning. A television operating according to the Euro-
pean television standard undergoes scanning with 625 lines over the display. In the
United States, the standard is 525 lines. The scanning frequency (Hz) indicates how
many times the screen is scanned in one second. The scanning frequency in television
is half of the net frequency — 25 Hz in Europe and 30 Hz in the United States. Because
of interlacing, the whole image needs two scannings. For computer displays, there are
different resolution standards from to .
At low scanning frequencies, the human eye can see the two sublimates. The
result is that still images flicker at scanning frequencies that are too low such as 25 or
30 Hz. Sufficient frequency for the human eye is 70 Hz. In computer displays, the most
relaxing frequency varies, but it is higher than that for a normal television screen.
CRT technology is mature, cheap, and reliable. It gives proper image quality. The
disadvantages are the large size and high level of energy consumption. This is why CRT
displays over 40 in. are difficult to construct and CRT technology is not suitable for por-
table displays.
Liquid crystal displays (LCD) are passive displays. The two types of LCD are
reflecting and transparent. The reflecting type needs illumination from the viewing direc-
tion. The transparent type needs back illumination.
Liquid crystal displays use the different optical properties of materials at different
temperatures. A current generates the temperature change. At the lower end of the tem-
perature range, the material is a crystalline solid, and at the upper end it changes into a
clear solid.
Plasma displays rely on the glow produced when an electrical current passes
through gas — usually neon. Free electrons and ionized gas atoms are present during
the discharge. Under the influence of the external field, the electrons acquire a high
kinetic energy. When they collide with the gas atoms, they transfer energy to the atoms
and excite them to an energy level above the ground state. The atoms then lose energy
by radiation and return to the ground state. Acceleration may be ac or dc, although ac is
more common.

Color image processing for printing

Reproduction operations for color pictures are the following? 1

. . . LTA.
.

- Color separation: The division of the object color into three or six signal com-
ponents.
- Color correction: The correction of the color errors due to nonideal tight
absorption properties of the printing inks.

- Halftoning or screening: Converting the continuous tone images into halftone

images consisting of halftone dots with a varying area coverage.
- Image restoration: Correction to equalize the output and input images or the
output and the original object.

85
CHAPTER 3

- Image enhancement: The intentional manipulation of images to obtain better

suitability for the end use of the image. With images for visual communication,
the final goal of enhancement is usually to improve visual image quality.

- Operations concerning the shape, size, and position of pictures: These

include cropping, merging, sizing, rotating, and mirroring the image.

- Image capture and output operations.

The reproduction operation for color images contains these operations. A typical
execution of the process uses the following order with the indicated devices:

- separation of colors in cameras or scanners

- correction, restoration, and enhancement of color with cameras, scanners, or

computers

- optical screening in cameras or scanners unless using digital halftoning

- shaping, sizing, and position using manual methods or computers

- digital halftoning in computers unless using optical screening.

The result of these image processing operations is a halftoned color separation

film or photographic paper ready for manual page makeup or processed, coded image
signals for combination with digital text information for display and output.

4.2 Color separation

Color separation occurs by measuring the amount of red, green, and blue colors at each
point of an original image. Obtaining the three components involves measuring the
original color image — a color photograph. With electronic cameras, the camera
separates the colors of the original object using three colored filters. The average
spectral absorptions of these filters are 440 nm for the blue filter, 540 nm for the green
filter, and 620 nm for the red filter.
Each color separation filter sees its own color in the object as white and its oppo-
site color as black. If the desired color is blue, its opposite color is the rest of the spec-
trum, i.e., yellow. From the color separation step until color printing or display, the
reproduction process operates as “black and white.” The color separation signal values
are of significance only on the scale that extends from white to black. Their use for gen-
erating the multi color picture in output depends on the imaging method. For additive
color formation, the “white” values of the separation signal are of interest. Controlling
the RGB signal allows control of the image. For subtractive image formation, the “black”
values are of interest. Controlling the opposite colors of RGB — the CMY values — allows
control of the image.
Color separation produces a signal transmitted for further processing as an opti-
cal signal, analog video, or digital signal depending on the equipment environment. Fig-
ure 26 shows the scheme of color separation for subtractive image formation such as
printing.

86
 Color and color images

Yellow Purple Bluegreen

The original color
of the object NY NY NY

The filter which Blue Green Red

sees the color filter filter filter
as black NY NV NV,

Separation signals

The ink determined NV NV NV,

by the signal Yellow Magenta Cyan
in printing

Figure 26. Color separation scheme for subtractive image formation.

For color separation, original pictures or objects contain colors that a particular
separation filter should see as black and colors that the same filter should see as white.
The following equation provides a calculation of the efficiency of color separation:

, _ 100(Dy~Ds)
Dy D,+D, (7)

where JD,Dp is the density of the color that a particular filter should see as black,
D Ss the density of the color that a particular filter should see as white.

Figure 27 shows that the total result of the separation depends on the following:

- energy emission of the light source used to illuminate the original image or the
illumination such as flash used for original objects

- reflectance and transmission properties of the original image or object

- transmission spectra of the separation filters

- spectral sensitivity of the light detector.

These factors influence the result multiplicatively.

87
CHAPTER 3

Zz Zz
SO OS
ap)

:in ‘ [N\\ a ee || ‘all

on —_—_—

fi
400 700 400 700 400 700 400 700
WAVELENGTH, nm WAVELENGTH, nm WAVELENGTH, nm WAVELENGTH, nm
Sensitivity Spectra Image Separation
of light of color spectrum signals
detector separation
filters

Figure 27. Formation of the color separation result from its spectral components.

Separation of black — the fourth ink in printing — uses sequential measurements

through the three color separation filters, measurement

- through a yellow “gold” filter, or

- computation from the three color signals.

The last technique is the for the present most common.

A dilemma exists in the selection of the color separation filter spectra. If the color
separation is done in the way the human eye sees the colors, the transmission spectra
of the filters overlap considerably. This leads to a situation where saturated colors are
difficult to reproduce because tight and wide spectra of the original colors appear similar
to the filters. For saturated colors, tight transmission spectra of the color separation fil-
ters are best, but then the system does not distinguish all the colors — those remaining
between the filters. A solution is the multi filter separation using more than three filters.
This is possibile in reproduction using high fidelity.
A basic difficulty in normal color separation is to see magenta as black and green
as white in the green filter separation. A general rule of thumb is that if the green filter
separation is acceptable, the entire color separation is accepable.

4.3 Color correction

Color correction is implementation of the practical operations in color matching. This is a

basic target in color reproduction. This section will discuss the reasons and practical
procedures of color matching for printing reproduction.
Color correction compensates for color distortions caused by incorrect and
insufficient light absorptions of the colored printing inks. A misunderstanding of color
correction is the collection of methods to correct existing faults in original images — color
restoration — or to improve color rendering of print or display — color enhancement. In
printing reproduction, if the paper used would reflect all the incoming light and each
color would absorb exactly one-third of the visible spectrum and be totally transparent
for the remaining two-thirds of the spectrum so that the absorption of the three inks

88
 Color and color images

would not overlap, no need for color correction would exist. Real printing inks do not
absorb a sufficient amount of the desired wavelengths of light and absorb too much
undesired wavelengths as Fig. 28 shows.

Magenta Cyan

TRANSMISSION
700 400 700 400 700
WAVELENGTH, nm

Figure 28. Ideal (dotted line) and real (solid line) light transmission spectra for one set of printing inks.

When color separating a gray original, the same color separation signal values
result from each of the three filters. If the printing were done using this uncorrected sig-
nal, the result for a typical ink set would be brown not gray, and the darkness would be
higher than that of the original. Color correction eliminates these deficiences. Keeping
the gray tone of original images gray in the print is a main criteria for a successful color
correction. This is gray balance. Each set of printing inks has an equivalent neutral den-
sity (END) curve. The curve shows how much of each ink requires printing on different
darkness levels to obtain neutral gray in the print. The END curves also depend on the
set of inks used and the order of printing. A general rule is that a cyan dot must be big-
ger for END than yellow and magenta dots.
The reason for the nonideal absorption of light is the statistical nature of light
absorption. Suitable dyestuffs with sharp absorption spectra are not available. For
example, magenta should absorb green light, but it also absorbs blue and red light.
Magenta therefore has a tint — it is slightly yellow and slightly cyan as Fig. 29 shows.
Printing inks therefore have desired absorption and undesired absorption proper-
ties. The basic principle of color correction is that the amount of ink (primary absorp-
tion) is reduced locally to correspond to the local amounts of tint of the two other inks
(secondary absorptions).

89
CHAPTER 3

x .
cutee:
i}
£ a
ZZ Zz ee
O ! On ee roa
= ' Fe ES= ek
fal 6 ! — Wo) aed
cr
O
Se
GBQ2QG!'
8 >Sg
aie
GLH
c | 8%
O 3 8
cs
So
a) Qae! Gae Lee a) == Oe
faa] SoH! ESO SOG a
<x S26'599 o8s <x
OBE! GRE OWE
|
= _—_——__
400 500 600 700 400 500 600 700

WAVELENGTH, nm WAVELENGTH, nm

Figure 29. The tinting of magenta ink.

Figure 29 shows tinting of magenta by yellow and cyan. Yellow and cyan signals
are consequently reduced locally by

Y=Y—My and ¢ =C—M, (8)

where Y_ is yellow separation signal
My yellow component in magenta signal
G cyan separation signal
M. cyan component in magenta signal
k a constant for the secondary blue absorption for the magenta ink used.
Three inks therefore require correction with two secondary absorptions for each
ink to be corrected. This means that six correction operations are necessary. Table 3
shows typical color separation filter densities for a set of printing inks.

Table 3. Typical filter densities of a set of printing inks with ideal filter densities shown
parenthetically.

Red filter density Green filter density Blue filter density wal
Cyan 1.30 (infinite) 0.40 (0) 0.15 (0)
Magenta 0.10 (0) 1.05 (infinite) 0.60 (0)
Yellow 0.01 (0) 0.07 (0) 1.00 (infinite)

As Figure 27 and Table 3 indicate, the light absorption of yellow is nearly the
ideal. This is typical for all commonly used sets of ink. The secondary absorptions of
cyan and magenta are considerable. By compensating in color correction for the two
most critical absorptions, the result is satisfactory.

90
 Color and color images

Due to deficient ink properties, the printing of pure cyan or magenta is not possi-
ble because there are no other inks to compensate the tint. Figure 23 shows that this is
why the color range of printing has a limitation.
Color correction corrects for tint errors inside the color range determined by the
three ink set used, but it is not able to widen the range. The range of printing is usually
smaller than that of original images. Those colors that cannot be reproduced are out-of-
range or out-of-space colors. Section 4.5 discusses their reproduction.

4.4 Screening (halftoning)

Screening or halftoning is the conversion of a continuous tone image to a halftone
image /9-#2. Screening originally used light diffraction in a narrow slit. Exposing a siit
network with even illumination gives a two-dimensional sinusoidal illumination. Exposing
this with a continuous tone photograph on a photographic high gradation film gives a
halftone dot structure. In a halftone dot structure, the relative dot size varies with the
darkness level in the original photograph.
Invention of screening occurred in approximately. The invention made possible
the printing of photographs inexpensively. Photography in 1880 was already approxi-
mately 50 years old, but it used cumbersome printing methods for the pictures. Most
pictures printed before the invention were hand engraved on wooden plates. They were
not true reproductions but an artist’s view. For almost a century, printed pictures had
much power as evidence as the phrase “photographs cannot lie” suggests. Some forg-
eries of printed photographs from this era are well known. From the beginning of the
1980s, picture processing has used computers. Forgeries are technically very easy to
make, and the printed picture has lost some of its power as evidence. Consider that
people can write whatever they want. Readers believe in some publications more than
others.
Screening originally used glass screens. These were networks of narrow slits
engraved on a glass plate. The user held the glass plate or glass screen at a given dis-
tance from the film on which a halftone image was exposed. In glass screening, the fol-
lowing variables affect the result:

- distance of the screen from the film plane

- the f-number (cf. Ch. 2) of the exposing system

- the ratio of black and transparent area in the slit network of the screen.

91
CHAPTER 3

Some automatic reproduction cameras today use glass screens. Figures 30 and
31 show the principle of optical screening.

Even exposure

vyyyy
2S S| aa
Sinusoidal
exposure

Image
plane

Figure 30. The generation of sinusoidal exposure for a halftone dot formation in optical screening.

Illumination in optical halftoning

i = i+(b/2)cos(v
2'7 mx)cos(
V2‘ Timy)
b = density range
m = screen frequency

Dot shapes are plane cuts of

the illumination surface

Figure 31. Intensity distribution generated by an optical screen and the corresponding halftone dot struc-
ture with halftone percentage = 100xdot area/total area.

A contact screen is a tone film exposed through a glass screen and developed
with low gradation. The intensity distribution generated by a glass screen on the film
plane has been stored on the contact screen film. A contact screen is put on the film or
photographic paper that is exposed in an exposing device. Contact screens find use in
reproduction cameras, contact frames, and scanners.

92
 Color and color images

Halftoning is done by
computer on digital image Light tone Middle tone Dark tone
signals. The basic compo-
nent of halftoning is a pixel {4-4 wt eg
matrix corresponding to a
halftone cycle. The matrix is
A. Conventional screen
filled with pixels that repre-
sent halftone dots. Halftone
dot structures that are na aaa -
impossible in optical screen-
ing can be generated. Due B. Dicital : ters
to the pixel structure, tone . Digital screen, 3 x 3 matrix size

rendering in digital halftoning Figure 32. Halftone dots produced by optical screening and by digi-
is always stepwise. tal halftoning.
Figure 32 shows schemati-
cally halftone dots produced by optical screening and by digital halftoning. Varying the
relative covered area generates different tones.
The following are the main variables of screening and halftoning:

- halftone dot size

- screen frequency (screen ruling, screen density)

- screen or halftone angle

- halftone dot structure.

The main purpose of screening is to convert the variations of the local density in
the original images into variations in the halftone dot area. In printing, a small halftone
dot (3%—10% depending on the printing proces) corresponds to the highlight point of
the original, and a large halftone dot (85%—95%) corresponds to the shadow point of
the original. Printing avoids areas having no dots (white paper) and solid surfaces
(100% dot) because these would result in stepwise tone rendering in the light and dark
ends of the picture.

Table 4. Screen frequency in printing.

[ Type of printing Screen frequency ae

Newspaper printing 18 — 34 I/cm
Offset printing, machine finished paper 40 - 50 I/cm
Offset printing, coated paper 60 — 80 I/cm
Gravue printing 60 — 80 I/cm
Digital printing 30 — 80 I/cm

93
CHAPTER 3

The screen frequency

expresses the spatial fre- Halftone dot Pattern of
quency of the halftone dots. cycle small dots
In impact printing, the
screen frequency that is suit-
able for printing depends on
the printing method used
and on paper properties. In
digital or nonimpact printing,
Screen ruling | = aie (lines/cm or lines/in.)
machine characteristics also a
play a role. The screen fre-
Figure 33. Optical screening for black ink produces a dot pattern at
quency is in lines/cm or in an angle of 45° as measured from the horizontal line.
lines/in. The latter results
from the former by multiplying with a factor of 2.54. Figure 33 shows how to determine
screen frequency. Table 4 lists typical screen frequencies in printing.
An increase of screen frequency improves the detail rendering in an image. The
following items limit the screen frequency:

- Information carrying capacity of paper. (Coated papers can carry more infor-
mation than uncoated papers and newsprint.)

- Image formation ability of the printing plate. (Letterpress and flexo plates may
be critical in this context.)

- Picture element size in electronic printing presses and laser printers.

Two halftone images on top of each other produce the Moiré-effect shown in Fig.
34. Moiré is a kind of aliasing phenomenon originating from interference of two equal
frequencies. If the angle of two equal halftone fields is zero, there is no Moiré. The fre-
quency of the Moiré pattern has a minimum, i.e., the period of the pattern maximum at
small angles. This is why multi color printing maximizes the screen angle difference. The
angles between screens for different colored inks is normally 15°.

Figure 34. Moiré-patterns generated by overprint of two screens with the illustration on the left having a
higher frequency pattern than the illustration on the right.

94
 Color and color images

The term halftone dot structure indicates the halftone dot shape inside a dot
cycle. Optical screening produces a diamond dot with possible minor modifications. Dig-
ital halftoning can produce distributed or dispersed dots as Fig. 35 shows. Halftoning
involves filling the matrix pixel by pixel in the order given by the numbers. The highest
tone is where pixel number one is black and the others are white. A dispersed dot is one
in which the relative area covered as indicated by the halftone percentage has several
“sub dots.” Figure 36 demonstrates the influence of a halftone dot structure on image
appearance. The nominal screen frequency is the same in all the examples.

Diamond dot Dispersed dot

|61]53] 41]33|37|52 |60|64| 4 [a3]9[41] 3 [35[11 43

4929] 21] 5|9 4 [28|
[3919] 11]1|3 [8[16/36 61 29[53[21|63 [31/55/23
4 [36[12/44] 2[34|10]42
aal27]24vol6 fe 30 [sal20]edleslsohs [sales
55[47]31]19[1426 46 r6lae| 8|ao)r| 46]6fs]
also) soaleale

Screen frequency = 1/a

Cycle frequency = 1/b
Pixel frequency = 1/c

Figure 35. Halftone dot structure and the definition of halftone frequencies.

In digital halftoning, the halftone dot matrix is filled by pixels according to given
rules. If the size of the matrix is K x L pixels, the number of the available gray levels is
2x Kx L +2 gray levels in the common case where matrices are used for forming a
halftone cycle and white (no pixels) is applied as one level. White (no pixels) and solid
black (all pixels black) are avoided in conventional printing. This reduces the number of
gray levels.
The variety of halftone dot structures in digital halftoning is infinite in limits of pixel
size and halftone matrix size. The basic structures are deterministic halftone, random
halftone, and frequency modulated halftone.
Figure 36 shows that the deterministic halftone simulates the conventional half-
tone dot structures. These are “diamond dot” structures. A halftone for black ink

95
CHAPTER 3

(45° screen angle) is reasonably easy to obtain in digital halftoning. Other screen
angles (+ 15°) are more complicated, because the pixel and halftone cycle grids do not
rotate. Most deterministic halftone dot structures are quite similar to print as optically
produced halftones, and the print quality will be equal. Some halftoning algorithms do
not properly indicate the screen angle. Under colored removal problems, the result is
Moire.
Random halftoning algorithms distribute the even middle tone halftone dots
inside the cycle to several sub dots. This increases the sharpness of the halftone image
and improves its visual quality. It gives better print quality at the same pixel size and
screen frequency level than with deterministic dots. In printing, random halftones are
more sensitive to fill in and more difficult to print than deterministic halftone dots.
Use of frequency modulated halftoning (FM halftoning) has increased. In this
method, screen frequency also varies inside a picture. The method gives excellent
image quality with particular use in high fidelity reproduction. The variables of halftoning
are the following:
- halftone dot size (coverage area)

- halftone dot frequency

- halftone dot structure.

All these influence mean density of a halftone field if the two others are at con-
stant level. This is why the density and color formation obtained by frequency modulated
halftones is difficult to predict for process calibration. Frequency modulated halftone
image with sufficiently small pixel size corresponds to continuous tone image. In the
future, print quality will improve markedly with frequency modulated halftones.

4.5 Restoration and enhancement

Restoration and enhancement are image operations to improve the reproduction final
result?
3° 1423 In a printed picture, considering the original scene would be restoration.
Enhancement considers the properties and preferences of the viewer. The following
result by definition:

- image restoration means equalization of the output with the input to the origi-
nal scene

- image enhancement operations try to improve the output image to give better
visual quality in the intended use.

Restoration concerns the maximization of information transfered from the original

scenes or images to the final images. Enhancement contrary to restoration imposes
purposeful differences in the final image compared with the input image. Enhancement
for one may be noise for another.

96
 Color and color images

halftone picture of 53 Ipi enlarged halftone by factor 2

ore

Hratatet.
reiatae
OUI) eo ete:
e
eter
arate
ererateleteners: Seseseseseiene
> ate.
cere.
tearereretere
Sec
pats

Petetatete
ont sta'e i)
SRO “.
.

Figure 36. Pictures at a constant level of screen frequency of 53 I/in. when using different digital halfton-
ing methods: a) is round, b) is diamond, c) is line, d) is dispersed, e) is square, and f) is cross.

97
CHAPTER 3

Figure 37 shows a flow chart of a control procedure.

Original Image Printed

image processing image

A posteriori A priori Control of

restoration restoration printing

Figure 37. The conventional control of printing reproduction.

In a printing reproduction process, image adjustment operations are implemented

in the image processing step. Printing remains constant by using control procedures.
This is a prerequisite to a priori prediction of the printing result and corrrect adjustments
of the images. The adjustment steps are as follows:

A posteriori restoration that corrects faults existing in the original image signal:
Such errors may be tinting of colors, blurring, or noise patches. For successful
a posteriori restoration, the formation mechanisms of the faults and their mag-
nitude must be known.

A priori restoration that compensates beforehand for faults that will occur in
the ensuing process steps: A typical a priori restoration operation is color cor-
rection. A priori restoration can be successful provided that the image forma-
tion mechanisms and the distortions that will be generated are known and that
printing is sufficiently stable.

Enhancement improves the visual quality of the printed result: The criteria for
enhancement are determined on the basis of visual tests on print quality and
the control of printing that keeps the printing conditions stable.

The following are the basic tools of restoration and enhancement:

adjustment of tone rendering

adjustment of color rendering

adjustment of detail reproduction.

98
 Color and color images

Tone rendering determines the darkness relationship between input image and
output images in an imaging process. It is controled by the shape of the tone rendering
curve or the shape of the image histogram in Fig. 38.
The density histogram of an images gives an expression to the relative area of
the image characterized by a given density ranging from zero to infinity. Each image has
a histogram of its own.

IMAGE
AREA,
%
OF
DENSITY
PRINT

DENSITY OF ORIGINAL DENSITY

Tone rendering curve Image histogram

Figure 38. Tone rendering curve (curve a enhances light tones, curve b is facsimile) and image historam
(histogram ais random, histogram b equalized).

The two commonly used image enhancement philosophies applied in printing

reproduction are the following:

- the equalization of the shape of the tone rendering curve for all images and
- the adjustment of the shape of the histogram. Equalization of the histogram for
all images produces contradictory results.

Equalization of tone rendering or the histogram causes the other to vary ran-
domly from one image to another.
A typical situation in printing reproduction is that the density range of original
images is wider that that of the printed images. This means reduction of the dynamic
range and reduction of the information transfer in the process. By adjusting the shape of
the tone rendering curve, one can adjust which parts (light tones, middle tones, dark
tones) of the image lose more information and which lose less information.
Because the probability of occurence of different density levels in an image varies
from one image to another, the image histogram is originally random. By equalization,
each density level becomes equally probable. This maximizes the information transfer
from input images to output images in an imaging process where the dynamic range of
the output is smaller than that of the input.

99
CHAPTER 3

Linear tone rendering

Histogram equalization

50% histogram equalization

50% linear tone rendering

Figure 39. The influence of tone rendering and histogram adjustment on the appearance of black and
white pictures.

100
 Color and color images

There is no scientific evidence determining how to enhance black and white

images for printing. Histogram equalization is better for “average” pictures. In certain
cases such as a black suited skier on white snow, it leads to an undesired result and
unacceptable quality. In imaging processes for reproduction, efficiency requires auto-
mation of algorithms for restoration and enhancement. Figure 39 shows that a system
that balances tone rendering and histogram adjustments so that the user adjusts the
balance is finding increasing use.
For multi color pictures, color rendering is adjusted. Color and the demand for the
adjustment are expressed as visual hue, saturation, and value (HSV) parameters. The
actual control may use the HSV, RGB, or CMY(k) color range.
The color balance of pictures may be distorted. Technically, this often happens
when one of the three primary colors is “tinted.” This could happen in color photography.
The color reproduction process seeks an anchor point. Tinting is corrected by adjusting
this to be correct in an image operation that correspondingly corrects for all the other
colors. Grays in the original scenes usually indicate color balance or the lack of it. The
adjustment makes grays in the original images appear gray in the print. This is gray bal-
ance. Figure 40 shows an example of gray balance adjusiment.
The term hue depicts the “visual color,” i.e., whether we see a color as blue,
green, etc. Color is not remembered very well, but the hue of certain colors can be eas-
ily recognized if it is wrong. Examples are blue sky that can easily turn to green due to
process errors. Green grass may turn to blue. Human skin is most critical. It can turn
reddish. These colors and desired trade mark colors can be fixed in a color reproduction
process. High quality reproduction uses more complex adjustments. Figure 41 shows
an example of standardization of skin color.

Figure 40. Gray balance adjustment of a color image.

101
CHAPTER 3

a) b) ; C)

Figure 41. The adjustment of the hue of skin color where a) refers to the original hue, b) to the European
standard color, and c) to the Asian standard color.

The term saturation depicts the amount of white light in a pure color. The range of
saturation of actual and original images (color slides) is much wider than that obtained
by printing. Many “out of range” colors exist. In the case of hue adjustment, the out of
range hues are pressed into the range so that the differences of the hues between
image will remain. Saturation has two philosphies:

- linear suppression of saturation

- printing all out-of-range saturations at the maximum saturation of the imaging

process.
The latter technique improves the general appearance of the saturation of an
image, but information is lost because the saturation differences of the out-of-range val-
ues are lost. Figure 43 demonstrates the situation.
The term value relates to the total amount of light. The adjustment of value is
equivalent to the adjustment of tone rendering. Histogram equalization is finding
increasing use for color images.
Image restoration is based on process profile measurements but image enhance-
ment is by far done as interactive operations. For the time being, the preferences of man
can be measured by means of visual tests, and enhancement can be taught to the com-
puter. Figure 42 demonstrates automatically enhanced images.

102
 Color and color images

_ ett ET Re aAda
(ibis

Figure 42. Automatically enhanced images.

The replacement of the gray component of images by black ink is under color
removal (UCR). A degree of UCR is given as a percentage. If the UCR is 100%, the sys-
tem has achromatic reproduction. Achromatic reproduction is therefore color reproduc-
tion in which the gray component (luminance component) is locally replaced by black
ink. This means that at most three inks are printed locally — two colored and black. Ach-

103
CHAPTER 3

romatic reproduction produces exactly equal color rendering as four ink reproduction
but with better printing stability. Figure 44 shows the principles of UCR and achromatic
reproduction. In achromatic reproduction, the locally remaining two colored inks are
usually decreased less than the amount of the whole gray component to improve color
rendering. This is under color addition (UCA).

Figure 43. The adjustment of “out-of-range” saturation with a) original, b) linear compression, and c)
maximum saturation

The main reason why

the range of color display is
wider than that of color print-
ing is that the absorption
properties of printing inks
are not ideal. This is why
high quality printing uses
more than four inks. The
Gray
number of subinks is usually component
seven — cyan, magenta, yel-
low, red, green, blue, and Figure 44. Principle of under color removal (UCR) and achromatic
black. This reproduction is reproduction.
high fidelity reproduction. A
seven unit press or several runs in a smaller press are necessary to print. High fidelity
printing does not have high use currently, but the introduction of high definition television
(HDTV) may put pressure on printers to improve quality considerably for commercial
purposes.

104
 Color and color images

5 Color formation
Basic color theory is very simple. In the reproduction of multi color pictures by printing,
several variables can distort the color formation from that predicted by a simple color
theory. The deviations depend on the nonideal spectral properties of the process
(nonideal transmission spectra of filters, nonideal absorption spectra of printing inks),
physical phenomena in the printing nip, and optical phenomena when the final
reproduction (intensity distribution of light) forms as the interaction of external light and
a printed halftone picture.
The most important variables whose influences require consideration in color
reproduction are the following:

- Nonideal spectral absorptions of the printing inks that partly limit the color
range of the printing process: This is partly because of the Moiré effect in multi
colorcolor printing. Moiré is visible as color noise on surfaces of an even tone
at lower frequencies than the frequency of the halftone dot structure.

- Halftone dot gain: An increase or decrease in halftone dot diameter in any pro-
cess step (output of halftone film/paper, page film exposure, printing piate
exposure, printing) will have an influence (cf. Section 4.3).

- Nonideal ink transfer in the printing nip: This is visible as ink transfer distur-
bance in wet-on-wet printing and mixing of different inks on the press because
of back transfer.
The result is that reproduction strictly according to the theory will not give a satis-
factory result because of the color distortions mentioned above. Each printing process
requires separate calibration to obtain an acceptable result. Image processing systems
are open, i.e., color image data can transfer from one system to another. This creates
the need to present color image data. This issue has two parts:
- calibration of input and output devices for color reproduction

- device independent color representation.

5.1 Color calibration

The calibration of the input and output that generally consists of continuous tone
RGB color values, halftone, color, and CMYK values is done for each imaging
device? 11-13,17,24.26 The purpose of calibration is to equalize output and input colors.
The calibration uses:
- statistical process model

- system theory based process model

- fundamental process model.

105
CHAPTER 3

Statistical process models use input-output measurements of the process. This

means collecting experimental data for the input colors as code values in computer
memory or as the relationship of the code values with the output colors of the print. In a
normal case, the measurement of several hundred different colors covers the color
range with sufficient accuracy. The experimental relationships are:

- original color — color value in computer

- color value in computer — print color.

Collecting these as tables with the LUT correction procedures generates the
desired, corrected result.
Experimental calibration has wide use. It is accurate if the number of measure-
ment points is sufficient. The disadvantages of the experimental calibration are the fol-
lowing:

- the measurement results are valid only in those conditions in which the mea-
surements were done (Extrapolation of results may lead to serious errors.)

- in four-color reproduction, there are infinite combinations of the four inks with
which a given color other than the pure cyan, magenta, or yellow can be
reproduced

- alarge number of measurements is necessary to obtain a reliable result (The

measurements will require repeating with changes in the process conditions
or materials.)

The masking equations belonging to the group of statistical process models offer a
way to computate the input-output relationship. The first order masking equations give the
relationship between an RGB signal stored in a computer's memory and a CMY signal:
= k,D,+k,D,+ksD,

ME = kyD,+ksD,+k.D, (8)

where JD, are color separation filter densities

C, M, and Y are cyan, magenta, and yellow densities in print.
The first-order masking equations represent experimental input-output relation-
ships. ener order masking equations including the interactions of filter densities, e.g.,
DD, and De . These can be transfer system models.
The Murray- Davies model predicts the average density of a print, D,,,,, from the
relative halftone dot size, a, (0 < a < 1) and the solid print density, D
fp)

D,,, = —loginp( 1-a+al0 °) (9)

106
 Color and color images

The Murray-Davies model assumes that the density of each halftone dot equals
solid density and computes the density value from the sum of relative reflections. The
constant density of halftone dots does not hold true. The Yule-Nielsen model takes this
into account by a factor n that compensates in a simplified manner for light penetration
into the paper between individual halftone dots (cf. Section 4.3):

SID)
fiz
D int = —f logio(1 —a+al0 ) (10)

The Neugebaur model is an exten-

sion of the Murray-Davies and Yule-
Nielsen models for color reproduction. It
sums the reflections of the eight possible
print and overprint areas in three color
printing as Fig. 45 shows.
The basic assumptions of Neuge-
bauer equations are that

- Anink prints a solid density or ay

does not print leaving the paper
surface visible. : nate
Figure 45. Three-color printing.
- The printing result can be char-
acterized by the relative covered area , i.e., halftone dot precentage.

- The halftone dots of the inks overlap or are side-by-side so that with three inks
2° = 8 color combinations occur.

In single color printing, the relative halftone dot area, f, results from the inverse of
the Murray-Davies model:
1 io

cate
ie Ss a (11)
Ore

If; s, and t (0 <x, s, t< 1) are the relative halftone dot areas of the three primary
inks representing the f values of Eq. 4, the dot coverages of the Neugebauer model are

f,= A-nd-s)d-?t)
(1
f= rd—s)\f, =9
=s- ),
nd— 2), f,= 11 -r)A-—s)
(12)
ee ah opal ad Roe ae)
fe = PSI

107
CHAPTER 3

If X;, Y;, and Z; are the tristimulus values of the plain paper and X — Xg, Y2— Yg,
and Z>—Zg are the tristimulus values of the overprints, the color of the print is obtained
as tristimulus values

X= > FX; Y= Day, Z= >, F4; (13)

When the Neugebauer equations are applied to four inks, the summation is car-
ried out from 1 to 16 (es = 16), respectively. The Neugebauer equations have under-
gone various modifications to improve their accuracy. The relative halftone dot area can
be calculated from the Yule-Nielsen model instead of the Murray-Davies model:

1 10 n”

oiey -D,/n (14)

1-10

Then the XYZ-tristimulus values for the n-modified Neugebauer equations are

The Yule-Nielsen and

Neugebauer models f(x)
assume that the maximum = JX
| |
density of halftone dots is
lower than the solid density
by a constant factor deter-
mined by n. This is not accu- _ Figure 46. Density profil of halftone dots.
rate. The absolute size of the
dot and the MTF or point spread function of the printing process determine the maxi-
mum density of the halftone dot. A dot profile model uses the definition of a point spread
function for the printing process. The approach therefore assumes that the output of a
printing system can be computed as a convolution of the input and the point spread
function. This means that the density profile of printed halftone dots, 2(x) (given for the
sake of simplicity in one dimensional form in Fig. 46), is then obtained as a convolution
of the input profile for a halftone f(x) dot and point spread function of the printing pro-
cess h(x):

g(x) = f(x)
x h(x) (16)
By writing the input function f(x) not only for a single dot but for a halftone area,
g(x) represents the density map of a halftone area. By using the same principle as

108
 Color and color images

applied in the Murray-Davies model, the following equation for the integrated density
Diy; results:

Ding = log yolk [108 dxdy] (47)

In Equation 10, the constant k can adjust the calculation to include an integer
number of halftone dot cycles. The model can apply to multi color prints by calculating
the three filter densities or the XYZ-tristimulus values of the print.
The dot profile model is a calibration model based on system theory. It deals with
the system as a black box. If the point spread function, its Fourier transform, or MTF of
the process is known, the dot profile model can predict print density and color in condi-
tions where prior measurements are missing.
To summarize, color calibration uses measurements of print color and the adjust-
ment of color rendering with corresponding reproduction operations. The intention is to
use computation and prediction methods that minimize the needs for measurements.
The methods can be the following:

- Look-up tables and first-order masking equations: These use only experimen-
tal data and give a result as good as the measurements.
- Higher order masking equations: These can be accurate, but the equations
are pure curve fits without theoretical assumptions.

- The Murray-Davies, Yule-Nielsen, and Neugebauer equations: These use a

basic assumption and experimental correction that are too simplified. A good
fit results with modified equations, but the error of the basic equations is too
big for calibration purposes.
- The dot profile model: This uses the assumption that the printing process is
linear and the tools of system theory can control it. Despite known nonlineari-
ties of the printing process, the dot profile model is suitable for engineering
purposes.

5.2 Device independent color

The technical color signal in printing reproduction is RGB or CMY(K) because these
signals can be displayed and printed respectively
'* 42426 The reproduction
equipment has until now been closed, computer systems including picture input,
processing, and output. This has meant that the transfer of images from one device to
another has not been possible. The need to transfer color image data from one system
to another has become necessary with the portable, microcomputer based systems and
writer-driven page and document manufacturing.
In television, color representation is device independent because of historical
reasons. The first step in the industry was the development and standardization of black
and white television. Then came the introduction of color television. The basic idea was
to use the same receivers. The black and white television sets show the color programs

109
CHAPTER 3

as black and white. The solution was device independent color. In black and white tele-
vision, one visual color component — luminance — already existed. This remained as
such with the introduction of the color signal. Addition of two color components provided
the full color picture. The color television signal is a visual or HVS signal. It cannot be
displayed as such. The normal television receiver converts it to the RGB signal for dis-
play. This visual color representation has two benefits in technical devices and systems:

- Color is easy to adjust interactively by a nonskilled person, because the

parameters are those easily seen and understood. Adjusting the color appear-
ance interactively by adjusting RGB or CMYK signal components demands
lengthy training.

- Color representation is device independent.

Input, RGB Transfer HVS

System 1 System 2
HVS color HVS color

Output Output
to display to print
RGB CMYK

Figure 47. Device independent color reproduction.

Figure 47 shows the procedure for device independent color reproduction. The
color separation signal that is the starting point of the multi color process is always an
RGB three filter signal. To compute one component of the visual HVS signal, all three
color separation signal values are necessary. The transformation from RGB to HVS is a
matrix operation

igh a\|R
Ve biG (18)
|S c| |B

and the transformation back to a technical signal for output

C d\|H
M\ = Vel) V (19)
Ne TAS

110
 Color and color images

6 Text processing and page make up

Other reproduction operations are the following:

- text operations and

- page makeup operations.

This section discusses these operations and the underlying principles.
Gutenberg made an invention around 1450 that allowed assembling metallic type
into letterpress printing. He did typesetting manually. Manual typesetting collects type
representing each character from type cases to a printing form. The printing form is the
printing plate in a flat bed letterpress. This was essentially the only text processing tech-
nology for over 400 years until the invention of a hot metal typesetter in approximately
1880. That invention mechanized hand typesetting, but the end product was exactly the
same — a printing form for letterpress. The technology dominated the “text industry” for
almost one hundred years. Then phototypesetting replaced hot metal typesetting.
In editorial text processing, the typewriter was the first technical step. The first
typewriter was constructed around 1857. It represented the only technology for over a
century. The first display terminals came to the printing industry in 1973 and to offices
about a decade later. In the office environment, computers totally replaced typewriters
in less than a decade. The only technical detail remaining from the original typewriter
technology is the order of characters on the keyboard called QWERTY. This term comes
from the six first characters of the second line from the top of the keyboard. The intent of
the original letter placement was to slow the typing speed to match the slow technical
speed of the first typewriters and to prevent typing adjacent letters simultaneously caus-
ing jamming. Now qwerty is a de facto standard that is almost impossible to change.
The steps of text processing include the“? following:
- input

- editing and correction of content information

- definition and embedding of structure information

- generation of layout information

- output.

Typing has retained its dominant position as the input method of text information
into the reproduction process. Optical reading or optical character recognition (OCR)
was popular with printers a couple of decades ago. Editors then prepared manuscripts
by typewriters. Printing houses retyped them into the reproduction system. Optical read-
ing seemed the perfect solution to the double typing problem. The actual solution came
from another direction. Editors started to use computer terminals for editorial text pro-
cessing systems. Then editors and freelance writers switched to microcomputers. The
problem of double writing does not exist anymore. Optical reading has some use to con-
vert old publications into databases. It is not much cheaper than rewriting because man-
ual proof reading and correction is necessary.

a
CHAPTER 3

Speech recognition as an information input method for communication is a possi-

bility. The present speech recognition systems may understand a few thousand words
with an error frequency of 1%. A feasible system (cheaper and faster than typing when
calculating costs of input, proof reading, and correction times ) would need to under-
stand ten times as many words and the error percentage would have to be approxi-
mately 0.1%. The enthusiasm for speech recognition as an input method to prepare
publications has also decreased.
Text is inputted normally as standardized binary text code, ASCII code. ASCII is a
7-bit character code (capacity 2’ = 128 characters) that includes text characters, num-
bers, and punctuation marks shown in Fig. 48. Because of the limited number of charac-
ters, there are different ASCII standards for different languages. Scandinavian ASCII
includes a, a, and 6. The 16-bit UNICODE is a possibility for an interlinguistic text code.
The code offers a capacity of 2'©= 65 536 characters. This code can include the kanji
characters of Chinese and Japanese. The transfer and processing time of text would be
double that of ASCII.

bf) D6 05) s04— Do. 2 pI

Code word

[Seer XeFoam | eens fy) ee

Scandinavian characters

Pscl0|
peorA

el > Pe ate
IcD4| $_| 4 | D |
%o
LSVN tees Ge eee
ICAN|(__| 8 | H |
EA |
SUB ceneiinespas
EGRET
PES=itaces24[ayy
cE Ges oe
Le ae
US eA ae
b1—b7: binary numbers
x/y: hexadecimal numbers

Figure 48. The ASCII text code.

112
 Color and color images

Editors and writers create content and do correction. Proof reading also occurs in
many processes as a productional step. Spelling programs for correction are common.
Writers have two basic philosophies in document preparation:

- interactive creation of the document appearance

- automatic layout.

Interactive document appearance generation is in computer terms the WYSIWYG

principle. This stands for “what you see is what you get.” It refers to a typographically
correct proof of the print on display. The WYSIWYG has achieved common use during
the last few of years. It places additional demands on the writer compared with simply
writing text. The writer must have information about the page structure of the publica-
tion, or the typography will require revision before publication. Automatic layout may give
to the writer ready WYSIWYG typography on the screen with instructions how to define
the different components with tags or only the tag orders and no typography on display.
The first methods used to define the document structure and typography were
typography coding. Typography coding is used by printers for the following reason:

- provide different structural components with tags

- define the character typography shown in Fig. 49.

Typography coding is partially process independent, but several typesetting

machines have their own coding. These typesetting codes developed to be move like
computer languages than a collection of code words.
Open systems and the use of the same content data for different output media gen-
erated needs for device independent methods to present the contents of documents.
Today, closed systems are still in use, but open procedures are increasingly used.
Structure generation and layout generation are separate from each other to
obtain multi purpose data bases. The structural information of a document does not
contain typographic information. It only defines different types of information such as the
following:

- title
- introduction

- normal text

- tables

- footnotes
- references, etc.

For the structural description of documents, device independent mark-up lan-

guages (ML) are used. The most common is Standard Generalized Markup Language
(SGML). These generate general purpose databases for different output media.

113
CHAPTER 3

Lay-out information determines the following:

- text typography (character faces and sizes)

- mutual location of text components, graphics, and pictures.

The layout information is coded by some page description language (PDL). The
most common is PostScript. A page description language may be specific to a device or
system, or it may be device and output resolution independent such as PostScript.
Development has headed for years to integrated text and picture production. Text
as output may be the following:

- photographs of typographically correct text

- , photographs of typographically correct pages including text and pictures

- database including content information, structural information, and layout

information.

Font measures Different fonts

Arial
Courier
Times
ALGERIAN
Character
beginning* Bookman
Braqgadocio
Se

>| k€Character space Brash Script

¢——_____—_—_>} Colonna
Character width Kidnap
Kino
Playbill
Font variables
Wide Latin
Font: Times, Arial, Courier

Face: plain, bold, italic, underlined

Size: e.g. 8-point, 10-point, 12-point, (72 points = 1 inch)

Figure 49. Text font variables.

The layout information is output using a computer to plate, computer to press, or

computer to print technology.
In a process producing text and pictures separately, the page makeup is manual
using “clip and glue” technology. In newspaper technology, both pictures and text are

114
 Color and color images

positive photographs on paper. These are glued in a page form. Assembled pages are
photographed in a reproduction camera to obtain page negatives. These are used for
offset plate exposure.
High quality printing uses photographs of text and pictures on film instead of paper
for quality reasons. Both positive (from positive photographs) and negative (from nega-
tive photographs) assembling is done. A page assembled on film may find use for plate
copy or an additional film step may provide even, flat page films without tape, etc.
Page makeup using computers may be totally interactive. This approach is closest
to the WYSIWYG technology. Completely interactive page makeup is not much faster on
displays than when done manually.
Automatic page makeup with computers is satisfactory when the layout is simple
as in products such as novels. Future development will gradually make automation of
page makeup of more complex products feasible.

115
 CHAPTER 3

MiReferenceess
1. Saarelma, H., Oittinen, P, Kuvatekniikan perusteet (Basics of Image Science).
Otakusantamo 842b. Otaniemi 1986. p. 101. (textbook, in Finnish)
2. Saarelma, H., Oittinen, P, Basics of Printing Technology. Otatieto 563. Otaniemi 1995.
Di2Za0,
3. Hunt, R.W.G., The Reproduction of color. Fountain Press. England 1987. p. 657.
4, Hunter, R.S., Harold, R.W., The Measurement of Appearance. John Wiley&Sons. New
York 1987.
5. Judd, D.B., Wyszecki, G., Color in Business, Science and Industry. John Wiley&Sons.
New York 1975. p. 500.
6. Laihanen, P, Color Science Approach to Digital Picture Reproduction. Dr. Tech.
Thesis. Otaniemi 1995.
7. Marr, D., Vision. Freeman. San Francisco 1982.
8. Nasanen, R., Effects of Halftoning Noise on Visual Perception of Spatial Signals.
Helsinki University. PhD Thesis. Otaniemi 1989.
9. Pinker, S., Visual Cognition. The MIT Press. Cambridge 1985.
10. Fairchild, M.D., Color Appcarance Models. Addison Wesley Longman, Inc. 1998.
p. 417.
11, Laihanen, et.al. Automatic Color Correction. 2nd IS&T and SID Color Imaginng
Conference. November 1994. Scottsdale, Arizona.
12. Oittinen, P, Saarelma, H., Elektroninen painatus (Electronic Printing). Otatieto 882.
Otaniemi 1994. (textbook, in Finnish)
13. Saarelma, H., Laihanen, P, Procedure and Apparatus for Maximizing the Visual
Quality of an Image in Electronic Form., International Patent Application
PCI/FI96/00339. 1996.
14. Saarelma, H., Kuvan ja tekstinkasittelytekniikka (Picture and text processing). Otatieto
578. Otaniemi 1997. p. 172. (textbook, in Finnish)
15. Inue, A., Tajima, J., NEC Research and Development 35(2):(1994).
16. Sddergard, C., Design and Development of a Set of Image Systems for Graphic Arts
Production. Helsinki University of Technology. Dr. Tech Thesis. 1994.
17. Yule, J.A.C., Principles of Color Reproduction. John Wiley&Sons, Inc. New York 1967.
p. 477.
18. anon., Proceedings and Seminars of SID 28 the Annual Conference, May 1977 and
previous editors.

116
 Color and color images

19. Widdel, H., Post, D.L., (eds) Color in Electronic Displays. Plenium Press. New York
1992. p. 335.
20. Kekolahti, P,, Digitaalinen rasterointi (Digital Halftoning). Helsinki University of
Technology. Lic. Tech Thesis. Otaniemi 1983. (in Finnish)
21. Saarelma, H., Image Formation in Halftone Photography. Helsinki University of
Technology. Dr. Tech. Thesis. Otaniemi 1979.
22. Ulichney, R., Digital Halftoning. The MIT Press. Cambridge 1987. p. 362.
23. Valovirta, A., Varin muodostuminen painatuksessa ja varivedostimilla (Color
Formation in Printing and Proofing). Helsinki University of Technology.
Lic. Tech. Thesis. Otaniemi 1994. (in Finnish)
24. Walowit, E. (ed), Device-Independent Color Imaging. SPIE Proceedings, California
2414:(1995).
25. Giorgianni, E.J. and Madden, T.E., Digital Color Management. Addison-Wesley, Inc.,
Massachusetts 1998. p. 576.
26. TAGA Proceedings 1997 and previous editions.
27. Anon., Encapsulated Post Script, File Format Specification. Adobe Systems
Incorporated. different versions.
28. Barnes-Lee, T., Hypertext Markup Lanquage. CERN Draft 1993. p. 35.
29. Barret, J., Reistoffer, K., Byte 12(5):171(1987).
30. Brushan, A., Plass, M., Computer 19(6):72(1986).
31. Kay, D., Levine, J., Graphics File Formats. Windrest Books 1992. p. 278.
32. Kay, M., The Open Information Interchange Report. Oll Spectrum 1(9):1(1994).
33. Kyongsok, K., A Future direction in Standardizing International Character Codes —
with Special Reference to ISO/IEC 10464 and UNICODE. Computer Standards and
Interfaces 14:209(1992).
34. Billmeyer, FW., Color Res. Appl. 12(4):173(1987).
35. Boynton, R.M., Human Color Vision. Holt, Rinnehart, Winston, New York 1979.
36. Granger, E.M., Heurtley, J.C., J. Opt. Soc. Am. 63(9):1173(1973).
37. Herwijnen, E. van, Practical SGML. Kluver Academic Publishers 1990. p. 307.
38. Herwijnen, E., van, The Open Information Interchange Report. Oll Spectrum
1(10):8(1994).
39. MacAdam, D.L., Color Measurement. Springer, Berlin 1981.
40. Tuuteri, L., Tunnistevarin automaattinen korjaus (Automatic Correction of Memory
Colors). Helsinki University of Technology. Lic. Tech Thesis. Otaniemi 1996.
(in Finnish)
41. Wyszecki, G., Stiles, W.S., Color Science, 2nd Ed, John Wiley, NY 1982.
42. Saarelma, H., Graphic Arts in Finland. 27(1):3(1998)

bé
BCHAIR _
Principles of printing

1 Printing in the Communication ChAin ...................ccesececececeeeneneeeseceseneneueeees 119

2 Printing production operations <y2es3;2s-ceee.c een. coecoerce ene eePee coe see ere 121
3 Printing ime@tinOdS 2. 5.crtcc.cacecre svecvercter teeter esse cece caenateee nce so: eens 127
OISCENICOS cereee8y cite ecctuns cite age uae toe RI ee a ea 131

118
 CHAPTER 4

BPrinciples of printing =
A discussion of printing can include the following approaches:

- printing in the communication chain

- printing production

- printing methods.

Discussions of communication often focus on evaluations and comparisons of dif-

ferent media — printed, broadcast, and network. The criteria in such evaluations can
come from various disciplines including technological, economical, sociological, and
ecological. Due to the many aspects and the inherent characteristic differences
between the media, no straightforward comparisons are possible.
Manufacturing type operations in the printing reproduction process encompass
printing, finishing, and the interfacial operations with the previous and subsequent steps
in the communication chain. Printing is undoubtedly the core step in making printed
products, and printing methods represent the principles on which printing relies. Paper
related discussions of printing often limit their focus to interactions of paper and printing
methods. The demands on paper are considerably more extensive. They extend to the
top hierarchical level, i.e., the communication process. When comparing products using
paper with other media products and services, a wide focus of discussion is necessary.
This chapter gives an introductory view of the common principles in the three
printing frameworks defined above.

1 Printing in the communication chain

Printing
in the
communication chain can Distribution Retrieval
: as prints as prints
have the role of producing ° P
products which then are
distributed physically or the Content : -
: ; ; F generation and Dissemination Retrieval
role of visualization in packaging
information retrieval as
Fig. 1 shows. Communication process
Distribution of prints
as products has been the Figure 1. Printing in the generic communication chain.
only alternative available
until recently. With data networks, distribution of content information as electronic data
has become feasible. This development underlines the need for technology applicable

119
CHAPTER 4

to distributed printing. This technology is digital printing. The following categories are
components of distribution:

- centralized printing: printing at one site with product distribution

- distributed printing (satellite printing or localized printing): data transmission to

the printing sites and local product distribution

- printing distributed to the end user: no distribution required.

Methods of the last named category are especially under development. A ques-
tion is how the different alternatives relate to each other considering the different criteria
such as cost and time. An approximate assessment of cost or time expenditure results
from adding the printing cost, data transmission cost, and product delivery cost.
Figure 2 illustrates the principle. The horizontal axis defines the degree of distribution of
printing, and the vertical axis is cost or time. Total efficiency changes in the opposite
direction to cost. With an increase in the degree of distribution, the data transmission
costs increase, the product delivery costs decrease, and the printing costs increase. A
variety of factors influence the behavior of the curve such as the size of the total edition,
the number of pages, and the level of picture quality. The availability of data transmis-
sion capacity, computers, printers, and finishing equipment at distributed sites and end
users are technically related boundary conditions for implementing distributed printing
commercially. It should be observed that cost and time expenditure are not the only rel-
evant criteria.
The distribution of
information from producers
to end users as products is
one-way or “top-down” com-
munication. Retrieval of the
information by the end-user
introduces a two-way ele-
ment to communication. The
change from a top-down to a OF
COST
TIME
Product distribution
bottom-up mode of opera-
tion reflects a very signifi- Data transmission
cant shift from a sociological
viewpoint. Globally Personal
centralized
When retrieving net-
DEGREE OF DISTRIBUTION
worked information by print-
ing, a shift in time or place Figure 2. The principle of cost or time expenditure in printing at dif-
usually occurs compared ferent levels of distribution.
with retrieval. An example is
the use of prints later or somewhere else. Retrieved prints can therefore represent a
mobile user interface to content information.

120
 Principles of printing

2 Printing production operations

Figure 3 shows the manufacturing operations included in printed media production. The
steps also include raw material and finished product handling and transfer steps. Film
and plate making occur between pre-press and press operations. When plate making
occurs in the press, it is a component of manufacturing. Ripping in electronic printing
often occurs in the printing device. Finishing is a manufacturing characteristic, and
dispatch is a logistical operation.

Printing production operations

Content
generation and SEricNe |Printing
packaging

Figure 3. Manufacturing operations in printing production.

Many different types of printing production are possible. Figures 4—6 define the major
types emphasizing the steps that are different and omitting those that are the same.
The term photocopying in Fig. 4 or copying for short means the information made
in multiple copies from physical pages called paper originals. In digital copying, the
page information is converted to digital form in the copying device for computerized
manipulation. For copying, it is converted back to analog signal or ultimately to an opti-
cal signal. Optical character recognition (OCR) allows the interpretation of text blocks as
text instead of images of text. This allows proof reading and correction. From this
description, digital copying obviously combines digital scanning and printing.

Digital Photo-
Paper
pages copying
originals

Paper
Originals
Fax
transmission reception
Digital

Figure 4. Copying (top) and faxing (bottom).

121
CHAPTER 4

A facsimile or fax device is an analog copier or a digital copier. Two fax devices
communicate with one another. The term fax implies the use of a given data format for
transmission. In digital faxing, the transmission is between two computers with printing
of the pages as desired. Fax-on-demand is a service prompted by a telephone call to a
server computer that stores the information requested. A fax-back program launches
transmission of the requested document. The trend to embedding processors in net-
work linked peripherals such as scanners and printers will enable them to communicate
directly without the intermediate role of separate computers. This development means
that separate fax devices will become obsolete.
The electronic or digital printing in Fig. 5 is printing directly from computers. The
technology has developed from making single copies — from a computer output or sys-
tems printing method — to making multiple copies (“mopying’”). Digital printing environ-
ments range from homes and offices to commercial printing establishments. Digital
printing is gradually replacing the procedure of digitally printing one copy and photo-
copying multiple copies.

Digital Digital
pages printing

Digital Digital ,

Digital Digital Master Mounting Mechanical

pages printing making on press printing

Figure 5. Use of digitally printed pages.

Digitally printed pages also find use to make plates for mechanical printing. In this
instance practiced in quick printing environments, the plates are typically paper based
masters for printing small editions with low quality requirements by the offset lithography
method (cf. Chapter 5). With an increase in the performance of digital printing, this pro-
cedure is gradually losing its popularity.
The conventional procedure in mechanical printing production includes exposure
of the page information on light sensitive film to make plates for printing as Fig. 6 shows.

le2
 Principles of printing

Digital Computer- Plate Mounting of Mechanical

pages to-film making plate in press printing

Digital Computer- Mounting of Mechanical

pages to-plate plates in press printing

Digital Computer- Mechanical

pages to-plate in press printing

Figure 6. Steps in mechanical printing.

With the development of computer plate making methods, the film making step
may not be necessary. Quality reasons can justify it if several plates are necessary from
a film due to large editions or several print runs occur at different dates. Mounting plates
in the press separately for each job becomes obsolete when making the plates in situ
called on press printing. It requires plate material supplied from a source residing within
the press or reusable plates. Comparison of Figs. 5 and 6 shows that the operations of
digital printing do not include a plate making step — the methods are “plate-less.”
Digital and mechanical printing represent different scales of print production. The
concept of scale includes the following:

- speed [linear speed (ms7!) and page speed (ppm, rph)]

- area that can be printed one-sided (simplex) or two-sided (duplex) with a run

- number of pages in a product that the printing operation supports

- size of the edition that can be feasibly printed.

In a small scale, the speed and size of the printable area are limited, and the
robustness of the devices does not support continual operation. This is because the
devices and their parts have a short life expectancy under continuous running condi-
tions. In the digital printing area, the principle of capacity multiplication by linking several
low volume printing devices has been successful. In printing directly from computers,
the data processing and storage capacity may limit the number of pages of a single job.
In web fed digital printing, the length of the print in the machine direction may be virtu-
ally infinite.
Figure 7 illustrates by order of magnitude the feasibility of different types of print-
ing using the number of pages in a job and the size of the edition. The types of printing
shown include different variants of digital printing and mechanical printing.

123
CHAPTER 4

O 8
e 10 Variable
information
<= printing
seg
i) AG
® Low end
x digital
LL Mid range
O 10*
Oo digital igh end
LU digital
jaa)
Distributed mechanical
=
=) Centralized mechanical
ZZ

10° 10° 10* 40° 10

EDITION
Figure 7. The scale of printing expressed as practical ranges of the number of pages in a job and the edition..

It is evident from Fig. 7 that the highest end digital printing has a smaller scale
than distributed mechanical printing. With technological development — digitalization of
mechanical printing and increase of scale of electronic printing methods, this is likely to
change. Variable information printing (VIP) is printing in which the edition size is usually
one — all pages are different, but the total number of pages printed can be very large.
Example include personalized forms.
In digital printing, the printable area is often the same or double the size of the
product. In mechanical printing, it is a multiple of the size. In the former, the printed
pages exit from the devices in the order of the page numbering. This means that digital
printing allows printing of single copies of a product. It also means that binding the
sheets into a finished product does not require collation and can potentially be inte-
grated in the printing device. Mechanical book manufacturing lines share some of these
features. Their area of application is volume production rather than single production.
Figures 8—10 show these differences.
In a digital printing line that integrates finishing, one copy of a product prints while
another is finishing. If the size of the printed area is different from the size of the prod-
uct, the sheets require folding before or after finishing. Printing can still be in the order of
the page numbers.

124
 Principles of printing

Printing Printing Printing

nN pages n pages n pages

1 copy 1 copy 1 copy

Finishing Finishing Finishing

>
TIME
Figure 8. Typical production procedure in digital printing.

In mechanical printing, the sheet size — same as plate size for the number of
pages — is typically much larger than the product size. Multiple pages are assembled on
a press sheet and printed simultaneously. Assembly of the pages follows an order that
guarantees correct page numbering after collation and folding. This is imposition. The
number of pages on the press sheet even considering both sides of the paper is smaller
than typical page counts in most products. With the exception of newspaper and high
volume magazine and book printing, the presses allow only one “in feed” of paper. This
means printing a product requires several! runs with different plates. In a given run as
many copies as necessary are printed. Figure 9 illustrates the production procedure.

Printing Printing Printing

k pages k pages k pages

m copies m copies m copies

Run n/k

Collation

TIME

Figure 9. Common production procedure in mechanical printing with n = number of pages in the product,
k = number of pages on plate, and m = size of edition.

125
CHAPTER 4

Plate making off-press can occur concurrently with printing, but the post-press
operations can only start after completing all runs.
In newspaper printing, the presses have as many parts and paper feeds as nec-
essary for printing the product with a run. Figure 10 shows the production procedure.

Plate
making

n/k plates Printing

k pages
m copies

Printing

k pages
m copies

Printing

n-xk pages
m copies
Collation Finishing

TIME

Figure 10. Production in newspaper printing for x + 1 total number of printing units.

Finishing and dispatch convert printed paper into products ready for delivery. Fin-
ishing is the physical equivalent of packaging of the information contents. It is mechani-
cal by nature and does not increase the information content of the product. Table 1
shows the most common finishing operations.

Table 1. Finishing operations.

Purpose
To produce product size from the sheet size of printing
Cutting To remove pages of a copy from a web, to separate the
pages and to trim them i
Collation To arrange the pages in the order of page numbering
Binding (stapling, adhesive binding, To attach the pages together to form a block
thread binding)
Case (cover) making To make a cover from its components and print it
Casing To combine the pages and the cover &)

126
 Principles of printing

Subscriber specific tasks in dispatch may include customized product assembly,

personalization, and addressing. Figure 11 illustrates the use of subscription data in dis-
patch. Future technical development will allow subscriptions to influence production fur-
ther left in the chain of Fig. 11. This will allow contents packaging or even creation
according to a client’s subscription. In such a case, the steps in the process chain would
merge. Current paradigms of news delivery in data networks support such workflow in a
limited manner.

Personal and address data

Subscriber/ | Subscription
client/product| Personal and address data; VIP
data base Subscription

Content Assembly Personali-

creation and Printing of segments zation
packaging Inserting Addressing

Figure 11. Subscriber specific dispatch operations.

Due to tightening delivery schedules, product mailing lines integrate with printing
lines.

3 Printing methods
Chapters 5 and 6 will discuss printing methods in greater detail. The methods belong to
the generic class of two-dimensional visualization methods. The class includes soft
copy and hard copy methods. The former are display methods, and the latter are
photographic and printing methods. Holography — a three-dimensional visualization
method produced by printing — lies outside the scope of this discussion as does two-
dimensional printing for uses other than visualization. Manufacturing of electronic
components falls in this category.
Figure 12 illustrates visualization methods in an attempt to show how their com-
plexity differs when considering the number of steps.

127
CHAPTER 4

Energy
Energy transfer to Energy Energy Energy
transfer ink which transfer transfer transfer
to display is transferred to paper to receptor to receptor
to paper

Display Direct methods;

technology e.g. Inkjet,
thermal Toner
transfer Deven iient development Development
in paper
on receptor

Imaging in paper;
e.g. silver based
electrography,
thermography Toner transfer Ink transfer
to paper to receptor

Receptor methods;
e.g. electrophoto-
graphy, electron Ink transfer
beam imging, to paper
magnetography

Display meres Non-impact methods 1| Mechanical l

methods

Figure 12. Steps in soft copy and hard copy methods.

The diagram distinguishes display methods, nonimpact methods, and mechanical

methods. Display methods and direct nonimpact methods have the smallest number of
steps. In direct nonimpact methods such as ink-jet and thermal transfer, the print is gen-
erated directly on the paper. Ink-jet is also a noncontact method, since no part of the
printing device other than ink contacts the paper at the moment of ink transfer. The
same is true for the methods that are noncontact where the print is generated by reac-
tion in the paper instead of ink transfer. Photographic methods fall in this category. So-
called receptor methods that are indirect methods constitute a large group of nonimpact
methods. Electrophotography is the foremost representative of this group.
Another way of distinquishing display methods from printing methods is to say
that the former are reversible and the latter are irreversible. Reversibility of printing
means the ability to switch the ink on paper on and off as desired without destroying the
sheet as in deinking. Development in this direction is in progress under the terms “elec-
tronic paper’ and “electronic ink.”
New inventions and the long-term development of known printing methods strive
to incorporate display type functionality. Display technologies on the other hand concen-
trate on the ability to handle them and their low energy requirement for printed matter as
targets of development.

128
 Principles of printing

Accomplishing the printing steps of Fig. 12 requires auxiliary functions*. These

include feeding paper and ink, transportation of paper through the printing device, and
take-off. Print drying is also an auxiliary function. Display methods obviously have no
material handling operations.
Paper feeding to a printing device can use sheets, web, or a continuous form as
Fig. 13 shows. Noncontact methods such as ink-jet and silk screen printing also allow
printing of irregularly shaped objects. Sheets may be taken off as such or as ready —
made products in the case of integrated finishing.

Printing substrate
at input

Object
(article)

Board 3 Plastic and

>150...200 g/m metallic films

Collection of Collection of
Collection of
Creasing webs, gluing, webs, cutting,
webs, cutting, and
and die Rewinding perforation, and folding, and
folding
stamping folding binding

Object
Printed matter
(article)

Printing substrate
at output

Figure 13. Input and output of paper in printing.

Paper in web form may be delivered as cut and folded sheets (signatures),
rewound on a reel, or in continuous form. In the two last cases, the paper may be
printed again or undergo some other conversion operation. In business form manufac-
turing, webs may be converted into continuous forms in a mechanical printing step. A
subsequent printing with a computerized printer would use continuous input.
The number of pages in sheet-fed printed products is commonly a multiple of the
number of pages that a run can print. The products usually have separate covers. These
product characteristics indicate that finishing in-line with printing is not practical. Deliv-
ery is necessarily as sheets. Finishing is also not necessary in package printing applica-
tions where further conversion operations occur when filling the packages in another
process. Finishing in-line with sheet printing is not as easy as in web printing. This is

129
CHAPTER 4

because web tension controls the transport of webs. Operations such as cutting and
folding in the direction of running are easy. To move sheets at press speed requires
holding the sheets by separate elements called grippers. This complicates simultaneous
finishing operations.
In web fed printing, the paper can be cut and folded in the press to the final prod-
uct size and even bound with staples or adhesives such as newspaper or pocket book
printing. Rewinding is practical in packaging applications.

130
 Principles of printing

iRefercnces as
1. Kipphan, H., Status and Trends in Digital Multicolor Printing. |IS&T’s 13th
International Conference on Digital Printing Technologies. IS&T Springfield Virginia
USA 1997. p. 11.
2. Saarelma, H. and Oittinen, P, Basics of Printing Technology. Otatieto no. 563,
Espoo 1995. p. 232.
BOHR ——
Mechanical printing

1 Printing Me@UhOGS® os. ccs serene yesceeee eaedeseecen een ene te eae noe ater sae eee renee 133
2 Using the metious 22% <n ccssccbectacn occas peanecsoroneetaee scenes ete ee een 139
3 Printing plates s2.2c8c ee oe wats eva soseemas ae are ae eee eae amen ane zene eee eee 140
ed imPSUCTN TAAL cesses cs eseeeter es ose acee ove ete onan ected age tec SE oe ee 140
See © LTMOGIADEG: DIALCS i secs seh oe aes Geo weet cesarean tebe am re me ee a ee 142
Oe VGTAVUTE: CYIINCONS <.Sects ceva cevese et ccxedaaecuaga acetic aseehuine ake’ vee ana tees ca eer ee 144
Oe “SEleKOgraphie: Plates. Sek xwed ates. n ete tantah a0 oss Oat cae setecee er 146
4 Printing nips PHENOMENAL se soc ie eadeter cae yay cscs seacnaulenceeet hc easceteae a eee 146
44 INK tranSter to: tN@ slates sx. ec cers eeaacet nds van eteaecs hanes eteece orate twine ee 147
A 2 coe iiksthan Siete tO The MADOr.ass xcieot exeesoeett atau neers ee 150
5 Drying of Prints ieee shee sesso eer eee es re eee 156
Sale =MechalisimcorSolvent-evaporatiOm: cscevecsucscsceeseseeseseece soe semen eee eee 158
52s» (Mechanisimrorctrenical dis inGicsasccsccseuersses-ccceeesee testace eevee se 159
6 Printing: PreSS@S 22.50.20. 5: 7-420. nonencsey acc consannwesanetoceuemauave Reeseeciessea enna aes 159
Gil GON STal ASPOCUS i cxise cienctaces ascccaek vs naam vetese cane aesea ae eee ete ce ee 159
G2 ar PG UUMUNFNCOSUDIMIES Se sheeracc. acre Soe secretin ies ox ess caetacs cetera cece eee are eve 160
GiBe, mE TOSSHOUINIALS eetiecss or vcae exc are tell ai eat es eau ck eetcl ee ene an 163
if Finishing in: Webifed: printing), Asccacoesccccesceanctcseccnceecceosees ne eee eee eee 166
FRETGREMG Sax fe ceece rt ohacct or aresesee caste sare heeded aa DN chee sear orate nr 169

132
CHAPTER 5

Mechanical printing =
Mechanical printing methods are a category of impact printing methods. Other impact
methods include matrix and daisy wheel printing. Mechanical methods differ from the
other impact methods by using a separate impacting surface — the printing plate — for
each job. In the other methods, the impacting surface composed of fonts or dots is
permanent. The term “mechanical” expresses the fact that the energy in image
formation in the printing step is pressure. The term distinguishes conventional printing
methods from computer output. These are the electronic or digital printing methods
discussed in Chapter 6.
The main factor that defines mechanical printing technology is the printing
method’~%. Section 1 of this chapter explains the principles of letterpress, offset lithog-
raphy, gravure, flexography, hot stamping, and silk screening. The methods differ in dif-
ferent representations of text and image information on the plate — the type of latent
image. Section 3 provides an overview of plate making. The subsequent sections deal
with the mechanisms in nips (Section 4) and drying (Section 5). The type of paper feed;
the type, format, and number of printing units; drying of ink; and the type of finishing
operations integrated in the press line identify printing press constructions. Section 7 of
this chapter discusses these major items.

1 Printing methods
Figure 1 gives a genera! imaging view of mechanical printing. In a narrow sense, the
imaging steps during printing include transfer of the ink to the plate, its transfer to the
receiving substrate or paper, and print drying. A broad definition also includes the latent
image or plate making step.

133
CHAPTER 5

Latent image on
printing plate

Ink transportation and Ink application to

Inking; unit; Rebs
metering printing plate

Paper feeding Paper transportation Ink transfer to Paper

devices and registering paper

Figure 1. The imaging steps in a press (top to bottom), supporting functions (left), and raw materials
(right) in printing.

Mechanical printing
uses the rotary principle. This
means that printing in! trans-
Impression
fers to the paper in a nip cylinder or
between two rotating cylin- offset
ders. Figure 2 illustrates this. cylinder
Ink transfer between flat sur-
faces or a flat surface and a Paper
roller is also possible. The Printing w
speed achieved is less than in cylinder or Printing nip
rotary printing which has offset
cylinder
made them virtually obsolete.
Printing pressure acts
in the direction perpendicular Figure 2. The printing nip is the contact zone between two cylinders
to the cylinder axis of the through which the paper travels (not to scale).
rotating cylinders. Ink trans-
fers to the paper in the printing nip from the printing cylinder — plate cylinder — or froma
separate transfer cylinder. The configuration in which the ink transfers to the paper from
the printing plate using a separate transfer cylinder — the offset or blanket cylinder — is
the “offset principle”. It has universal use in lithographic printing. The method is offset

134
 Mechanical printing

lithography or offset. Direct transfer lithography — a little used method — is dilitho. The
offset principle can in principle also be applied to other methods. For instance some
electrophotographic and solid ink-jet devices use it.
Ink resides on the printing plate spatially distributed according to the printing
(image) and nonprinting (nonimage) areas. The printing area consists of alphanumeric
text, graphics, and halftone dots. The dots have the smallest linear dimension. They can
be less than 20 um. Figure 3 shows a classification of printing methods by the type of
printing area.

Latent image on
printing plate

Depth modulation Height modulation Surface chemical poy

modulation modulation

Packaging Publication Intaglio Flexo- Letter- Hot Lithography Silk screen

gravure gravure graphy press stamping

Figure 3. Classification of mechanical printing methods.

In letterpress, flexography, and hot stamping, the printing and nonprinting areas
form a geometrical profile in the z-direction of the plate surface as Fig. 4 shows. The
printing areas are above the mean depth of the profile, and the nonprinting areas are
below. Letterpress and flexography or flexo differ in the characteristics of the plates and
the inks. In flexo *°, the plate profile is higher, and its polymeric material is softer than in
letterpress. Flexographic inks are two orders of magnitude less viscous than letterpress
inks. Under nip pressure, a softer plate conforms more easily and achieves contact with
a rough printing substrate. This process is necessary for ink transfer to occur.

_- Ink

Halftone dot

Figure 4. The principle of letterpress and flexography. Magnified z-direction view of printing plate.

135
CHAPTER 5

Squeezing of the inks toward the

edge of dots or characters gives distinct
features for recognition of flexography
and letterpress. Figures 5 and 6 illustrate
such appearance.
Hot stamping is a decoration
method in which ink coated on a film trans-
fers by heat and pressure to a receiver
sheet. The raised parts of the profile con-
tact the film, and the resulting heat flow
causes liquification of the ink. Thermal
transfer printing, a nonimpact method, is Figure 5. Letterpress printed character.
a digital counterpart of hot stamping.

Figure 6. Flexoprinted halftone area.

In the gravure tech-

nique, cells — a negative | Gravure (area modulation)
relief engraved or etched in
a copper layer on the print- ae as ee |
ing cylinder — form the print- is
ing area. For better
durability, copper cylinders Gravure (depth modulation)
receive a subsequent Si 7
chrome plating. Figure 7
shows that different levels of
darkness result in halftone Ng Ink
image printing by varying the
area of the cells or by vary- Figure 7. The two principles of gravure — area modulation and depth
modulation — showing a magnified z-direction view of the printing
ing the depth of the cells.
cylinder.

136
 Mechanical printing

The technical procedures of cylin-

der making, however, lead to some varia-
tion in area and depth. Note in gravure
that text characters are also cells as ——~ cell structure
Fig. 8 shows. If the engraved areas were-
the size of individual characters, surface
|
~~ character
tension could not keep the very fluid gra-
vure ink within the cells. Running of ink
Figure 8. The principle of cell structure in gravure.
and smearing of the print would result.
A form of gravure called intaglio printing uses viscous inks. This technique allows print-
ing characters engraved on cylinders without halftoning.
As mentioned above, mechanical
energy is the main source of energy to
transfer ink from the printing plate or off-
set blanket to the paper. Gravure also
uses electrical energy®. This occurs from
a high voltage source that generates a
field between the impression cylinder
and the printing cylinder and is called
electro assist. The electrical potential
induces electrical charges in the ink and
in the conductive rubber surface of the Figure 9. Gravure printed character.
impression cylinder. As a result the ink is
drawing upwards from the cells especially at the edges, i.e the meniscus is concave.
This aids contact with the paper surface.
The cell structure is often visible in gravure printed text as Fig. 9 shows. Halftone
dots have the appearance of a doughnut due to the concave meniscus.
In letterpress, flexo,
and gravure, a printing nip
prints one side of the paper.
The uneven pressure gener-
ated by pressing two profiled
plates together prevents
simultaneous two-side
“ printing.
The printing area in
offset lithographic printing is
on the plate as a surface
chemical pattern. This
means that there is little
height difference between Figure 10. Gravure printed halftone area.
the printing and the nonprint-
ing areas. The process functions because of a greater affinity between the printing
areas and the printing inks than between the nonprinting areas and the printing inks.

137
CHAPTER 5

Figure 12. Offset printed characters. Left: coated wood free paper, middle: LWC paper, right: newsprint.

Using highly viscous inks

hamper spreading of the
printing ink to the nonprint-
ing areas. Water called
dampening or fountain water
is also applied to the plate.
Water adheres on the
nonprinting areas. Continuing
development is making
lithographic printing without
water possible. This issue is
further addressed in Section
4.1 of this chapter. In offset
printing, both sides of the Figure 11. Offset printed halftone area.
paper can be printed
simultaneously. This is blanket-to-blanket printing or perfecting. The even surface of the
offset cylinders aids two-sided printing.
Compared with the other printing methods, offset printing does not have distinc-
tive properties as Fig. 11 and 12 show. Letters are uniform in density and their edges
are smooth. Defects only become visible under magnification. Better quality gives
higher magnification that is necessary to reveal nonuniformity.
Silk screen printing or
serigraphy is not a contact Doctor blade
printing method like the
other mechanical methods. Ink
Figure 13 shows the princi-
ple of silk screen printing. A
stencil replaces the printing
Printing area
plate. The stencil has differ-
ent porosity in the printing
Figure 13. The principle of silk screen printing.
and nonprinting areas. Ink is

138
 Mechanical printing

pressed through the stencil to the paper or other substrate positioned below the stencil.
Treating the stencil by photochemical means makes it porous in the printing areas but
not in the nonprinting areas. The term silk screen dates from the time when the stencils
used silk. Today metallic and polymeric stencils are used. The term screen refers to the
filter function of the stencil.

2 Using the methods

Presses using a given printing method can differ from each other widely. Press
construction is determined by the type of products for which it has been designed. The
major press variables are paper feed, paper take-off, and dying. Table 1 shows the
major classes in different types of printing for different types of presses. The table also
indicates the primary product categories for each type of press.

Table 1. Different types of printing.

Printing method Paper in Paper out Drying

Letterpress Web Newspaper No drying
(newspaper printing)

(books, business forms) Sheet Sheet No drying

Web Form No drying
Web Book No drying

(small scale utility printing) Sheet Sheet No drying

Newspaper offset Web Newspaper, book, | Cold set; no drying (normally)
(books, directories, business forms) form
Heat-set offset Web Cut & folded Hot air drying
(magazines, catalogs) sheet (=signa-
ture)
Sheet fed offset Sheet Sheet No drying or UV-drying or
(special interest magazines, IR-drying
advertising, books)
Gravure Web Signature Warm air drying
(magazines, catalogues, packaging) |Web reel (Packaging)
Newspaper flexo Web Newspaper No drying
Packaging flexo Sheet Sheet No drying (Absorbent sub-
(solid board, corrugated board, Web Web strates)
paperback books, plastic films) Warm/cold air drying
Silk screen Sheet Sheet Different drying methods;
Web Web no drying/UV/IR
Object Sheet
Object

The names of the offset printing methods, newspaper offset, heat-set offset and
sheet fed offset, reflect the importance of paper feed and dying. Newspaper offset print-
ing is also cold set offset. This refers to the absence of external drying capacity in the
presses. Web fed magazine offset is heat set. Drying is by hot air. The type of paper

139
CHAPTER 5

feed relates to the scale of printing and the basis weight of the substrates. Large scale
printing and printing on low basis weight substrates is web fed.
The type of drying technology in Table 1 correlates to the demands that the print-
ing method places on the ink properties and the desired quality level of the final product.
Inks containing a large proportion of solvent such as gravure, flexographic, and
heat-set offset inks typically dry by evaporating the solvent by air flow. Depending on the
temperature of the air, drying may be “hot air’ as in heat-set offset, “warm air’ as in gra-
vure, or “cold air” as in flexographic printing. The drying mechanism can also be a func-
tion of the ink as in oxidizing and polymerizing inks. These are called — “high quality”
inks. Polymerizing inks are mainly used in sheet fed offset, but it is likely that their feasi-
bility to be used in web fed printing will increase. News inks do not dry at all during the
one day life time of newspapers.

3 Printing plates
3.1 General
In mechanical printing, printing plate make-up links the information processing and
printing steps of the printing production chain. Plates can be made either from digital
page information or from page film or paper originals as Section 2 in Chapter 4 explains.
A separate industry makes the printing plates — the base material in plate making for
printing. The companies have their expertise primarily in polymers and light sensitive
systems. Gravure cylinders are an exception. Printing houses copper plate and recycle
them.
Imaging using printing plates has two steps. The first is transfer of the text and
picture information to the plate. This latent image formation only occurs once for a job.
The second step is the transfer of ink via the plate to paper. This physical image forma-
tion occurs many times.
The first imaging operation is commonly one of the following:

- radiation sensitive

- mechanical

- nonimpact printing.

Radiation sensitive plate making methods include light and UV-sensitive photo-
graphic and photomechanical methods, and IR-sensitive thermal methods. Photo-
graphic methods use silver halide coatings. In photomechanical plate making, the
radiation sensitive coatings are organic photopolymerizable materials. After exposure
and possible development, a mechanical operation such as mechanical removal of
exposed or alternatively unexposed material occurs. Thermal methods use IR-sensitive
monomers. “Mechanical” in the context of plates means that the printing surface results
from engraving or some corresponding operation. In nonimpact printing, the printing
and nonprinting areas on the plate are generated by putting material on the plate.
Figure 14 defines the basic categories of plate making in three generic steps —
modulation of plate properties, selective allation, and ablation. These steps may include
several operations. Typically, a smaller number of steps in plate making requires a larger

140
 Mechanical printing

Modulation of
plate properties Allation Ablation

Radiation sensitive
Radiation induced chemical Development, removal
Photographic reaction in radiation sensitive >| of exposed/unexposed
— lithographic coating ® coating @
Photomechanical
— lithographic
— gravure
— flexographic
— letterpress
— silk screen

Thermal

Mechanical
— gravure Laser/EB/diamond
— flexographic head engraving ®

Non-impact
— lithographic Light induced changes in Toner
photoconductivity ® ? development @

Ink-jet printing,
reaction

Laser/heat induced
coating transfer,
reaction

Figure 14. The basic process of plate making. The numbers refer to the order of steps.

energy in the first step and vice versa. In masking methods such as making multi metal
plates — discussed later — or in silver masked methods, the number of steps is a multiple
of the number of the basic steps. This allows more versatility and increases the degrees
of freedom of optimization.
Modulation of plate properties means external energy such as radiation forms an
image of the printing pattern on the plate by varying some characteristic of the plate. For
this to occur, the radiation and the plate properties must match. In photographic and
photomechanical plate making, radiation induces chemical reactions. These are irre-
versible meaning disposing of the plates or recoating them for further use. The use of
photoconductivity as in nonimpact electrophotographic plate making is an example of a
reversible change. A goal of plate research is to develop materials whose properties
can be activated by external energy reversibly for direct use as printing surfaces.
Allation refers to selective addition of material on a plate surface to form a printing
pattern. Nonimpact printing methods such as toner development, ink-jet printing, or
thermal transfer can accomplish this. Satisfactory durability of plates requires that the
inks undergo some chemical reaction with the receiving plate surface or are post-
treated by heat or radiation. Practical implementation of plate making may also include
nonselective allation and ablation as explained below.
Ablation involves selective removal of plate material. It is the single step in
mechanical engraving and the second step in photomechanical plate making.
Figure 14 suggests that photomechanical methods are the most versatile type
because they have application in several printing methods including lithographic and

141
CHAPTER 5

flexographic printing. Direct engraving is the standard method in gravure and also finds
application in flexography. Nonimpact methods for lithography are in the early stages of
development.

3.2 Lithographic plates

The base material in lithographic (offset) plates is aluminium. Polymer and paper based
plates have application in small format printing of short runs. A coating on the
aluminium base functions in printing as the ink receiving surface, and the pure
aluminium — usually in oxidized form — acts as the water receiving surface. The
durability of a plate depends on the durability of the coating and the hardness of the
grained and oxidized aluminium. The run length of a conventional, photomechanically-
made offset plate is typically 10°-10° impressions depending on the construction of the
plate and the mechanical condition of the printing press.
A consequence of the variety of plate making processes applicable to offset
lithography is that there is variety also in the implementation. Figure 15 shows that litho-
graphic plate making can use the following:

- exposure in a frame as in a “computer to film” operation

- direct exposure from digital data as in “computer to plate” operation

- direct exposure from digital data with the plates mounted on the press as in
the “computer to press operation.”

When the exposure occurs in a frame, the intensity of the explosing radiation is
modulated as the radiation passes through a page film positioned between a plate and
a light source as Fig. 15 shows. Exposure of the total page occurs simultaneously.
Lithographic plates operate negatively or positively. The definition depends on the
type of photosensitive reaction that the exposure causes. In negatively operating plates,
exposure causes hardening of the radiation sensitive coating. In positively operating
plates, the reverse occurs — the coating softens and becomes soluble in a developer
solvent. The user can remember this easily by knowing page negatives are used to
expose negative plates and page positives to expose positive plates.
After exposure, a plate undergoes manual or automatic processing followed by
manual mounting or robotic mounting around a cylinder in the printing press. Process-
ing includes development to finalize the reaction initiated by exposure and fixing to
remove unexposed or exposed — depending on the type of reaction — material from the
surface. Plate making that uses exposure in a frame is “conventional offset technology.”
In “computer to plate” making’, Fig. 15 shows that the plates undergo serial
imaging by scanning. This means that the total time of plate exposure is the product of
the time needed to expose a single pixel and the number of pixels per page. Page size
and pixel resolution determine the number of pixels per page. To keep the total expo-
sure time short, the energy needed to expose a pixel should be as low as possible. This
is why the plates are more sensitive than the conventional plates. Plates that would not
require any treatment after exposure are under development. Computer to plate tech-
nology has reached the commercial stage after a lengthy development phase.

142
 Mechanical printing

The computer to press

technology of Fig. 15 obso- Plate exposure in frame
letes the separate plate
making step. It is an answer
in offset printing to the chal- / Page film as original
(product of text and
lenge of nonimpact printing image processing)
for short runs”. One advan-
tage is color registration
since the plates do not move Even exposure

after exposure. A disadvan-

tage is that the press is ina T+ — Page film
stand-by state while making ——— oa Light sensitive
| vo printing plate
the plates that causes a loss
in productivity compared to
Computer to plate
off-press plate making. Mak-
ing all the plates can occur
= _| Computer | —
simultaneously. Computer to — = _ Exposure
press technology is a “pre- ee a \_) head
step” to an on-line interface _ Disc iy) ——,, Scanning
Prom ay plate
Page as ey
between computers and con- data base CSU
ventional printing presses.
The technology of the first Computer to press
generation uses ablation
methods. Nonimpact printing =—— |Computer | SS SSS

based plate making also has aaa KC) .@)

potential because the height Disc
differences on offset plates Page as
aoiainace
a eTOffsetRTEpressaT
are insignificant. This means
the amount of material requir- Figure 15. Printing plate exposure in a frame, computer to plate,
ing deposition on the plate is and computer to press. Offset plates and polymeric letterpress and
small. Nonimpact printing flexo plates use exposure in a frame and computer to plate. Only off-
methods can transfer small set plates now use computer to press exposure.
amounts in a spatially con-
trolled manner. To achieve durability, post-treatment operations are necessary. This
restricts the reuse of plate material. Resolution is still critical today.
The light sensitive coating on a conventional offset plate is usually a photopoly-
mer or a diazo material. The exposure uses radiation in the UV and visible wavelength
ranges. The broad class of materials called photopolymers also includes species that
are sufficiently sensitive for use in computer to plate imaging. Besides photopolymer
based imaging, a variety of other principles find use for plate making from computers.
Additional materials are under development. Photographic methods as a class have the
advantage of high sensitivity. Thermal methods have the potential of not requiring any
development step after exposure by IR radiation. They also offer the advantage of day-

143
CHAPTER 5

light handling that plates with sensitivity in the visual range of wavelengths do not have.
Nonimpact based plate making technology offers erasibility in the press and reuse.
Electrophotographic methods of plate making have been available for a long time. Ink-
jet and thermal transfer methods are new developments.
Erasability and reuse could facilitate variable information printing by allowing
change of data at every revolution using the offset lithographic method. To make this
possible, plate making and plate conditioning for reuse would have to be accomplished
at press speed. This means that plate making would have to occur in the time scale of
milliseconds.
Masking methods refer to the category of plate making methods in which imag-
ing occurs in two steps. As mentioned above, this offers more freedom in matching sen-
sitivity and durability to the requirements of plate making and printing. Typically, plates
for masking methods consist of three functional layers. “Ordinary” plates only have two
layers. In the first step of plate making, an image is generated on the top layer. The
image protects an underlying layer that experiences exposure or etching in the next
step. The image may also be generated by nonimpact printing such as ink-jet and ther-
mal transfer.
Multi metal offset plates or bimetal plates also use the masking method. The
plates consist of two metals and a radiation sensitive coating. One metal has hydrophilic
properties, and the other metal has oleophilic properties. In printing, hydrophilic areas
are water receptive, and oleophilic areas are ink receptive surfaces. The information to
be printed is exposed on the light sensitive layer that polymerizes. In the second step,
the polymerized layer controls the etching of the underlying metal layer. After etching,
the polymer layer is washed off. Multimetal plates offer high durability.

3.3 Gravure cylinders

The printing areas on gravure cylinders have cells that meter and transfer ink to the paper.
Engraving usually generates the cells, although etching is also possible. Figure 16 shows
the processes. After engraving or etching, gravure cylinders receive a chrome plating to
improve surface durability. A chromium plated gravure cylinder can print millions of copies.
Gravure is therefore the printing method for the longest runs. Gravure cylinders are
recyclable by treating the surface of the cylinder after printing for a new engraving.
Engraving machines are special scanners equipped with diamond heads. For
engraving, a copper cylinder is mounted in the machine as the output drum. A computer
may feed Information directly to the engraving heads, or it may come from a page film
mounted on an input drum. The former method finds increasing use. The scanners can
read continuous tone page films and convert the signal to instructions for the engraving
heads. Using halftone films made for offset is also possible, although the gravure dot
structure is not similar to the halftone dot structure in offset. This property is very useful
when using the same advertisement in several periodicals. An advertising agency can
deliver the same films to the printers whatever the printing method.

144
 Mechanical printing

Cells transferring ink

(depth 10-100 um)

Copper cylinder with

chromium treatment

Cylinder etching Cylinder engraving

. Page data
Page positive from from
+ scanner database

Exposure on pigment paper

. Relief in pigment paper . Cylinder is engraved by means

Etching of diamond heads
solution
vY
Pigment-
paper

Copper cylinder

. Diffusion controlled etching . Chromium plating of cylinder

through pigment paper of cylinder

. Chromium plating of cylinder

Figure 16. Gravure cylinder making by etching (left) and engraving (right).

Etching by acid belongs to the class of masking methods. Pigment paper made
light sensitive by gelatin controls the etching. The paper undergoes exposure through a
positive, continuous tone page film and a special gravure screen. After exposure, wash-
ing the paper with water provides the relief. The paper relief wrapped around a copper
cylinder controls diffusion of the acid and the cell depth.
In an etched gravure cylinder, the area of each cell is equal, but the depth varies
with the density level in the page film as Fig. 16 shows. In engraved cylinders, the cell
area and the cell depth vary. Due to the different cell structures, etched and engraved

145
CHAPTER 5

cylinders give different printing results. If the technology in a gravure plant changes from
etched to engraved cylinders, all the reproduction parameters require readjustment.
Besides engraving and etching, other less common methods of gravure cylinder
making are possible. An autotypic gravure cylinder is coated with a radiation sensitive
layer like radiation sensitive offset plates. Exposure occurs through page films. Etching
occurs similarly to etching of multi metal offset plates. The difference from etched cylin-
ders is that the pictures in autotypic cylinders are normal halftones, and text is not
screened. Autotypic cylinders find frequent use in gravure for packaging applications.
Laser engraving of gravure cylinders on copper has been unsuccessful. The rea-
son is roughening of the surface due to melt metal leaks. Etching or engraving a cylin-
der to maximum depth, filling the cells with plastic, and partially burning the plastic with
laser exposure has proven practical. Some systems of this type are currently in use, and
a renewed interest is under way.
Direct cell engraving by an electon beam is possible. During engraving, the cylin-
der must be under vacuum. This makes the equipment extremely expensive.

3.4 Flexographic plates

The versatility of the flexographic method is reflected in the multitude of methods
applicable to flexographic plate making. The plates may be made from page films or
directly from computers. In the former case, the methods are photomechanical, and in
the latter they are mechanical.
Photomechanical plates have a metal or a polymer base. The light sensitive layer
contains a photopolymerizable material. The steps of plate making are the same as
those in lithographic, photomechanical negative plates — exposure, development, and
removal of the unexposed layer. An additional step may include nonselective exposure
to harden the plate profile.
Flexographic materials suitable for engraving include polymers and rubber that
can undergo engraving by mechanical heads or by laser ablation.

4 Printing nip phenomena

The phenomena in the printing nips of presses and in drying determine “physical print
formation” in imaging technology. Optical phenomena occur when prints are visualized
by light for viewing. In imaging technology, this is “optical print formation.” Perceptual
image formation is the ultimate step. From an imaging viewpoint the result comes from
cascading the influences of these steps. This section discusses the physical print
formation in nips as it relates to the principles of printing methods. Optical print
formation is the topic of Chapter 7. The subject here has two components:
- ink transfer to the plate

- ink transfer to the paper.

The phenomena in transfer to the plate and transfer to the paper depend very
much on the printing method.

146
 Mechanical printing

4.1 Ink transfer to the plate

In gravure, ink filling the cylinder cells before transfer to the paper uses flow mechanics.
Ink flows as the cylinder revolves in an ink pan. An alternate procedure uses an
additional cylinder to lift the ink from the pan. The ink transfers from the additional
cylinder to the gravure cylinder in the nip between these two. For flow into the cells, the
ink must have low viscosity. The speed of filling or the degree of filling is approximately
proportional to the inverse value of viscosity times printing speed:

Degree of
filling ~ (viscosity x speed)| (1)

This is because the resistance of ink to flow is proportional to the product of vis-
cosity [Pas] and speed [ms] divided by film thickness [m]. Ink also lifts on the nonprint-
ing areas. A doctor blade wipes it to leave only a thin layer of solvent. This layer acts as
a lubricant between the metal doctor blade and the cylinder.
In letterpress and flexo, ink transfer to the plate occurs using the height differences
between the printing and nonprinting areas and on splitting of the ink between the plate
and an ink feed roller. The principle of transfer is the “contact and split” principle.
In offset lithography, ink transfer to the plate uses surface energetic phenomena.
Surfaces have free energy since molecular forces do not balance at interfaces as
Fig. 17 illustrates. The composition of surfaces may also differ from that of the bulk.
In offset, ink transfers
to the printing area of the Z = Force unbalance at interface
printing plate by the principle i
of contact and split. This \V Interface
requires that the adhesion
force between the ink and yA
the plate (A, print
Vrinz) |S E orce balance inside
ee a material
greater than the opposing < z
force, i.e., the ink cohesion
(C;,,,)- Ink should not trans- V
fer to the nonprinting areas. Figure 17. illustration of the origin of surface energy.
This means that adhesion
(A inkmonprint) Should be smaller than ink cohesion. Figure 18 shows the principles of ink
and water transfer to the plate in offset printing. Dampening water applied to the non-
printing areas enhances lithographic printing areas. The presence of water lowers ink
adhesion to nonprinting areas. More precisely, the force required to split the water film in
the nonprinting areas is much less than the force required to split the ink film, so no ink
remains on the noimage area after separation of the inking rollers and the plate.

147
CHAPTER 5

Forme roller Water transfer roller

mie tieietsr eee Seay ween ink --; = = - = - - —- Water

Printing Nonprinting Printing Nonprinting

area area area area
Transfer condition: Ainkiorint > Cink Transfer condition: A water/nonprint 2 aes
printing area (print) nonprinting area

Nontransfer condition: Aininonprint < Cink Nontransfer condition: Avjsteriorin. < C water
Nonprinting area printing area

Splitting line

Figure 18. Simplified principles of ink and water transfer to the plate in offset printing.

After transfer to print-

ing areas, the ink should not Adhesion Cohesion
migrate sideways. A spread-
ing parameter, S, defines the ink/nonprint ink

tendency of ink to spread — —>

sideways from printing areas
Va
_
ix iia”;
to nonprinting areas. S' is the
difference between ink adhe- Nonprinting area Printing ink
sion to the nonprinting area
and ink cohesion as Fig. 19
S = Dink nononnt f Cink
shows. Ink must spread
against its cohesion. This is Figure 19. Definition of spreading coefficient.
why the net energy is a dif-
ference. S should be negative to prevent spontaneous spreading.
Since water®19
reduces ink adhesion to the Initial situation Result
nonprinting areas, it also

aes
decreases the spreading 1
coefficient. When the condi-
tions of no ink transfer on Pe
2
nonprinting areas and no Surface energy Surface energy
spreading of ink occur with- of system of system
out water, water loses its pri-
mary function. Water also
1GWane: Y42
has secondary functions. It
Change in surface energy = Ay
acts as acoolant anda
washing liquid for the nonim-
Figure 20. Adhesion defined as surface and interfacial energies.
age areas of the plate and

148
 Mechanical printing

blanket. In dry offset printing as offset printing without water is called, the rollers have to
be cooled to prevent harmful temperature rise.
A thermodynamic approach’? relates adhesion and cohesion as discussed
above to surface tensions or energies (y) and interfacial energies (y;>). Adhesion is the
change in surface energy of a two layer system when the layers come into contact:

Arg = (%1 + Y2) =V2 (2)

Figure 20 illustrates
the relationship. For a liquid
and a solid layer, the interfa-
cial tension, y,>, also relates
to the contact angle, o, that
forms when placing a drop of
liquid (index 1) on a solid
surface (index 2). According <<
to the classical Young-Dupre
Figure 21. A drop of liquid on a solid surface.
equation for the terms in
Gigazdle

Yo = Yin
t+¥, cosa (3)

By inserting the contact angle force balance in Eq. 3, the following results for the
adhesion between a solid and a liquid:

Ags =e + Yo-Y
+ Yo
Cosa = YC + cosa)
(4)

The highest adhesion occurs when o = 0, i.e. the liquid wets the solid. The lowest
adhesion occurs when « = 180°.
When a = 0,

Ya= 12 (5)

This is acommon rule of thumb to predict adhesion. It says that a liquid will wet a
solid only if it has lower surface energy than the solid.
Consistent with the thermodynamic approach, cohesion is the energy required to
split a material layer. This energy is equal to twice the surface energy of the material as
Fig. 22 shows.
In inks, factors other than surface energies contribute to cohesion. Splitting
involves more than division of a layer. Before ultimate splitting, flow and possibly elastic
deformation occur in an ink layer. Ink cohesion is therefore apt to be considerably higher
than predicted by surface forces alone.

149
CHAPTER 5

Using a more detailed

thermodynamic discussion, Initial situation Result
surface energies decom-
pose into components
according to the types of , ‘ ———
forces responsible for sur- ye —a one Se {
face energy. One approach
uses a polar component and Surface energy Surface energy
a nonpolar component!®. of system of system
Another uses acid-base 0 9
;interactions
parcels '’. It should be if
noted however that these Change in surface energy = C 1
thermodynamic consider- C,= 2y,
ations do not apply when
external force is applied. Figure 22. Definition of cohesion using surface energy.
Thus, the amount of ink
transferred cannot be influenced by changing the surface energetics of rollers or paper.

4.2 Ink transfer to the paper

In the printing nip, ink and paper come into contact under a force that acts perpendicular
to the roller axis. Nip pressure is the force divided by nip area [Pa] or the force divided
by press width (N/m; line pressure). The latter finds common use because nip widths
are not measured. Nip pressure is approximately 5-30 kN/m. Nip width is approximately
5-20 mm. This means that nip pressure is 0.1—-10 MPa. Nip pressure causes the
materials to deform. Roll surfaces, inks, and paper have a viscoelastic character —
deformation varies with the speed of printing. Deformation becomes less at higher
speeds. The deformations in the paper are reversible at the pressures that occur in
printing nips. Permanent deformation may result from the influence of water transfered
in printing nips.
The following expression depicts elastic deformation under pressure:

at

where ¢_ is _ relative deformation

P nip pressure [Pa]
E elastic modulus [Pa]
At the duration of the deformation [s]
n viscosity [Pas].

150
 Mechanical printing

Viscoelastic deformation results from a combination of elastic and viscous defor-

mation.
Pressure is necessary in the nip to overcome roughness of the material surfaces
by deformation and to bring the surfaces into contact. Contact is necessary for ink trans-
fer. Deformation in all the materials, ink, paper, plate, and offset blanket counter balance
the pressure. Total absolute deformation should equal the sum of roughness of the plate
and paper for contact between the two to occur:

Lei = roughness (8)

l

where I.; |S layer thickness for nip material i.

Deformation in roller and paper surfaces causes changes in volume; it is identical

with compression. Inks are noncompressive and they flow.
Printing methods differ depending on the material layers in the nip that deform.
Figure 23 illustrates the alternatives. For the sake of simplicity, the figure assumes that
only the paper surface and not the plate surface or offset blanket surface is rough. The
figure does not show the cell structure of the gravure roil. The same principles apply
even without the assumption of only one rough surface. The top of the diagram shows
the initial situation with no deformation — the profile of a rough paper and a layer of ink
on a cylinder surface. Because of the scale magnification, the surfaces look flat and not
curved. The other parts of the diagram illustrate three cases of deformation — deforma-
tion of the cylinder surface so it conforms with the paper profile, flattening of the paper
surface, and flow of ink from contact points.

Paper

Ink
Initial situation
Plate or
offset blanket
Paper

Ink Deformation in
the plate on
Plate or
offset blanket
offset blanket

Paper
TEE ae] —— |hk Deformation in
the paper
Plate or
offset blanket

a o 'i = i ——— |Pelore)

Deformation
ae lal obtained by ink
Plate or flow
offset blanket

Figure 23. Different deformation conditions in the nip.

151
CHAPTER 5

The plate is largely

responsible for deformation in Nip pressure
flexographic printing, and the
blanket cylinder is responsi-
ble for the phenomenon in
iia
offset printing. Compared
with gravure where the plate
cylinder is metallic, this
means the smoothness and 1 Contact and adherence
compressibility requirements 2 Penetration
on the paper are less impor- 3 Splitting
tant. In gravure, the neces-
Figure 24. Ink and paper phenomena and pressure distribution in a
sary deformation results from printing nip. -
the contributions of the paper
and from ink flow.
Contact between ink
and paper is a requirement
for ink transfer to occur. The
second condition of transfer
relates to adhesion with the
Splitting controlled
concepts discussed in the
previous section applying.
The surface energy of the ink Immobilization controlled
should therefore be lower
than the surface energy of the
paper. Contact and adhering Contact controlled
start increasing as the paper
X
enters the nip and encounters
increasing pressure. Figure 25. Ink transfer curve.
Nip pressure also
causes ink to penetrate into the porous structure of the paper. This occurs primarily in
the middle of the nip where nip pressure is highest as Fig. 24 shows. At the exit side of
the nip, the ink film splits.
Modelling of the nip phenomena can use an ink transfer equation’? '*. The classi-
cal Walker-Fetsko model dates from 195579.
Figure 25 shows the ink transfer curve. The model relates the mean quantity of
ink transferred to the paper in the solid area printing (v) to the mean quantity offered on
the plate (x):

y = AlDB
fa — DB) (9)

152
 Mechanical printing

Because the model concerns mean quantities, it is a macro scale model. Instead
of formulating the model in quantities per unit area of paper [x, (gm“2)], one can use ink
film thicknesses [h, (tum)].The relation is as follows:

= =
x[gm~] = p[kgdm]h[um] (10)
where p_ is the density of the ink.

Because p is commonly approximately | kgdm’?, the x value in gm? and thick-

ness in um are numerically close.
In the model, A expresses relative contact area (0-1). In the solid area printing,
ink should cover the entire paper area. A equals one. Uncovered areas occur on the
micro scale due to roughness. With increasing ink film thickness, uncovered spots on
the paper surface disappear approximately consistent with the following:

AS ea (11)
where k_ is asmoothness parameter.

In other words, by increasing the ink film thickness, the contact between the
paper and the ink is improved. The better the printing smoothness of the paper, the
smaller is the thickness of ink needed in printing.
An immobilization parameter, b, and immobilization function, B, give the penetra-
tion into the paper structure in the model:

esl ae
(12)
Penetration increases with an increase in x toward the immobilization capacity of
the paper with given printing conditions depicted by b [gm]. The capacity is an interac-
tive property of paper and ink. It also depends on printing speed and nip pressure.
Higher speeds give less time for penetration, and the immobilization capacity appears
less. Ink pigments do not move deeper into the paper structure after the nip by the
action of capillary pressure of the pores. Volatile ink components such as the solvent
can be absorbed, if evaporation does not efficiently remove them. The capillary suction
of the pores is a surface energy force. Ink binder may be imbibed with the solvents. To
prevent this the binder should be attached to the pigment particles.
The fraction, f, of the quantity of unpenetrated ink transfers to the paper by splitting.
The ratio of ink transfered and ink offered (transfer coefficient, /,,,) can also depict the
magnitude of ink transfer when plotted against x:

ink transferred to the paper

fiot = 13
quantity of ink on the plate le)

153
CHAPTER 5

Typically f,,,reaches a
maximum in the range
where immobilization con-
trols ink transfer as Fig. 26
shows. On smooth surfaces,
the transfer coefficient is >|x Increase in
near 0.5. With an increase in II porosity
roughness, the coefficient i)
decreases. Porosity of the
paper has an inverse influ-
ence. Increase in
roughness
Quantitative measure-
ments of ink transfer are not X
made during production runs
of printing. Gravimetric meth- — Figure 26. Ink transfer depicted by y/x = f;,;-
ods can be coupled with lab-
oratory scale discontinuously operating printing tests to provide quantitative
measurements of ink transfer.
In multi color printing,
inks in the 2—4 printing nips _ Desired splitting line
must adhere to the previ- EAS
ously printed layers. Consis- a Second ink
tent with the thermodynamic
approach, this means that First ink
the surface energy of the
inks should decrease in the Paper
order of printing. Because of Transfer condition:
other factors noted above, C, < C,

an experimental quantity General transfer condition

called ink tack has been Con Ce
defined. It predicts the over- ne
printability of inks. Tack is Figure 27. Overprinting.
the resistance of an ink film
to splitting. At the moment of overprinting, Fig. 27 shows that the tack level of an under
printed ink layer (C;) should be higher than that of the ink to be overprinted (C;,,) so
that splitting can occur in the latter layer.
The degree of ink transfer in overprinting compared with transferring the same ink
on unprinted paper is trapping. Trapping is 100% when no difference exists in the two
transfer cases. Trapping is zero when no ink transfers on a previously printed layer.
Measurement of trapping can be densitometrical or colorimetrical. Figure 28 shows
trapping of magenta on cyan.

154
 Mechanical printing

C 100 % C 100 % C 100 % C 100 % C 100%

MO0% M 25 % M 50 % M 75 % M 100%

Figure 28. Trapping of magenta on cyan.

Trapping is not a problem in printing where the ink layers dry between every printing
nip as in gravure. The situation is called “wet-on-dry” printing. It can be a problem in offset
where the transfer conditions are “wet-on-wet.” This is especially true when the time inter-
val between the nips is short. The time interval is influenced by press construction and
printing speed. In wet-on-wet printing, some ink transferred to the paper in previous print-
ing units may back transfer on the blanket in a subsequent printing unit as Fig. 29 shows.
Trapping problems
contribute to a reduction in First nip Second nip
the color range reproduced.
The number of colors printed
at any one spoton the paper —_— Transfer Trapping Back trap
and the trapping problems will first ink second ink first ink
be less when using under
color removal. This means
replacing some or all gray
component of color with black
as Section 4.5 in Chapter 3 Figure 29. Illustration of trapping and back trap.
discussed.
When using opaque inks, the most opaque ink is printed first because it will then
disrupt color formation least. Opinions about the positions of the least and most significant
colors of yellow and black differ significantly. The argument for good trapping favors print-
ing black first and then yellow last. The argument for sharpness favors printing black last.
Printing nips differ from many other nips because printing is a spatial process. It
reproduces details which can be as small as the resolution limit of the human eye. Most
other nips transfer spatially continuous layers or mold substrates. If the roll surfaces in
the nip region rotate at different speeds, the surfaces slide relative to one another and
distort the halftone dots. The speed conditions are “rolling.” Rolling is ideal when no
speed differences occur. Roll diameter is a parameter which can be used to adjust roll-
ing. The adjustment are made on the micro meter scale by packing. A method is spraying
a foam between printing plate or blanket and the cylinder in which these are mounted.
Other sources of distortion are interactions of the ink, dampening water, and paper.

155
CHAPTER 5

5 Drying of prints
In mechanical printing methods, ink is in fluid state in the press. After transfer to the
paper, the ink should solidify rapidly to avoid spreading sideways or becoming
excessively absorbed into the paper with consequent losses in printed quality. Drying of
ink on paper has two steps — setting and drying depicted schematically in Fig. 30.

Setting Drying
Printing
ink is converted
ink is in fluid ink is converted
into semisolid
form into solid form
form

Figure 30. Drying of ink.

An ink layer is set when it is touch-proof. Figure 31 shows that tests in which
another surface presses against a print with a controlled pressure measure this prop-
erty by considering the optical density transferred to the counter surface. An ink layer is
dry when it can withstand shearing without ink debonding from the surface. The test is a
rub-off test.

Pressure

Shear

e—— Print e— Print

Set-off test Rub-off test

Figure 31. Testing smearing by touch-proof and rub-proof procedures.

Tables 2 and 3 summarize the setting and drying mechanisms in different types
of printing.
The technologies used to dry prints are fairly mature. Requirements for higher
print durability, faster drying, and lower energy requirements call for new techniques.
Radiation curing technologies offer considerable promise. The use of UV and IR radia-
tion curing will probably spread from sheet-fed offset printing to other methods.
As already noted, drying the ink in traditional newspaper printing uses no external
energy. The inks also do not contain any internal drying mechanism. In the strict sense
of the word, the inks do not dry at all. A consequence is a considerable smearing ten-
dency. Because newspaper presses are multi web, the installation of dryers is not feasi-
ble. In newspaper printing presses equipped with only a few reel stands, IR drying to
enhance news ink setting is a theoretical possibility not presently widely used. News
inks may contain binder components that dry over long periods. The time scale is very
long compared with the useful life of a newspaper.

156
 Mechanical printing

Table 2. Setting and drying mechanisms in offset.

Printing method Setting i Drying

Cold-set offset Penetration of ink, separation of oil No separate drying mechanism
(mineral oil, vegetable oil) from the ink
Heat-set offset Separation of oil (mineral oil) from the ink | Evaporation of low molecular weight
oils
Sheet-fed offset Separation of oil (mineral oil) from the ink | Oxidation or
UV polymerization of binder
(IR polymerization of binder) |

Heat-set inks contain more volatile oils than news inks. Blowing air with a temper-
ature of 200°C-—300°C against the web in an oven situated after the printing units
removes the oil from the printed layer. Some ink setting occurs before the print reaches
the oven. The ovens resemble air foils used in paper coating technology. The ovens con-
sist of modules whose temperatures are independent of each other. Besides removing
the solvent from the ink, hot air also causes evaporation of water from the paper. This
has an adverse effect on paper properties such as smoothness and folding strength.
The paper also warms to 100°C—150°C. The proportion of energy consumed by evapo-
ration of these oils is only a fraction of the total energy applied. Hot air drying is not an
energy efficient drying process for printed paper. After exiting the dryer, the web travels
around chill rolls that cause a decrease in temperature. This “consolidates” the ink layer
and prevents smearing in the folder at the end of the press. Remoisturization of paper
may occur before the folder to restore folding strength.
Conventional sheet fed inks dry by oxidation of the binder. Oxidation is a slow
process. It can take a few hours before the sheets become rub-proof. Radiation by IR
can accelerate the process. Quick setting inks set by separation of a mineral oil compo-
nent by absorption into the paper. Radiation curing by UV and the use of UV sensitive
inks helps drying in the press. UV inks are especially useful when printing paperboard
and other surfaces that must be highly rub-proof such as book covers. Varnishes prima-
rily used in-line with sheet-fed offset printing have similar compositions to sheet-fed inks
except for the absence of pigment.
In gravure, Table 3 shows that drying of the inks by air stream occurs immediately
after each printing unit. This decreases the time and likelihood of solvent penetration
into the paper before the web enters the dryer. In gravure, setting and drying therefore
merge into one step.

17
CHAPTER 5

Table 3. Setting and drying mechanisms for different printing methods.

Printing method Setting Drying

Gravure Evaporation of solvent Evaporation of solvent
Flexo
plastic films Evaporation of solvent Evaporation of solvent (packaging flexo)
fiber based packaging |Separation of water Absorption of water
newspaper Precipitation of binder Precipitation of binder, evaporation of water
Letterpress Separation of oil, penetration
of ink Oxidation of binder
Silk screen Separation of oil, evaporation of |Oxidation of binder, evaporation of solvent (water)
solvent (water) UV polymerization of binder

In flexographic printing, the drying technology varies according to the type of pro-
duction shown in Table 3. In printing on nonabsorbent polymeric films, air drying is the
normal procedure. Printing on absorbent substrates such as corrugated board does not
necessarily require any external drying. The same is true for newspaper flexo. News-
paper flexo inks are aqueous and have a high pH to solubilize the binder. Newsprint is
mildly acidic. As the ink comes into contact with the paper, the acidic environment
causes the binder to separate from the water, i.e., precipitate and attach to the fibers.
UV curing flexo inks are also possible.

5.1 Mechanism of solvent evaporation

In dying by solvent evaporation, the factor that controls the speed of drying changes in
the course of the process. In the first step, the volatility of the solvent is the controlling
factor, and the evaporation rate is constant as the left side of Fig. 32 shows. During this
step, solvent concentration through the ink layer decreases linearly as the right side of
Fig. 32 shows. As the concentration of solvent depletes from the top of the print, the rate

Solvent

Solvent/ink

Solvent/paper
LAYER
INK

OF
RATE
EVAPORATION
IN
CONSENTRATION
SOLVENT

Figure 32. Evaporation of solvent.

158
 Mechanical printing

at which solvent diffuses to the top surface through the ink layer becomes critical.
Evaporation becomes diffusion controlled, and its rate gradually diminishes. In the third
step, retention mechanisms acting between the solvent and the paper such as capillary
suction become operative, and evaporation slows further. The rate becomes zero at a
point where the print may still contain a significant portion of the solvent.
Solvent concentration in the printed layer after drying should be as small as pos-
sible for best rub proof results and other characteristics. The paper and the ink layer
should exert no retention action on the solvent. No functional relationship occurs
between solvent concentration and rub-proofness in arbitrary sets of prints.

5.2 Mechanism of chemical drying

Drying by oxidation and polymerization occurs by radical polymerization. Section 9.6
discusses this mechanism in greater detail. The process consists of initiation,
propagation, and termination. In the initiation step, catalyst — initiator or drier — becomes
activated. In oxidation, this occurs by ambient oxygen. In radiation, polymerization
occurs primarily through UV radiation.
In the future, electron beam (EB) curing /® will become more common for special-
ized printing and varnishing applications because of durability requirements. EB has the
advantage of penetration through thick layers of ink and varnish because of linear
instead of exponential attenuation and energy of radiation.
Ink composition and paper properties influence the rate of oxidative drying. The
reaction generates water. If the paper has a high moisture content, the reaction shifts
toward the left, i.e., the initial state and drying slows. A low pH level has an unfavorable
effect because of reactions between catalysts and acidic components in the paper.

6 Printing presses
6.1 General aspects
Industrial printing by mechanical methods can have a wide range of scales. Estimates
for the scale result from calculating how many pages or square meters of paper can be
printed in unit time multiplied by the number of colors. Figure 33 shows the variables in
the calculation.

Web fed printing

Web speed Web width (m)

Scal z. (m/s) or or Aseniees Number of
reas a press speed press format webs
web
(rph) (pages/cylinder)

Sheet fed printing

Sheet size Press speed Total number
Scale = |(pages/cylinder)| * | (rph) X | of colors

Figure 33. Scale of printing.

159
CHAPTER 5

The scale of mechanical printing production covers a range of approximately

1:100. The cost of a new press installation varies in the same range. Considering this
wide range, the industrial environments using printing production also differ widely.
In high volume web fed printing, web speeds go to slightly less than 20 ms’! or
70 000 rph. Web widths determined by the press format (Section 6.3) can be over 3 min
the case of gravure. The number of colors is most often four, and the number of webs
can range from one to as many as 15. In sheet-fed printing, the speed may reach
15 000 rph. The total number of colors is 4-8, and the largest sheet sizes are typically
AO or its equivalent.
Printing presses are fine mechanical devices due to the requirements of printing
for spatial accuracy. Printing presses consist of

- printing units

- equipment for feeding the paper in and out of the press and between the
printing units

- dryers

- finishing equipment.

The printing operations also require control systems. Their cost is approximately
one-tenth of the press cost, although this number is increasing. A press typically con-
tains parts from several manufacturers each of which has their own area of expertise.
Printing units are the core of printing presses around which the rest is built. The
actual printing occurs in printing units. Sheet-fed presses can handle one stream of in-
fed and out-fed paper. Web presses can be single or multi web constructions. The webs
are typically combined after printing to give a single out-feed stream.
The following text discusses different types of printing units and the principles
related to press formats. The construction of printing units and the number of pages
printed per revolution of the printing cylinder relate to the printing method. Chapter 8
covers aspects related to the operations that the paper undergoes in the press including
finishing.

6.2 Printing units

Printing units consist of
- one or several printing cylinders including offset cylinders and impression
cylinders

- ink and water feed units

- drives for running the press.

160
 Mechanical printing

Printing units are “self-standing” entities. Structural differences arise depending

on the following factors:

- number of colors printed on the two sides of the paper

(marked m+n or m/n)

- direction of passage of the paper in the press — vertical or horizontal.

The following symbols
can illustrate the compo-
nents of the construction:
Printing units such as Po NER
in Fig. 34 where one or more —@O
colors can be printed on two ( C)
sides of the paper with the ~ a
web traveling up from a reel FO ee
stand are called stacks or Flexo Newspaper Offset newspaper,
towers in flexography. In off- “four high”
Set printing, they are two- Figure 34. Stacked printing units. © printing plate, @ offset cyl-
highs or four-highs depending _ inder, © impression cylinder.
on the number of colors.
Four-high newspaper
offset presses allow full color
printing on both sides of the “ee ( ee
paper web in one printing unit. Se Oye
The satellite printing eae Z @
units in Fig. 35 also known Flexo Offset Offset
as central impression (Cl) newspaper, newspaper,
printing units in flexography “9 unit! “10 unit™
allow four color printing on Figure 35. Satellite printing units. — dryer.
one side in one unit with the
web supported against a central
cylinder.
With some variation in the configuration of the basic satellite unit, 2 + 2 printing or
printing on two webs 1 + 1 and 1 + 1 becomes possible. Such variability as provided by
the “9-unit” and “10-unit” constructions is desirable in newspaper production because of
daily variation in the number of pages. Units with a vertical paper path are efficient for
using minimal floor space. This is particularly valuable in newspaper printing where the
presses consist of a large number of units. Figure 36 illustrates the construction of a
four web newspaper press that contains stacked and satellite units.

161
CHAPTER 5

$ | =a
eh
|
8G
bod! hod! thod
3
Figure 36. Four web newspaper press.

A characteristic of sectional printing units is horizontal passage of the web

through the printing unit. In a heat-set offset press such as Fig. 37 shows, a web travels
through four — or more if including varnishing — blanket-to-blanket units. Each prints one
color on both sides. Drying of the ink occurs after the printing units.

IR thermo-
meter

eS Lee Folder
Reel stand i sae
rene) Printing units Berea «

OQ)
// //
II

10]
e,

Figure 37. Heat-set offset press.

162
 Mechanical printing

Gravure integrates drying into the

units as Fig. 38 shows. In sectional gra-
vure presses, eight units are necessary
to produce full color printing on both
sides. A ten unit press with two reel
stands allows the following: Figure 38. Gravure printing unit.

- double-sided four-color printing on one web and black on the other web
- single-sided four-color printing on two webs with black on the other sides.
Figure 39 illustrates schematically the construction of a gravure press.
A vertical paper path is not practical in heat-set offset and gravure. In the former
case, the reason is the bulky ink feed units and the desired time interval between the
nips; it should not be too short for ink setting to occur. For gravure, the dryers occupy a
fair amount of space preventing stacked configurations.

Rae esi a es Ie) all ee El ealieal eles

ome | ee ee | ee | ee | ea |e ee ee

Figure 39. Construction of a gravure press.

Of printing methods, sheet-fed print-

ing is by far the most common today in off- ;
ox
set printing. Typically, one or two colors are (
) )

printed on one side of the paper in a single

»)
unit as Fig. 40 shows. Sheet transfer cylin-
oS
SoSS
x ) (
Sy NOY)
ders form an integrated part of the printing SS

units. Two single color units allow 2 + 0 or

1 +1 printing. The latter requires turning Figure 40. Sheet-fed offset printing units.
the paper between the units. © transfer cylinder.

6.3 Press formats

The format or size of a press determines how many pages can be printed at a time on
one side of the paper. The format is always a multiple of a folded page: 1 x 2 or 2x 1.
The first figure refers to the number of pages mounted around a plate cylinder in the
machine direction of the press. The latter number represents the number of pages
across the cylinder width — the cross direction of the press. In the volume production of
newspapers, a typical format is

AEXATEXI2 een Ko

163
CHAPTER 5

may"
ty a A
nt N|N|N/N wl | =="
} N|N/N|N a MiM SE
= O M|M =|s
> Q =
CD OF PRESS = Re
Figure 41. Typical newspaper press for- CD OF PRESS
mat (“broad”) with MD = machine direc-
tion and CD = cross direction. N is Figure 42. Narrow heat-set formats. M is magazine page.
newspaper page.

where two newspaper pages N (broadsheet or other standard size) are wrapped around
the plate cylinder and four across the width of the cylinder as Fig. 41 shows.
The press can run in a double production or straight run mode in which the same
physical pages, i.e., pages with the same page number of the product (two folded
pages) are positioned twice around the cylinder. In double production, two newspapers
result at every revolution. The number of pages with a different page number per web is

2 sides/web x (2 x 4)/2 pages/side = 8 pages/web.

In the single or collect production mode, the number of pages per web doubles
compared with the straight run because all pages have a different page number and is

2 sides/web x (2 x 4) pages/side = 16 pages/web.

At the same web speed, production speed or the number of newspaper copies
per hour for double production is twice the production speed in collect production. The
total number of pages in the product determine the number of webs — the minimum
number of printing units in use in newspaper printing any given day. The straight run
mode is preferable because it reduces the printing time by half compared with collect
production. The total number of printing units in the press may limit the use of the
straight run mode. If the page count is large, the collect production mode may have to
be used and the printing time doubles. An alternative solution is printing one or more
sections of the newspaper in a separate run and inserting into the rest of product on-line
with printing the main part.
In newspaper presses, the pages of a normal size newspaper orient vertically on
the cylinder. This is long grain, “portrait,” or standing orientation. Tabloid pages orient hori-
zontally. In heat-set offset printing and gravure printing, pages may also orient horizon-
tally. This is short grain or “landscape” orientation. In web fed printing, paper runs through
the press oriented in the same way as in the paper machine, i.e., in the machine direction.
Fibers in the paper have an orientation in the machine direction. This means the fold runs
in the machine or long grain direction when the pages orient vertically. When the pages
orient horizontally, the fold is across the main orientation of the fibers.

164
 Mechanical printing

The narrowest heat-set presses are narrow web presses. The cylinder format is
double or quadruple the size of a folded page as Fig. 42 shows:
2x(1x2) = 2x 2 (portrait)
4x(2x1) = 4x 2 (landscape).
Here the number of
pages across the width of the 4
web is two. In the former 2
case, the pages have portrait LL MIM
orientation. In the latter, they re M
have landscape orientation. O
Page size is typically A4 or S
B4. Because 2 x 2 is a practi- a
cal plate size, the 4 x 2 land- GENO NESS
scape press is suitable for Figure 43. Heat-set press formats.
double production. Other typ-
ical heat-set formats include that of Fig. 43:
4x(1x2) = 2x 4 (portrait) and
8x(2x1) = 4x 4 (landscape).
The latter finds common use in the double production mode.
Gravure presses are the widest of all including such formats as Fig. 44 shows:
24x (2x1) = 6x8 (landscape)
30 x (2x 1) = 6x 10 (landscape)
28 x (1x 2) = 4x 14 (portrait).
The widest gravure presses are more than 3 m wide. The basic formats of offset
sheet fed presses are multiples of the A4 or B4 (or other standard) size indicated in Fig. 45:
A3 = 2x A4
A2 = 2xA3
etc. so AO = 16 x A4

iy =
uw =
or
oO
LL
oO
Qa
=
6x8

CD OF PRESS

Figure 44. Gravure formats. Figure 45. Sheet size.

165
CHAPTER 5

Press formats in flexographic newspaper printing are the same as in offset news-
paper printing. In packaging flexography and gravure the formats are not restricted by
the standard publication sizes. Variable size presses are common.

7 Finishing in web fed printing

Paper webs undergo cutting, folding and binding in the machine and cross directions.
The operations can be in-line with printing using machine elements that function in the
MD-z and CD-z planes, respectively (z is the web thickness direction i.e. perpendicular
to the paper surface). One example is a slitter using a spherical blade fastened around
the circumference of a cylinder as Fig. 46 shows. Slitting occurs as the cylinder rotates
and the blade presses against the moving substrate.

Blade

Cutting line
= = +» MD

Figure 46. The principles of cutting in the MD (left) and CD (right).

Cutting in CD occurs with straight blades extending over the width of the web or
sheet. For in-line cutting to achieve the necessary frequency of one or more cuts per
revolution of the press, the blades sit on the surface of rotating cylinders. In off-line cut-
ting, the movement of the blades can be one-dimensional instead of rotational.

Stopper
CD
MD va
hm

74 Z

Figure 47. The principles of folding in the MD (left) and CD (right) directions.

The MD folding as Fig. 47 shows by feeding the web with the two edges brought
together into a nip. The cylinders of the nip are aligned in the MD-z instead of the MD-
CD alignment of press nips. Folds in the CD are obtained by feeding the edge of a sheet
in a stopper and pressing the middle of the sheet into a nip that generates the fold and
causes the sheet to travel in the CD-z plane.

166
 Mechanical printing

Slitting

Guiding
roller Turner
bars

Web drive Registering Machine

rollers direction
folding
Printing
Cutting direction
unit
folding, cutting

Figure 48. A schematic of the superstructure in a newspaper press.

Figure 48 illustrates the construction of the folder part of a press. In newspaper

printing, this is the superstructure.
In the folder, the webs are first slit and collected on top of one another before MD
folding. Figure 49 illustrates folding of three webs in one folder.

Web 1

© Web 2
O
Web 3

CD
cutting

Figure 49. MD folding of three webs. Figure 50. Cutting and folding of newspapers.

Figure 50 further clarifies the principle of cutting and folding the webs. A four
page wide web is the example. The web is first cut in two page wide webs. Cutting in the
CD gives the format of the product.

167
CHAPTER 5

In the basic configuration, the size

of a press sheet is always halved when
folded. Assuming the press size to be
m Xn, the number of times (x) folding MD folding
must occur to produce the signature size
relates to press size by
Cutting
(mxn)/2 = 1x1 (14)
++ 1 — — — CD folding
or

x = log,(mxn) (15)

If the number is an integer, folding in

the press can occur without slitting the
: : MD folding
web. By folding three times, the sheet of a (= parallel folding)
2 x 4 heat-set press converts into the sig-
nature size as Fig. 51 shows.
The press sheet of a6 x 8 gravure _ Figure 51. Cutting and folding in heat-set offset.
press cannot be folded into the signature
size without cutting the press sheet first into smaller entities often called ribbons. An AO
size sheet fed printed sheet reduces to A4 by folding four times.
By inserting, one or more preprinted sections can be combined with the product
or slipped between its pages.

168
 Mechanical printing

bReferencess a
1. Printing Fundamentals. Ed. A. Glassman. TAPP! Press, Atlanta 1985. p. 388.

2. Hinds, K.F., Offset Lithographic Technology. Goodheartwill, August 1995. p. 720.

3. Adams, J.M., Fauz, D.D., Rieber, L.J., Printing Technology, 4th ed.. DELMAR
Publishers, 1996. p. 646.
4. Tuttle, D., Flexographic Printing Presses. in Chemistry and Technology of Water
Based Inks. (Ed. P, Laden), Blackie Academic & Professional, London 1997.
pp. 43-110.

5. Siconolfi, F., Flexography, principles and practices. 4th ed. Ronkonkoma, NY 1991.
Foundation of Flexographic Technical Association. p. 549.
6. George, H.F., Oppenheimer, R.O., Electrostatically Assisted Gravure Ink Transfer.
12th IARIGAI Conference. IPC Science and Technical Press. 1974, Guildford UK.
7. Kipphan, H., Direct Imaging in Theory and Practice — Computer to Press vs.
Computer to Print. TAGA 1996 Proceedings. Technical Association of the Graphic
Arts. Rochester NY 1996. pp. 589-612.
8. McPhee, J., An Engineer's Analysis of the Lithographic Printing Process. 1. TAGA
1979 Proceedings. Technical Association of the Graphic Arts. Rochester NY. pp.
237-277.
9. Karttunen, S., Lindqvist, V., Water Flow and Surfactant Effects in Offset Litho.
Advances in Printing Science and Technology (ed. W.H. Banks). Pentech Press,
London. 15:176(1980).
10. Strém, G., Wetting Studies Related to
Offset Printing. PhD Thesis. The Royal Institute of Technology, Stockholm 1988.
11. Contact Angle, Wettability and Adhesion. (Ed. K.L. Mittal), Zeist Netherlands 1993.
pp. 3-36.
12. Micale, F.J., lwasa, S., Lavelle, J., Sunday, S., Fetsko, J.M., The Role of Wetting in
Printing. TAGA Proceedings 1989. Technical Association of the Graphic Arts
Industry, Rochester 1989. pp. 309-329.
13. Aspler, J.S., Leoutre, RP, The Transfer and Setting of Ink on Paper: a Review. Preprint
1991 Symp. on Paper Coating Fund., Montreal 1991.
14. Mangin, P, Lyne, M.B., Page D.H., DeGrace, J.H., Ink Transfer Equations —
Parameter Estimation and Interpretation. Advances in Printing Science and
Technology (ed. W.H. Banks). Pentech Press, London. 16:180(1981)..

169
CHAPTER 5

15. Walker, W.C., Fetsko, J.M., Am. Ink Maker 33(12):38,69(1955).

16. Bly, J.H., Electron Beam Processing. International Information Associates,
Tardley PA 1988. p. 167.
17. Berg, J.C., Nordic Pulp and Paper Reserach Journal 8(1):75(1993).
18. Good, R.J., Contact Angle, Wetting and Adhesion; a Critical Rewiew. In Contact
Angle, Wettabilily and Adhesion. (Ed. K.L. Mittal), Zeist, Netherlands 1993.
pp. 3-36.
19. Watkiss, PJ., A Review of Direct to Plate Systems in Lithographic Printing. in
Chemical Technology in Printing and Imaging Systems. (Ed. J.A.G. Drake). Royal
Society of Chemistry. Special Publication No. 133. 1993, Cambridge UK 1993.
pp. 50-72.

170
 Mechanical printing

171
BOR
Electronic printing

1 General! principles: ces siccsceveies Fei sek es ea ee eeen ones coe eeee ee a eeaen ae eee 173
Tete Over allicharacteniStleSe.. avcustecdbuct vss teaccatesecscucescaataamieecees aeeeteere er ete eee 173
We2e Method Categories ssc c-s 3,sca viesscatccsscsoncoeeeieca odeAnti anea aeeenees ae eee eee 145
2 Electrophotogranty vcecoccc access ca taseeseracs eteeees ieee ine Cesena seer eee 178
tts | BEPUNTGHONGS soeFesoti cotter ees cetne ood at cts ane reb ares ioe Bate es ee ene 178
2 i202 © SOMENS ANC DAD CIS: aconles tar Sreeahersvacsuecvsua ees walled gened acray chopaeneeeee ete ee 182
3 PBK =]OLSonn us ew oh ev atv daodie aa eects acta en ee a ee 184
OOP IING DIOS ereAeahh rnb una C8 es ett oils Mama teaad age ae a 184
Gee, GMS PANG DADCTSx.castes cae athens celts conus danas ade eaeaee a eee gece te ee 186
4 Thertival: methods 0. c.c0s1)/scavsenetstaeduss2 esdeus voy teeeeasa scenes eae ee ee 188
FRETEENCES teruey tesa Sustrcat sant aie aaraeae tent eacc ese See ee eee 190

172
CHAPTER 6

Electronicprintn =
The term electronic printing used here is synonymous with digital printing. The terms refer
to printing where the page data are input directly from a computer system to the printing
device. In principle, this allows every printed page to be different. The definition means
that the methods of electronic printing are nonimpact. An exception to the definition is
photocopying or copying using electronic printing methods that are not necessarily digital.
Electronic printing methods ’~> differ considerably in the principles they employ.
The first section of this chapter explains the general characteristics. Separate sections
discuss the principal methods, electrophotography, ink-jet, and thermal methods and
include topics related to paper and ink.

1 General principles
1.1 Overall characteristics
For many centuries, the only technology available for producing reading matter for an
audience of more than one hundred people was mechanical printing. Copying changed
this some decades ago, and the development of electronic printing technology further
expanded it.
In the longterm future, high end electronic and mechanical printing will approach
and even merge with each other. In the first stage that is reality today, single-use printing
plates for mechanical printing can be imaged in the press as the term computer-to-press
suggests. The vision is that in the second stage of development the plates will become
multi use and not require removal from and mounting in the press for each job. Plate mak-
ing constitutes a separate step due to the energy requirements of plate making and the
time taken to provide the press with page data. The third stage involves development of
receptors that are multi use — but require so little energy that the latent image can be
changed at printing speed provided the ripping and transmission speeds are sufficient.
Nonreceptor methods such as ink-jet are challenging nonimpact receptor methods.
Figure 1 shows that the milestones in electronic printing can be identified as co-
occurances of inventions and development on the nonimpact printing methods and
computers and computing.

Principle Electro- ; Laser Multi color

of electro- photographic Ink-jet printing laser
photography copying coding on paper printing Ink-jet
1 1 j
rH
1940 1960 1980 2000
Computer Computerized Computerized Resolution Automatic
text image independent page _
processing processing page production
description

Figure 1. History of electronic printing.

173
CHAPTER 6

The impetus to develop nonimpact computer printers came in the 1960s from the
area of high volume data processing, i.e., from systems printing environments. These
included banking and insurance. Personalization of printed products by addressing and
other coding applications followed next. The decade of the 1980s was the time when
typographic text printing came to offices. Development of resolution independent page
description, as facilitated by page description lanquages, was a key factor. It means that
a document, a printing job, may be output on any printer which supports decoding of the
given page description code. Currently, the use of computer-to-print technology as an
alternative to mechanical printing in the production of printed matter such as books, bro-
chures, magazines, and advertising in network environments is a rapidly evolving area
in black and white and color.
Electronic printing has three segments: low end, middle range, and high end.
Separation of the segments relate to factors such as functionality, quality, and speed.
Low end printing refers to personal printing in offices and homes. When a printer also
supports scanning, copying, and faxing, it is a multifunction printer (MFP). Middle range
printing uses distributed production such as found in organizations and schools and in
the future possibly also in bookshops and libraries. High end production competes with
low end mechanical printing. The speed of electronic production from low end to high
end spans approximately three orders of magnitude.
Until recently, the resolution of the printing device largely controlled the quality of
electronic printing. Resolution is the number of pixels (dots) printed per inch. Higher res-
olution provided better quality. In fact, it was not only the nonimpact methods that limited
the rise in resolution. Data transmission and data processing speed were others. Over
the last decade, the level of resolution in electronic printing has increased exponentially,
and other quality factors have emerged as relevant.
The number of gray levels per pixel influences perceived quality. A better measure
of quality is (resolution)? x logo(gray levels). From the visual viewpoint, these two proper-
ties are interchangeable to a limited degree, although a large number of gray levels can-
not compensate for resolution below approximately 300 dots per inch. This is because
individual pixels are visible and cause deterioration in perceived visibility quality.
Electronic printing today primarily uses scanning, i.e., printing one pixel at a time.
This means that the time taken to print a page relates to the number of pixels per page.
The latter is inversely proportional to the square of resolution:

pixels per page

Number of = BOSEALEG .
pixel area
7
= (resolution) X page area.

174
 Electronic printing

Besides the number of pixels per page, page speed depends on the speed of
printing a pixel:

speed per pixel

Page speed =
number ofpixels per page

=O XX
]
eo)
(resolution) Xx page area _ printing time per pixel

At a given level of technology, the time taken to print a pixel can be assumed con-
stant. This means that page speed inversely relates to (resolution)* as Fig. 2 shows.
The same relation
applies to data transfer. The
time required by data pro-
cessing may depend even
more strongly on resolution.
The speed of electronic
printing has developed expo-
Different levels of
nentially during the last ten speed per pixel
years. This development will
SPEED
PAGE
probably continue. When
replacing scanning by imag-
ing with a page wide array,
the influence of resolution
decreases because it only
acts one-dimensionally.
RESOLUTION
Speed and quality are
not the only performance Figure 2. Relationship between speed and resolution at a given level
factors of digital printing. of speed per pixel in scanning type printing.
Currently, the color capability
of electronic printing is an issue. Others include the capability of duplexing. Additional
issues arise from system variables such as how efficiently printing integrates with infor-
mation processing and finishing.

1.2 Method categories

Figure 3 shows the steps in electronic printing. Decoding of the page data for printing —
ripping — is the first step. The raster image processor (RIP) that does the ripping may be
in the printing device or separate. Ripping can also occur in the user’s computer. This is
an option in personal printing where a low printer price has paramount importance. A
disadvantage is that ripping keeps the computer processor occupied for the time it
takes. Equipping the printer with the RIP is standard practice in the low end and middle
range printing techniques.

175
CHAPTER 6

In middle range color

printing and most high end

Pan
Contents, structure and eae
printing, separate RIPs are lay-out data Publication data

common. This indicates two

facts. As printing speed
Raster image Ripping
increases, continuous run- processor
ning is necessary, and high
speed devices do not switch
| 2-D page data |
on and off instantaneously.
This means feeding the
ripped data to the printer ata Modulation and
constant rate. Ripping the Printing head deflection of writing
energy
entire printing job is neces-
sary before printing starts. !

RIP units are separate so

the printer will not be idle
!

during this time. The fact that

RIP makers are often differ- Energy transfer on
receptor
ent companies than printer
manufacturers also favors
separate RIP units.
A job ticket that spec-
ifies the number of copies,
settings of the RIP, type of Development
system
paper, and finishing comes
with the document content
data. The RIP unit produces Transfer to paper
system
data as two-dimensional bit-
map data. This modulates
the energy level input into
the printing head. The modu-
lated energy stream
deflected to generate the
second dimension repre-
Printed paper |
sents the “aerial image” in
photography.
Energy transfer to a
Figure 3. Imaging steps in electronic printing: ea steps,
receptor surface generates a
“latent image.’ The latent L__] outcome, LJ printer part, ed raw material.
image is generated at print-
ing speed. In some nonimpact methods such as ink-jet, no two-dimensional latent
image is generated. In those methods where a latent image is generated, it uses inter-
action of energy and the receptor surface. The receptor surface may also be paper.

176
 Electronic printing

The next step is image development — again consistent with photographic termi-
nology. An inherent characteristic of mechanical printing is the use of mechanical pres-
sure to transfer ink to the plate and from the plate to the paper as Chapter 5 discussed.
In electronic printing, interactive forces between the ink and the latent image influence
ink transfer to the latent image or image development more than mechanical forces. The
type of the latent image determines these. Transfer to the paper — the next step — may
use mechanical forces. It may also use the offset principle. Drying of the print or fusing it
concludes the printing cycle.
A fundamental imaging classification of nonimpact methods uses the type of
latent image. Figure 4 shows the major categories:
- electrical charge based methods

- magnetic methods

- thermal methods

- ink-jet methods.

Latent image

Electrical charge Magnetization Thermal

pattern pattern pattern

——

Electro- Electro- Electron Direct Thermal

photography graphy beam imaging |
Maonctogrenny thermography | | transfer

CS = Continuous stream
DOD = Drop on demand
ont Electrical Corona Magnetic Thermal
g current discharge field current

Writing energy

Figure 4. The major categories of non impact printing.

The electrical charge pattern is the most versatile type of latent image. The
generic name for the group of methods in this category is electrostatic methods. In elec-
trophotography, the charge pattern is generated by the interaction of light from a printing
head with electrical charges deposited on a receptor surface. In electrography, the
charges deposit directly on a receptor that may even be paper using electrode matrices

177
CHAPTER 6

as printing heads. The paper must possess dielectric properties so that the deposited
charges do not leak away before development. In electron beam imaging, corona dis-
charge heads generate charged particles. In electrostatic methods, development of the
latent image into a visible image occurs with electrostatic attraction of toner particles.
In magnetography, a magnetization pattern is produced using magnetic printing
heads like those in magnetic recording on a magnetizable surface. Toner is attracted to
the latent image through magnetic forces.
In the group of methods called thermosensitive or thermal methods, the latent
image results from variations of temperature. Such an image can result from a thermal
printing head on specially coated paper as in direct thermography or on an ink film as in
thermal transfer methods. In direct thermography, the heat energy of the latent image
triggers a chemical reaction or a physical phase change that leads to local variations in
contrast. In other words, the imaging behavior is a component of the paper. In thermal
transfer, the thermal latent image causes ink transfer from a coated film to a receiver
paper by changes in viscosity or by phase changes.
In ink-jet methods, no two-dimensional latent image forms. The two varieties of
the method are continuous stream (CS) ink-jet and drop on demand (DOD) ink-jet. In
the former, a continuous stream of drops emanates from a nozzle. Drop selection for
printing involves selectively charging the drop sand steering them electrostatistically in
much the same way as electrons are steered in cathode ray displays. In DOD ink-jet, all
drops that are generated go to the paper. A thermal signal commonly controls drop gen-
eration by thin film boiling. An alternative is piezoelectric generation.
Those nonimpact methods that use light as writing energy can function as analog
copying methods (print-to-print). All the methods can function as digital copying tech-
niques (print-to-computer-to-print).
Each nonimpact method has a characteristic functionality, quality, and speed
potential. In simple terms, the methods using receptors e.g. electrophotography and
magnetography have speed potential. The nonreceptor methods e.g. ink-jet and thermal
transfer are inherently suitable for color printing. No demarcation line exists between
receptor and nonreceptor methods when considering quality. A further functionality
aspect is integration with other printing or paper converting lines. Ink-jet printing being
noncontact is suitable for integration.

2 Electrophotography
2.1 Principles
Electrophotography® 2 or xerography is the main method of electronic printing today.
Discovered at the end of the 1930s, the patented invention covers converting a two-
dimensional electrical charge image into a visible image by toner development. The first
electrophotographic product was an analog copying machine. It is still on the market
today. Introduction of the first laser printer occurred in the middle of the 1970s. Multi
color laser printing has been available since the end of the 1980s.
Electrophotography is undoubtedly the most mature black and white nonimpact
technology. Global development activity guarantees continual improvements in color

178
 Electronic printing

technology. Photocopying was the impetus for early development. Several factors fuel
the development today. Color is their common denominator. High end color electropho-
tography challenges sheet-fed offset printing. An additional bonus is its variable data
capability. At the low end, higher complexity prevents color electrophotography from
competing with color ink-jet.
Long term concerns in development include memory photo conductors. Today,
the latent image must be formed on each revolution. Memory photo conductors such as
magneto ceramics would make that unnecessary.
Currently, commercial electrophotographic printing at the high end can print at a
speed and resolution that correspond to a data streaming rate of up to gigabit per sec-
ond range. The standard resolution in binary printing is 600 lines/in., but this is increas-
ing. Color electrophotography uses multiple levels of tone per pixel.
Figure 5 shows that electrophotography is a multi step process. The steps include

1 charging of a photo conductive receptor surface

2 exposure by page information to give an electrical charge pattern correspond-

ing to a printed page

3 conversion of the charge pattern into an optical image by development with

a toner

4 transfer of the toner to the paper

5 fusing the toner on the paper

6 erasure of the remains from the receptor surface before repeating the
process.

Mirror

Exposure 4 | Toner

Illumination
Charging] a Development

Q Original eracuie

|
Fusing
Transfer! Paper
Figure 5. The steps in electrophotography (analog electrophotography).

In analog photocopying, reflection from a page original uses mirrors to expose

the photo conductor drum. In digital electrophotography, the exposure results from mod-
ulating a light source such as a laser by digital page data as Fig. 6 shows. In digital pho-
tocopying, the digital page data results by scanning from a page original and analog to
digital (A/D) conversion.

79
CHAPTER 6

Photo conductor drums are

Light semiconductor materials
that are electrically resistive
Ee NGduiationand in the dark — capable of
Digital data °

scanning maintaining a charge — but

conductive when exposed
with light’”. Charging of the
Exposure | + E Toner
+
photo conductor in the first
step of electrophotography
+
t+
primarily uses corona. The
principle involves ionization
of air between a corona
Paper roller attached to a high
Figure 6. Digital electrophotography. potential source and a
dielectric surface. In electro-
photography, the surface is a photo conductor. lonized air conducts electricity that leads
to migration of charged particles such as electrons and ions from the corona to the
photo conductor.
The exposure of the photo conductor in digital electrophotography most com-
monly uses scanning with a laser light source. This explains why electrophotographic
printing has the common term of laser printing. Simultaneous exposure of the entire
width of the page is possible by using parallel modulation of a light source built as an
array such as a light emitting diode (LED). The exposure causes neutralization of elec-
trical charges in the nonimage area. Different levels of density per pixel result by modu-
lation of the laser power. Eight bits per pixel modulation is common in color
electrophotography.
The electrical field generated by the latent image charge pattern controls transfer
of the toner to the photo conductor’. Because the toner has an electrical charge, the
field exerts a force in the toner particles amounting to

force = electrical field x toner charge (3)

This force pulls the toner to the photo conductor. The type of development deter-
mines the opposing forces and the force balance that influences the amount of transfer.
Transfer of the toner from the photo conductor to the paper’?-’5 in the next step
results by

- Charged roller opposite in sign to the toner

- corona charge on the back side of the paper opposite to the toner

- mechanical transfer by use of an offset cylinder or belt.

180
 Electronic printing

Sear P

Age
ih= eeeoe
vr]
:
'
PC PC

Corona Nip and Offset cylinder

electrical field
P = paper
T = toner
PC = photo conductor

Figure 7. Transfer of toner from the photo conductor to the paper.

Transfer by generating an electrical field using a corona electrode positioned at

the back of the paper as Fig. 7 shows is the classical technology. New types of toners
have also given impetus to application of other principles. These include using a nip or
an additional transfer medium such as a drum or a belt. Their operation resembles the
principles used to transfer printing ink to paper in mechanical printing.
Toners usually consist of fine solid particles. They are powders. To improve adhe-
sion of the particles to the paper, fusing which is analogous to drying, follows the trans-
fer step ’®,Fusing occurs by heat and pressure called hot nip fusing or by radiation or by
pressure alone. The last named is called cold nip fusing. Heating causes binder materi-
als in the toner to soften and melt with subsequent spreading on and into the paper.
Pressure also causes some temperature rise in the toner with resulting softening and
toner penetration into the paper. Could fusing has limited application at the low end of
the speed and quality scale.
Figure 8 shows the implementation of color printing by electrophotography in var-
ious ways. The technique depends foremost on the desired printing speed. At low
speeds, the page data representing the four colors can be imaged in succession on a
drum. The time to print a copy is the sum of the time to print each of the four colors. The
toners may accumulate on the photo conductor, transfer drum, or belt and transfer
simultaneously to paper. Alternately, one color at a time can develop on and transfer
from the photo conductor. The transfer belt then must contact the photo conductor four
times.
If the printer has a separate unit for each color as in mechanical printing, imaging
the four colors occurs simultaneously. The time taken to print successive copies equals
the time taken to print one color. Such a “tandem” construction is more expensive than
the single drum construction and therefore primarily suitable for middle range and high
speed printers.

181
CHAPTER 6

yN
Accumulation on paper
< 2

Accumulation

e eal
—- ri fel llr
= Accumulation (PC)
= on photoconductor ae

T| See PC = Photo conductor

4;
ey+
im
| | |

1 10 100
SPEED, page/min
Figure 8. Four color electrophotographic printing.

2.2 Toners and papers

As noted above, electrophotographic toners are commonly powdered materials. The
practical size is typically 5-10 um. A narrow distribution has been found to be
advantageous. With an increase in the nominal resolution, particle sizes have
decreased. Approximately, the particle size is a fifth of pixel size. At 600 dpi, the particle
size is thus about 8 um. A printed pixel contains several particles.
Liquid toners are also possible. They have application when the resolution
requirement is high because toner particles dispersed in a carrier Such as a kerosene
like solvent can be small — less than 1 um. After transfer to the paper, the carrier phase
becomes superfluous. When using an offset cylinder, the liquid can evaporate from the
cylinder before transfer to paper.
Toner particles consist primarily of pigment to impart color, thermoplastic resins
to bind the pigment to the paper, and charging additives to enhance the charging capac-
ity. In percentage, the resin content is by far the largest often more than 90.
Ink flow and ink splitting on roller surfaces implement the transportation of inks in
mechanical printing. Powdered toners require different techniques. They transfer by
adhesive, electrical, and magnetic forces. Another factor in the use of toners is how to
charge them electrically. Considering these two issues, two major solid toner technolo-
gies are possible /2:
- mono or single component toners

- two or dual component toners.

182
 Electronic printing

Mono component toners may contain magnetic materials to allow movement of

the particles from a supply to the development region on the photo conductor by mag-
netic forces. If the particles are not magnetic, rollers transport them. Electrical charging
provided by the transfer system controls the transportation.
Two component toners contain toner particles and carrier beads that are approxi-
mately an order of magnitude larger than the particles. Carrier beads are metal particles
coated with a thin polymeric layer. Charging of the toner particles occurs in contact with
the carrier beads. The mechanism is triboelectrification. Transport to the photo conduc-
tor is electrostatically controlled.
The most important properties that characterize toners include
- particle size

- charge to mass ratio

- surface chemical properties

- viscoelastic properties.

Small particles give good sharpness, resolution and even solid areas but are sen-
sitive to the conditions of development. Process stability may suffer. The charge to mass
ratio must be within a tolerance range for the process to function without particle depo-
sition on the background area and their visibility as optical noise. Scatter of toner parti-
cles near printed edges is a characteristic feature electrophotographic prints as Fig. 9
shows’”. It is related to both particle size and charge to mass ratio.
Fusing behavior of toner particles
determines the mechanical properties
and surface texture of the toner in print.
The behavior is controlled by surface
chemical and viscoelastic properties. The
fusing conditions should cause the binder
to soften and adhere to the printing sub-
strate. Melting is necessary to achieve Figure 9. Electrophotographically printed character.
print gloss.
Electrophotography requires specialized properties from the paper. Because of
the use of electricity in the process, the runnability of a paper is critical. The electrical
conductivity of the paper should be sufficient so that excessive accumulation of electric-
ity on the paper does not occur despite charge removal conductors in printing devices.
Accumulation of charges causes sheets to adhere to one another and jamming. Image
quality requires electrical resistivity in the paper, because the paper must hold the elec-
trical field during transfer of the toner. These two factors determine a tolerance range for
paper resistivity. All toner should transfer to the paper from the photo conductor to pre-
vent accumulation after the erasure step. Accumulation will cause a ghost image on the
next print. With a rise in moisture content, resistivity drops exponentially.
In solid areas, the amount of the toner on paper is 5-10 gm". If the paper is too
conductive, the toner spreads sideways causing losses in detail. The moisture content

183
CHAPTER 6

of paper is critical as is friction. The significance of moisture content arises from its influ-
ence on paper resistivity. Friction must be within a specific range since the transporta-
tion occurs commonly by friction.
The thermal conditions in fusing determine the requirements on the dimensional
stability of the paper and the strength of the surface. The higher the temperatures, the
more stringent are the requirements. Thermal properties also have a role. A lack of
dimensional stability can cause curl, cockle, and image depletion in duplex printing.
Softening of the paper surface in fusing may cause adhesion on the counter roll loosen-
ing particles from the surface and accumulation on the rollers of the printer.
Good toner adhesion requires surface chemical compatibility of the paper and the
toner in the surface chemical sense in Chapter 5. With an increase in the resolution of
printing and the related decrease in toner size, the smoothness requirements’? on
the paper become increasingly important. With better resolution, other aspects of qual-
ity requirements also increase. These include requirements for small scale evenness of
structural, electrical, and thermal properties of the paper.

3 Ink-jet
3.1 Principles
In ink-jet printing?’7, a drop is the basic imaging element. Figure 10 shows a
classification of ink-jet methods. The principles of generation of ink drops from a stream
of liquid emanating from a nozzle have been available since the late 1800s. These
principles form the basis of continuous stream ink-jet printing available as a commercial
method since the 1960s. Key issues in the development of continuous stream ink-jet
concerned the mechanisms applied to select drops from a liquid stream with control by

Ink-jet

v v
Continuous jet Drop on demand jet
(CS) (DOD)

: Vv v
Piezoelectric jetting, Piezo- henna
selective charging, and electric ‘ettin
deflection jetting J g

Pee
|| Liquid |
!
Liquid
ink ink

Figure 10. Classification of ink-jet methods.

184
 Electronic printing

the page data. In discontinuous, i.e., drop on demand ink-jet, all generated drops go to
the paper. The history of DOD ink-jet covers a period of about 30 years and is much
shorter than that of CS ink-jet.
The areas of application of continuous ink-jet include high speed printing such as
addressing, personalization, and coding and high resolution color printing such as
proofing. Publication printing in full color is much awaited. DOD ink-jet finds primarily
use in office, home, and wide format printing. Color pictures have an increasingly impor-
tant role in home and wide format printing. Digital photography is an important driving
force behind color printing in homes. Wide format printing of banners, signs, and display
advertising benefit from the one-off printing capability of ink-jet.
Due to the light weight and compact construction of ink-jet printing heads, the
method has potential for integration in-line with other printing methods or processes.
Expectations are that ink-jet will develop into a versatile color printing method despite
the considerable technical obstacles. The time required for development will probably
be quite long. A factor behind this observation is the very specialized and highly multi
technical nature of ink-jet know-how.
According to a fundamental thermodynamic principle, a liquid stream as gener-
ated in CS ink-jet breaks into drops“** if the external area of the stream is greater than
the external area of the drops. This is because droplet formation reduces the total sur-
face energy of the system. Nozzle size, pressure in the nozzle, and perturbation in the
nozzle influence drop formation and drop size. Figure 11 illustrates the principle. A
piezoelectric crystal in the nozzle chamber generates perturbation. “Printing” drops are
selected from the stream of drops by charging with a signal controlled by the data-
stream and electrostatic deflecting. Implementation may also be the reverse, i.e., the
nonprinting drops are selected by charging and deflection.

Modulation by
data stream

Deflection
Radio frequency Charging of of ink Direction of travel
excitation ink drops drops =of the pap
paper

oft Printed line (the device

i 5 O & is capable of printing
OPOEGeOL OBO S a line of pixels)

Collection of
nonprinting
drops
Figure 11. Basic principle of continuous stream ink-jet.

185
CHAPTER 6

Continuous stream ink-jet can reach a speed of approximately one million drops
per second. With a nozzle size approaching 10 ym in diameter, the speed of the drops
can be approximately 50 ms~'. Drop size is approximately 10 piko liters. Gray levels are
generated by halftoning or by firing 0-n drops per pixel where n may be 2°. This results
inn +1 possible gray levels. Halftoning increases the number of tone levels.
The most common type of drop on demand ink-jet today is the thermal ink jet?° or
bubble ink-jet that Fig. 12 shows.
In this process, a ther-
mal heating element con- A Drop
nected to the digital data A
heats the ink periodically. Hey EN
Near the heating element, Ink
ink vaporizes instanta- Bubble ' é
neously to generate a bub- XS 4f nese layers
ble. A given mass of liquid /STL ea Res rea
esistor
has a bigger volume in the ~~ Thermal resistor
vaporized state than as a -—— Background
fluid. The increase in volume
in the ink nozzle causes Figure 12. The principle of DOD thermal ink-jet showing a highly
some ink to squeeze from magnified view of the nozzle.
the nozzle and form a drop
that travels and deposits on the paper. The thermal element cools before another cycle
begins and another pixel prints.
Variation of the heating period can be used to modulate drop size to a degree.
One drop represents one pixel. This means that by modulating the heating period a few
gray levels can be generated without halftoning. The drop generation speed is ten thou-
sand drops per second approximately two decades below that in continuous ink-jet
printing. Resolution reached in DOD can exceed CS by a factor of five. The largest
nozzle matrices contain about a thousand nozzles so that the speed per printing head
reaches ten million drops per second.
A commercially viable alternative to thermal drop generation is pressure genera-
tion by pietzoelectric crystal. In piezoelectric ink-jet, a crystal mounted behind the noz-
zle expands and shrinks at the rate of an incoming acoustic signal. Page data
modulates the signal. Piezoelectric ink-jet can use fluid and phase change inks. The lat-
ter are solid at room temperature and fluid at the temperature of jetting.

3.2 Inks and papers

In ink-jet development, inks and papers have played a key role. One reason is the need
for the very small nozzles through which the ink emerges required to reach high
resolutions. The inks should not clog the nozzles when in a stand-by state or form a
crust. In thermal DOD ink-jet, rapid thermal gradients in the heating period impose high
demands on the thermal stability of the inks. Cogation is lack of stability.
A characteristic of ink-jet inks compared with inks in other printing methods is that
they are very fluid. This is especially true for the continuous stream ink-jet method

186
 Electronic printing

where rapid drop formation requires viscosity near 1 mPas. In thermal jetting, viscosity
is commonly less than 5 mPas, 10 mPas is the upper limit. Table 1 gives a summary of
the main components of ink-jet inks

Table 1. Composition of ink-jet inks.

Component Purpose
Dye or pigment Coloring material
Binder Binding the pigment on the paper (not always used)
Liquid phase Carrier phase for the dye or pigment and binder
Continuous jet: water+ alcohol DOD jet: water | Allows the desired low viscosity of the ink and prevents ink
+ glycols from drying in the nozzle
Charge generation additives (continuous jet) | Enhances the electrical charging capability of the ink
Preservatives Prevents bacterial decomposition

The use of pigments instead of soluble dyes as the source of color in ink-jet inks
is increasing due to their lightfastness. Low levels of viscosity and freedom from nozzle
clogging can, however, be more easily reached with soluble dyes. Dye concentration is
considerably below 10%. Ecological aspects, low viscosity and desired drying proper-
ties are the main criteria in selecting solvents for the inks. Water is preferred especially
in home and office enviroments. In CS ink-jet, rapid charging of the drops requires a
high electrical conductivity. This is the reason for the charge generation additives.

Figure 13. Ink-jet printed dots. Top left: uncoated paper, top right: coated paper, bottom: white opaque
papers.

187
CHAPTER 6

Because of the fluidity of the inks and the high velocity of impact on the paper,
ink-jet papers ’® 19 require characteristics that are matched with the inks and the drop
volumes. This is especially true at the higher end of the quality level. Ink-jet papers must
be smooth so that drops impacting on the paper spread evenly. Figure 13 shows some
examples.
Sufficient and even porosity composed of small pores is necessary to absorb the
solvent quickly and to counteract the spreading tendency. The amount may be 102 gm”.
Dyes are trapped close to the surface of the paper. This can be aided by cationic dye fix-
atives in the paper which capture the anionic dyes and allow the colorless vehicle to be
absorbed into the paper or coating. The use of aqueous inks places demands on the
dimensional stability of ink-jet papers. Cockling and curling tendencies are critical fac-
tors.
The impact of drops on paper causes enlargement typically by a factor of two.
This and limits to decrease drop size constrain efforts to raise resolution although the
commercially implemented levels keep going up and have exceeded 1000 dpi. Printed
pixels are most visible in light tones where they are far apart. Using low concentration
inks alongside full concentration inks can reduce the visibility of resolution in light tones.
This practice is common in ink-jet printers for photo printing. The diluted inks include
cyan and magenta.
Another modification is the use of more than three colored inks that helps reduce
the solvent that transfers to the paper in multi color printing. This means supplementing
the three process colors cyan, magenta, and yellow with additional color inks such as
orange and green. They also bring the advantage of a wider color gamut.

4 Thermal methods

Imaging by thermal methods*° consists of two main categories. When color and
contrast are obtained by heat induced chemical or physical reaction, it is direct thermal
imaging. With heat induced transfer to paper of ink coated on a film, it is thermal transfer
imaging.
Thermal methods are small scale methods — typically less than 10 pages/min.
New applications that would require higher speeds are not emerging. Direct thermal
imaging has its main application in telefax and label coding. The former is a diminishing
area due to “plain paper’ alternatives such as ink-jet and electrophotography. Thermal
transfer methods are capable of high quality color products required in proofing and
photorealistic printing.
Commercial thermal transfer methods exist in the following forms:
- thermal ink (wax) transfer
- thermal dye transfer or dye diffusion thermal transfer (D2T2) or dye sublima-
tion.

188
 Electronic printing

Both forms involve the use of a ribbon on which a continuous film of ink has been
coated. The ink transfers to the paper by heat from a digitally driven printing head. Fig-
ure 14 illustrates the principle. Color results from using ribbons consisting of page wide
segments of the different colors.

*=-— -aper
| Cyan Magenta Yellow _*——— Ribbon

Thermal head

Digital data

Figure 14. Principle of color thermal transfer (not to scale).

Thermal ink transfer is a binary method. It generates two levels per pixel (1 bit:
“black or any other color’ or “white”). Gray levels result from halftoning. In thermal ink
transfer, heat causes release of the ink layer from a ribbon. Release within a narrow
temperature range can occur in two ways. Wax type compounds may be components of
the ink or applied as a separate coating between the ink layer and the ribbon.
In thermal dye transfer, the amount of ink transfer can be controlled by variation
of the heating period of the head. This allows generation of several gray levels.
Thermal transfer methods require very smooth papers compared with printing
papers to achieve contact between the ink coated ribbon and the paper under the lower
pressure used. The pressure is of the order of kilopascals instead of megapascals as in
mechanical printing. A special coating enhances migration of ink into the paper. Papers
made from synthetic fiber or synthetic film are common. Synthetic films may be lami-
nated to papers from wood.

189
 CHAPTER 6

MReferencess aa
1. Anon., Handbook of Imaging Materials. (ed. A.S. Diamond). Marcel Dekker, New
York 1991.
2. Anon., Output Hardcopy Devices (ed. R.C. Durbeck and S. Sherr). Academic Press,
New York 1988.
3. Jonson, J.L., Principles on Non Impact Printing. Platino Press, Irvine Ca 1986.
p. 345.
4. Levy, A.V., Biscos, G., Non-Impact Electronic Printing. InterQuest, Charlottesville Va
1993. p. 314.
5. Oittinen, P, Saarelma, H., Elektroninen painaminen (Electronic Printing; In Finnish).
Otatieto 882, Otaniemi 1992. p. 220.
6. Schaffert R.M., Electrophotography. Halstead Press, NY 1975.
7. Scharfe, M., Electropotography, Principles and Optimization. Research Studies
Press. Letchworth, England 1984.
8. Williams, E.M., The Physics and Technology of Xerographic Processes. John Wiley,
New York 1984. p. 288.

9. Pai, D.M., Springett, B.E., Reviews of Modern Physics 65(1):163(1993).

10. Schmidlin, FW., IEEE Trans. Electron Devices. ED-19(4):448(1972).
11. Law, K-Y., Organic Photoconductive Materials: Recent Trends and Developments.
Chem. Rev. 93:449(1993).
12. Schein, L.B., Electrophotography and Development Physics. Springer Verlag, Series
in Electrophysics vol 14, Berlin 1988. 2. korjattu painos 1992.
13. Hosaka, Y., Xerographic Toner Transfer Analysis. Japan Hardcopy ’88. pp. 123-126.
14. Schleusener M., Modelling of Electrostatic Toner Image Transfer. Proceedings of the
IS&T Seventh International Congress on Advances in Non-lmpact Printing
Technologies. Vol. 1(1991), pp. 167-175.
15. Yang, C.C., Hartmann, G.C., IEEE Tans. Elec. Dec. ED 23(3):308(1976).
16. Lee, L-H., Thermal Fixing of Electrophotographic Images. Adhesion Science and
Technology (ed. L-H. Lee), Plenum Press 1075, pp. 831-852.
17. Scharfe, M.E., Schmidlin, FW., Charged Pigment Xerogaphy. in Advances in
Electronics and Electron Physics, 38(31):31(1978).

190
 Electronic printing

18. Oliver, J.F., Wetting and Penetration of Paper Surfaces. Colloids and Surfaces in
Reprographic Technology (ed. M. Hair, M.D. Croucher), ACS Symposium Series
200, 1982, pp. 436-453.
19. Oliver, J.F., J. Imaging Tech. 14(5):144(1988).
ZO, Kulmala, A., Kopio- ja laserpaperin painettavuus (Printability of copy and laser
printing paper). Licentiate of Technology Thesis. Helsinki University of Technology,
Laboratory of Paper Technology, Otaniemi 1996. 83 p + Appendices.)
a ~—. Heinzl, J., Hertz, C.H., Advances in Electronics and Electron Physice. Academis
Press 65:91(1985).
22. Loye, W, J., Taub, H.H., Ink Jet Printing. Output Hardcopy Devices (ed. R.C. Durbeck
and S. Shrerr). Academic Press, Boston 1988. pp. 311-370.
PaE Rayleigh, J.W.S., Proc. London Math. Soc. 10:4(1978).
24, Weber, C., Zeitschrift fur anfewandte Mathematik und Mechanik 11(2):136(1931).
Bey Lee, F.C., Overview of Thermal Ink Jet Technology. SPIE Hard Copy Output
1079:342(1989).
26. Hann, R.A., Thermal Dye Diffusion Printing — How Does it Work. Proceedings of the
8th International Congress on Non-Ilmpact Printing Technologies. IS&T,
Williamsburg Va 1992, pp. 361-363.

191
BOAR ki
Optical imaging in printing

1 Overview of optical’phenomena 22.32 2.20.i0.ccacs onasesescsas cscs oeacseenseseneeerssseeenete 193

2 SUraGe TELEGTION os -cc ccc ccr ces cen ceseee coe ee erates eee eae eee eee 194
3 infernal reflection seca. sheesh aca eevee eaters ce ete eae 196
Sil eh SOU GIDITES Bee, chor. couse eter ct ot dans OO eS ey ee eee ee oe ee 196
2, AAR UO EOFS eek occ sceteeek vets acs cue tosace toel tees eae ateeret etter see eae ease nan ne ee 199
4 Colorimetrig.Golots cc: ttre cckethewancc ect eadendans Soeeesaae te eo eeaaeee ee 201
4) Optical behavior of the substrate ............... ccc cc cccceceeseeceeseeeeeeeeenaeeners 203
6 PRIN NOISE Weeccc ts sane asnetay ovat aan necnaascnace teers cateuecue ree ceed ei nea seen eee ee 207
FROTGTOIICESS, czalares cro accsnieet aon eo eer aR ey est SC Sa on ee 210

192
CHAPTER 7

Optical imaging nh
printing
Imaging in printing from a physical viewpoint was an earlier topic. This chapter explains
the basic principles of optical image formation — interactions of light and print. Chapter 3
introduced some concepts applied here to printing and prints. Its Section 2 discussed
the human visual system which also has relevance to the topic of this chapter.

1 Overview of optical phenomena

Interaction of light with a physical print generates an optical image.The term light means
the power of radiation or intensity, and intensity, times its duration or energy. Optical
print formation is the step where light interacts with a physical print. A print or any other
hard copy such as a photograph does not exist as an image without this interaction. A
physical image is characterized by the distribution of some light absorbing material such
as silver grains, toner particles, or printing ink and its components in the x,y-plane and
the z-direction of a matrix structure such as paper. In printing, the interactions of the ink
or toner and the substrate cause the coloring material to be spread and penetrated
according to statistical mechanisms. The variation resulting from the randomness
covers several orders of magnitude on the spatial frequency scale. The characteristic
scales of the imaging elements such as pigment particles determine the smallest scale
of variation. The largest scale of variation typically relates to functioning of the printing
process. This chapter first examines the optical phenomena on a macro scale. Sections
2-5 cover this. Section 6 examines random variations and the micro scale.
Light and image interactions include several phenomena’. Figure 1 shows these
using a print consisting of an overprint of two ink layers as an example.

1 Incident light beam

2 Surface reflection distribution

3 Llight travelling in the ink layer, absorption

4 Scattering of light in the ink layer

Ink layer es di caused by ink opacity
5 Scattering of light in the paper causes
Ink layer : optical spreading

Paper 6 6 Reflection of light from the paper

. 7 Internal surface reflection

Figure 1. The phenomena of light interaction with a print.

193
CHAPTER 7

When light impinges on a physical image, surface reflection called the first sur-
face reflection occurs from the interface between the image and air. Surface reflection
from prints of any color has approximately the same spectral composition as the incom-
ing light. In other words, it is white. On metallic surfaces, the spectral composition of sur-
face reflection differs from the incoming light. This is due to higher refractive indices of
metallic surfaces than of prints. The proportion of light that is not surface reflected is
refracted into the color layers.
In the color layer, light absorption occurs. Light absorption in black inks is spec-
trally nonselective, and in colored inks it is spectrally selective as discussed previously.
Absorbed light energy converts to heat. In unfavorable conditions, the absorbed energy
may cause a chemical reaction leading to color changes and fading.
A printed ink layer should not scatter light. Scattering causes light to change its
direction of travel without any influence on light intensity. In a multi layer print, light trav-
els from an overprinted ink layer to the underprinted ink layers and the paper without a
significant surface reflection at the interfaces. This is because the refractive indices do
not differ much.
In the paper substrate, light scattering occurs. Reflectivity of paper uses succes-
sive scattering events. More pronounced scattering results in a smaller depth to which
light penetrates before backscattering. Light scattering causes light diffusion. It loses its
directionality and polarization becomes randomized*. Scattering also causes light to
travel sideways in the paper. This means that even when illuminating a paper surface
with an infinitely narrow beam of light, the resulting image is an enlarged spot. A conse-
quence is that infinitely sharp printed dots or lines do not look infinitely sharp. This phe-
nomenon is optical spreading?®. In optical spreading, the density profiles of dots
therefore flatten, and the effective halftone dot areas increase. Because light scattering
depends on wavelength, the influences differ in various spectral ranges®. Section 5 in
Chapter 3 briefly discussed modelling of dot spreading by a point spread function in the
context of color calibration. Section 2 of this chapter discusses optical spreading further.
Light absorption also occurs to some degree in the paper. Light scattering and
light absorption determine paper opacity. The spectral selectivity determines the paper
color. The combined result of scattering and absorption is that a given proportion of light
reflects from the paper and penetrates back into the ink layers.
When light reaches the interface between the uppermost ink layer and air from
within the print, some undergoes surface reflection back into the ink layer. This is inter-
nal surface reflection. Steps 3-6 of Fig. 1 then repeat. The intensity of light that passes
many times through the ink layers decreases rapidly and its signicance is small.
In summary, light reflection from prints consists of reflection coming from the sur-
face and reflection coming from within the sample — internal reflection.

2 Surface reflection
The surface reflection behavior of any surface ''*:7-9 has two factors:
- the relative proportion of surface reflection of the incoming light intensity
- the angular distribution of surface reflection in the half space above the print.

194
 Optical imaging in printing

The relative propor-

tion of light that is surface
reflected depends on the
angle of the incoming light
relative to the surface nor-
mal. Figure 2 shows the
refractive index of the sur-
face. The curve in Fig. 2
depicts reflection of ran-
domly polarized light’. The -:

reflection coefficients vary

with polarization. Polariza-
COEFFICIENT
REFLECTION
tion is preserved in surface ae
reflection from papers and 0 45 75 90
Surface normal
prints“. With an increase in
the angle of incidence, the ANGLE OF INCIDENCE,°

proportion of surface Figure 2. Relative surface reflection as a function of the angle of

reflected light increases. At _ incident light. Refractive index 1.5.
perpendicular incidence
(zero angle of incidence), the fraction of reflection on paper and print type surfaces is
approximately 4% (reflection coefficient 0.04). It is approximately 26% at 75° incidence.
When a surface is illuminated in its plane at 90°, all light is surface reflected.
Refractive index is a fundamental optical material parameter. With increase in the
refractive index, the magnitude of surface reflected light increases at any given angle of
incidence. In the class of paper making materials, titanium dioxide has a higher refrac-
tive index at 2.7 than the other raw materials. Most fall into the narrow range of 1.5-1.6°.
The refractive indices of inks of different color and paper are approximately the same
magnitude. This means there is no optical interface between them, and interfacial sur-
face reflection is negligible.
The angular distribution of surface reflection is the surface reflection indicatrix.
The topological surface profile characteristics determine the distribution of the reflec-
tion. On mirror-like surfaces, all surface reflection is directed in the specular angle B = a
in Fig. 3. Even extremely smooth papers and prints that reflect almost specularly do not
give the impression of mirrors because the surface reflection coefficients are more than
an order of magnitude smaller than those of the metallic materials used in mirrors. With
an increase in roughness, surface reflection becomes increasingly diffuse. Figure 3
shows a diffuse distribution.
The angular distribution of diffuse reflection has the functional form of RcosB
where R is reflection coefficient. A diffusely reflecting surface looks equally bright from
all viewing angles even if the illumination is directional. This is why reading from dif-
fusely reflecting surfaces is comfortable. The color of such a surface is also independent
of the viewing angle. This result comes from multiplying the distribution RcosB by the
area seen by aperture A at different angles — A/cos. On directionally reflecting sur-

195
CHAPTER 7

faces, color varies with the angle of view because the proportion of white light in all light
is not constant.

Specular reflection Diffuse refelction

Figure 3. The two extremes of surface reflection distributions — specular and diffuse.

Measurement of the gloss of papers and prints?‘” is usually by specular gloss —

the degree of surface reflection directed in the specular angle. Registered gloss read-
ings relate to a smooth surface scale that has the same refractive index as the sample.
A smooth surface has a gloss of 100%. Gloss is therefore a measure of smoothness. A
diffusely reflecting or matte surface has no preference of reflection at the specular angle
by definition. The reading obtained with a gloss measuring device is small making
assessments of the degree of matte difficult. Other means of measurement are needed
for that. The magnitude depends on the aperture sizes.

3 Internal reflection

3.1 Solid prints

Toners and printing inks are primarily light absorbing materials. A light absorption
coefficient can depict their behavior. According to Lambert-Beer’s law, the intensity of
light (7) when it penetrates through a layer of infinitesimal thickness, dx, is attenuated in
proportion to the product of the intensity (7) and absorption coefficient (k,)
dl
aa ap (1)
Integration over a finite thickness, x,, gives an exponential relationship to trans-
mitted intensity vs. thickness:
—k Xx,
Tee toe (2)
where J) is_ incident intensity
ky the absorption coefficient.

The ratio, //Io, is transmittance, T,. Instead of ink film thickness, Eq. 2 can use the
amount of the coloring material per unit area by scaling the absorption coefficient, ky,
with the inverse of specific weight, p, as follows:

196
 Optical imaging in printing

In solid prints, transmitted light penetrates into the background substrate in which
a portion depicted by coefficient, Rape reflects backward. Scattering events in the sub-
Strate diffuse the light and randomize polarization. Back reflected light penetrates into
the absorbing layer and attenuates again consistent with Lambert-Beer’s law. The frac-
tion of light emitted from a continuous tone image is as follows:

aa: —k x, R —k x, —2k x,
mae () paper cs lGRaner’ (4)

In absolute density units, Equation 4 becomes

Is 2k, x log (= log (o paper (5)

Calibrating the measurement to give a zero value for density measured from the
substrate eliminates the final term. Density is then

i
De =N08 107 aa = log oR int = 2k,x,log
ie (6)
0** paper

A model shown in
Fig. 4’ predicts that density A
increases proportionally to ?
the ink film thickness. This is |
true at the low and medium Controlled
parts of the ink film thickness by gloss
scale. With an increase in
ink film thickness as the —— Controlled by
light absorption
magnitude of internal reflec-
tion diminishes, the amount OF
DENSITY
SOLID
AREA
of surface reflection reaching
the detector of a densitome- INK FILM THICKNESS

ter increasingly controls den- _ Figure 4. Relation between ink film thickness and image density.
sity. Density is normally
measured using incidence at 45° relative to the surface normal and detection at 0°
(45°/0°). This means a perfectly glossy surface has no surface reflected light reaching
the densitometer. If the surface is matte, 4% of surface reflected light registers in reia-
tive units. The level of gloss therefore determines the saturation level of density reached
with an increase in the ink film thickness as Fig. 4 shows. On a matte surface, the level

197
CHAPTER 7

is 1.4 (= log 0.04) density units. Polarization filters can eliminate the influence of sur-
face reflection in density measurement. This is due to the previously mentioned fact that
the direction of polarization does not change when light is surface reflected but does
change in internal scattering events.
To predict the density
of two or more multi color Incident light Reflected light
overprints, Figure 5 shows lo
that Lambert-Beer’s law
applies successively to lay-
ers using light intensity
transmitted from a previous Le
layer as the incident light Te
intensity to the next layer. R
paper
The intensity of light pene- apes es
trated through two layers is
as follows: Figure 5. Reflection from a multi color overprint.

Eee
La ere ee TeTT (7)
where subscript 2 refers to the bottom layer.
As light travels backward through the print, light intensity is similarly attenuated
so that the intensity of reflected light is the following:

ye
[= 1)1,\T,R paper Pi Ty = Lo T Raper (8)

Print density calculated relative to the paper becomes

D = -log510 = =l6e (lf. DDS (9)

Io paper

where the densities of the single layers without the influence of surface reflection are D,
and D,. By neglecting the surface reflection, an additive law for the density of over-
printed layers results. Overprinted density is the sum of the densities of the ink layers.
The approach extends to any number of layers that do not scatter light noticeably.
For colored inks, the light absorption coefficients, k, are functions of the wave-
length of light. Strictly speaking, the above expressions therefore refer to densities
determined by monochromatic light. By defining average absorption coefficients, the
expressions can apply to measurement of density through red, green, and blue filters.
The values are filter densities.

198
 Optical imaging in printing

3.2 Halftone prints

Examination of optical image formation in halftone prints?°:®'9can use macro and
micro scales. Macro scale reflection is averaged over an area that is sufficiently large to
include a statistically representative number of halftone dots. Reflection is the sum of
reflection from the dots weighted by the relative halftone area and reflection from the
substrate between the dots weighted by the proportion of unprinted area. Mean density
is the logarithm of the sum reflection scaled relative to the incoming light intensity or
reflection from the paper. The Murray-Davies expression in an earlier chapter uses the
assumption that the dots are ideal and the substrate is a perfect scatterer.
The Murray-Davies model does not consider the fact that light in the paper dif-
fuses sideways. Diffusion has the consequence that light that has hit the print between
the dots may emerge through the paper and vice versa as Fig. 6 shows.
The extreme case of
light diffusion is total diffu-
sion. This means that light {
reaching the paper does not A
“remember” whether it pene-
trated through the dots or =meee
through the unprinted area Ege pe a
of the paper. By using “a” to Figure 6. Illustration of light diffusion in paper.
denote relative halftone area
and k; and x, in the same meaning as previously, the total relative amount of light trans-
mitted into the paper in a halftone print is the following:

I trans _
ee ea) (10)
Io

The total lack of memory means that fraction a of the amount given in Eq. 10
goes into the dots and is attenuated by K.,,,,.exp(—k
x,) but fraction (1 — a) is reflected
from the paper. This gives the following for total reflection scaled relative to the paper:

el fo = [1-a(1-T,)} (11)
I,R paper

Comparison with the Murray-Davies equation shows that light spreading in the
paper causes a drop in total reflection and a concomitant density increase. The trend is
the same even without assuming that diffusion is total. Total diffusion is likely to occur
only at high screen rulings where the dots and distances between the dots are short.
The mean distance of spreading and the shape of the spreading function are properties
of the paper.

199
CHAPTER 7

An effort to depict spreading by a single parameter has led to the Yule-Nielsen

model? in Chapter 3, Section 5.1:
—D./n (12)
iol (1 — a) at ald é |
Dins = —nlog

The model reduces to the Murray-Davies model when the spreading parameter,
n, equals unity, i.e., there is no spreading. Relative halftone area calculated by the Yule-
Nielsen model from measured densities gives an approximation of the halftone area.
The accuracy of the model is still deficient because it uses simplified assumptions.
A better approach depicts optical spread by a point spread function®.

Solid 7 | Filling-in
density
it | _——

Se _No spreading Z_
[= xt
a) G} /
=z =
Oo _ Input =
_/ percentage

Output
_}~ percentage |
50 100 i ; 100
HALFTONE, % HALFTONE, %

Figure 7. Definitions of dot gain and filling-in (left) and a typical dot gain curve (right).

Optical spreading leading to a rise in density causes an increase in the halftone

percentage similar to other types of point spreading phenomena. In general imaging ter-
minology, the phenomenon is blurring. The increase in effective dot percentage is dot
gain. The adjective “effective” is important because the dot percentage increase is cus-
tomarily determined indirectly by calculation from optical measurements as Fig. 7
shows rather than quantitative measurements of the amount of ink. The curves in Fig. 7
are tone reproduction curves. The curve marked “no spreading” represents the Murray-
Davies model.

D = -logjl1-4;,+4;,10 —Ds
7] (13)

The other is a measured curve. D is halftone density, D, is solid density, and q;,, is
relative input halftone area. Dot gain is the difference a,,,,, — aj, at given levels of density.
Mathematically, dot gain is the following:

wane 1-10" |, aK
out un fie 10 — in

200
 Optical imaging in printing

Spreading inevitably causes halftone dots above a given dot percentage to

merge. This condition is filling-in.
In multi color halftone printed images, halftone dots have mostly random posi-
tions relative to one another because of color register variations as Fig. 8 shows. If the
inks were completely transparent and the paper were a perfect reflector, positioning of
the dots would have no influence on color. This is not the case and overlap infuences
color.
Assuming randomness the relative
halftone area of a color is its probability of |
occurence. The probability that a given ©
ink, 7, occurs in a unit area is thus a; In
the case of a two color halftone print, the
probability that inks 1 and 2 overlap is the a
product of the respective relative halftone
areas, ad). The probability that ink 1 is Figure 8. Illustration of random overlapping of half-
printed as a single color equals its proba- tone dots.
bility, a,, times the probability that ink 2
does not occur, (1 — az). Using this argument for two color printing gives the relative
areas and respective reflectances indicated in Table 1.

Table 1. Different combinations of area in two-color halftone printing.

Relative area Reflectance

Only ink 1 aay Ri |
Only ink 2 ay (1—ay) R>
Inks 1 and2 aay Ri2 =R Ro
Paper (1 — ay) (1 — ay) Raper

Reflection of the print is the sum of the reflectances weighted with the relative
areas:

R = (relative area; x reflectance) (16)

i
The reflectances are either continuous functions of wavelength or vectors with
three components that represent reflectances measured through three filters. The
model in Table 1 is that of Neugebauer in Section 5.1 of Chapter 3. Density is the nega-
tive ten base logarithm of reflection calculated from Eq. 15.

4 Colorimetric color
The earlier discussion of optical image formation considered a densitometric viewpoint.
Density values measured from monochrome black images give a good prediction of
how dark the image looks. Filter densities of colored images do not represent perceived
color well. Colorimetric color values explained in Section 3 of Chapter 3 agree more with
visual perception.

201
CHAPTER 7

Determination of colorimetric color coordinates can use the following three meth-
ods in order of improving precision:
- measurement of RGB filter densities and their conversion into colorimetric
coordinates using experimental calibration matrices

- reflection measurements through tristimulus filters by colorimeters

- reflection measurements of wavelength spectra by spectrometers and compu-

tation of colorimetric values.

In solid prints, spectral reflection consists of surface reflection that is spectrally

nonselective and internal reflection that is colored. The spectral transmittance of the ink
and the spectral reflection of the paper influence the internal reflection in a multiplicative
fashion. Using p for the coefficient of the first surface reflection while neglecting the
internal surface reflection, spectral reflection from a print is as follows:

TA) = Ig AEp += poe Rayer] (16)

Because the surface reflection is white, its exclusion from the viewing angle of
specular reflection favorably influences the color. Color of matte prints does not vary
with viewing conditions because the influence of surface reflection remains the same.
A consequence is that the spectral composition varies with the thickness of the ink layer,
x,. The color does not remain the same.
The color of the background paper substrate also influences color of solid prints.
This is because the colorimetric color coordinates defined in Section 3.3 are computed
without normalization to the paper background. In otherwise similar conditions, print col-
ors are less chromatic on low brightness papers. Density values are usually expressed
by calibrating the substrate to give zero density. This difference is due to a stronger ten-
dency of the human visual system to adapt to the background luminance rather than the
background color.
The influence of the background is still smaller than what it is in halftone areas
where the effect of the reflection from the paper is not multiplicative but additive as
expressed by the Neugebauer model. Written for a single color print, the model takes
the form:
x halftone Sirsa yeaa
paper dot

Leralfione = (1 Ss a) Papen Tt AY gor

(16)
Ziatione = (1 i a)Z paper ae dot

where the XYZ-values are the CIE XYZ color coordinates. The values can be
transformed into the CIE Lch (Section 3 in Chapter 3) coordinates for easier
comprehension. Figure 9 illustrates'*how the hue angle value changes with an
increase in the halftone percentage.

202
 Optical imaging in printing

=. 180
Lu

=
FE
fe)

WZ
O¢
LU OL
On 90
FE Color
Lu 5 of solid
raap) printing
Le
wS
Ot
ob a
s

Wu) ie 5 100
az(ge HALFTONE, %
Za
= [al
Lu
=)
a5 ce)j=)

Figure 9. Hue angle, /, in halftone printing.

Light ( < 50%) and medium (50%—70%) tone halftone prints do not have the
same hue as a Solid print. The hue is intermediate between the hue of the paper and the
hue of the solid ink. In an example where the paper has a yellow shade and the ink is
magenta, light tones are orange, and medium tones are shifted towards red. The phe-
nomenon is a weakness of halftone printing and causes a decrease in the color range in
light and medium tones. Note that the hue of paper here means the almost inevitable
hue of any paper not colored substrates.
An ideal case is one where the print attains the hue of the solid print indicated by
the dashed line in Fig. 9 even at infinitely small percentages. If the hue difference
between the paper and the ink is zero or the hues are complementary — hue difference
180°, the ideal case results. The reason why complementary hues are ideal is that the
color of the ink compensates for the paper color and paper color does not influence the
hue.
In reality, the paper color and the color of the inks cannot meet the ideal condition
at the same time for the cyan, magenta, and yellow inks. It is possible to optimize paper
color if some order of priority can be given to reproduction of different colors. For
instance if light reds are important, the paper color should be red or cyan.

5 Optical behavior of the substrate

The optical behavior of printing inks was characterized above by an absorption
coefficient consistent with Lambert-Beer’s law. If a material absorbs and scatters light

203
CHAPTER 7

as paper does as discussed in another book of this series, the simplest approach is to
characterize the optical behavior using two material parameters: absorption coefficient,
k, and scattering coefficient, s. More elaborate approaches are also possible ’°.
The Kubelka-Munk model is a solution’? of the general radiation transfer equa-
tion. It assumes irradiation of a substrate with a diffuse, monochromatic illumination.
The transfer of light in a plane parallel layer such as paper uses two differential equa-
tions. The first equation relates to the transfer of light downward in the layer, and the
second equation relates to the transfer upward:

au
7 = -(k+s)l+sJ
- (18)
— = -(k+s)J +s]
dx

where J __ is the light intensity travelling downward

d) the intensity moving in the opposite direction across an infinitesimal
layer thickness, dx, of the material.

Figure 10 provides an
illustration.
Intensity, 7, decreases
Illumination
by light absorption and scat-
tering from the, /, flux to the,
J, flux and increases by scat- xed
tering from the, J, flux. The, Ps
J, flux is attenuated by |
absorption phenomena and oS
ES J
amplified by scattering from
the, J, flux. x=0
Solving the equation’
Figure 10. Definition of coordinates in the Kubelka-Munk model.
gives two reflectance terms:

- reflectance of a layer of thickness, d, measured against an ideal nonreflecting

background denoted by Ro
- reflectance of an “infinitely” thick pile of the substrate denoted by R...

Equations 19 and 20 show the results for each case:

k RY oR
R Gn (*) +2- (19)
R) 5

204
 Optical imaging in printing

_ exp[sd(1/R,,-R.,)]-1 (20)
om (yr Dexp(sdd/R.—R_))—R.

The significance of the Kubelka-Munk model is that measurement of Ro and R..

with a photometer allows calculation of absorption and scattering coefficients, k and s,
for a given substrate. Both coefficients are in units of [( ,um)~'] or [m2g~'] when replacing
the thickness, d, in the model by the grammage in gm. This is the convention for
paper.
Measuring R., with blue light having its peak intensity at 457 nm gives ISO or
TAPPI brightness. Brightness closely correlates to human perception of reflectance’.
The corresponding CIE Y-value is called lightness. Paper luminance is proportional to
the cubic root of lightness. The degree of paper opacity is the ratio of Ry and R...

opacity = R)/R,, (21)

In logarithmic units, as —l0g19(Ro/R..),gives the lower limit for the visibility of print-
ing through the paper. This is because penetration of ink reduces Ko of the paper sheet.
The human visual system is assumed to respond logarithmically to reflection.
Expression of paper color may use any color coordinate system, but the CIE Lab
representation is most common. A fully achromatic paper would have zero chroma, and
the hue angle would be undefined. The hue of real papers usually deviates from neutral.
Investigators have suggested several expressions for calculating the whiteness of paper
from CIE color coordinates’°.The formula recommended by CIE (1981) is as follows:

Wot S007 x) 1 7/00(y, + y) (22)

where xandy_ are the CIE chromaticity coordinates

x,andy,, the coordinates for the perfectly reflecting diffuser in
the D65 illumination.

In the optical sense, the sideways light diffusion in paper limits the detail repro-
duction capability of papers. The phenomenon is the point spread function defined in
Section 4 of Chapter 2 and discussed in the context of color calibration in Section 5.1 of
Chapter 3. The optical point spread function shown in Fig. 11 is the distribution of light
reflected from the paper at given distances from the point of incidence.

205
CHAPTER 7

tee Reflection
Illumination distribution

Figure 11. Illustration of the light spread function, h(x,y), of paper where w is the mean distance of
spreading.

Light traveling further before backscattering gives flatter distribution. The mean
distance of spreading, w, is the distance at which the distribution has diminished to eo!
times its peak value. A simplified derivation suggests that

Ro
0 50 (23)
when Ro and s are as defined above of w and 9 is the specific weight of paper. The
values, have an order of magnitude of 10-104 um. Earlier discussion considered optical
point spreading as a source of dot gain. This causes broadening and blurring of lines
and losses in detail and sharpness. The influences can be mathematically predicted
using the assumption of system linearity as outlined in Section 4 of Chapter 2.
The loss in detail rendition induced by optical point spreading also occurs as a
loss in optical information capacity. | Information capacity essentially expresses the
number of tones given as the log» value, i.e., bits that can be rendered per unit area.
The resolution of density and color levels and spatial resolution influence this. Optical
spreading causes deterioration in spatial resolution.
The ratio of the reflectance range, S, and noise, N, determines the resolution of
density levels at zero spatial frequency. At finite frequencies,f, the available reflectance
range, is reduced by the MTF to

SC) = (0) MIC) (24)

206
 Optical imaging in printing

Light absorption in paper and less than ideal brightness, Ro, limits the reflectance
range within which information may be stored to

where Rjog is reflectance of a solid print.

Optical noise in papers is fairly small — typically somewhat above one percent of
the reflectance. This suggests that optical S/N for paper is at best approximately 10° (40
db). At worst, it is approximately 10 (20 db).

6 Print noise

Noise /%4 is uncontrolled —

deterministic and random — Signal Noise
variation of a signal. In
prints, noise is variation of eres Bee Gloss
the image signal in the x,y- reflection |---..
plane of the image. In optical
image formation, the input
signal is illumination, and the
' ; Internal
output signal is reflection Density
emitted by the image.
Reflection consists of
surface reflection and
internal reflection. Image
noise may arise from
variation in either or both of
these signal components. Figure 12. Types of noise.
Figure 12 shows that images
may be noisy in gloss, density, or color. Variations in surface reflection contribute to
gloss noise and may also influence densitometric and colorimetric noise.
The root-mean-square variation of a signal (rms-noise or standard deviation o
or N) is the basic measure of noise.
The common expression of image noise is the signal (S) to noise ratio in loga-
rithmic units (base 10) as decibels

SNR = 20logio(S/o) (26)

or as the inverse of the linear signal to noise ratio, the variation coefficient

var coeff(%) = 100(0/S) (27)

207
CHAPTER 7

The distribution of noise variance with respect to spatial frequency is the power
spectrum.
Noise occurs on a scale that is characteristic of the origin of noise called the
noise source. It is practical to identify the following scales:

- macro scale noise characterized by a dimension called wavelength from 1 mm

upward

- micro scale noise with a characteristic dimension between 10 um to 10° um

- invisible scale where the dimensions are smaller than 10 um.

The power spectrum provides a continuous representation of noise at different

scales. Section 6.2 in Chapter 8 discusses this issue further from the viewpoint of print
uneveness. Figure 13 shows the three scales in relation to the visual contrast sensitivity
function. The function gives expression to the inverse of the threshold contrast required
to generate a visual response at a given wavelength. A high value of the function means
a small threshold, and a zero value means infinite threshold. When noise is greater than
the threshold it is visible.

Visual contrast
sensitivity function

AVAVAVAVAVAVA
Micro scale noise

vw

Invisible noise
FREQUENCY —>
<— WAVELENGTH
Figure 13. Illustration of noise of different size scales.

The sources of noise in prints@° relate in Fig. 14 to the signal itself, the imaging
process, and the materials. As an example, a halftone signal is noisy because of its spa-
tial discontinuity. The image processing steps cause noise whose magnitude can be
computed. The magnitude can only be influenced by a different choice of imaging
parameters such as the number of signal quantization levels.

208
 Optical imaging in printing

Macro scale noise

sources in printing pro- TS
cesses include long term | orig
variations in ink feed. In 5 | Ink-paper
printed matter, these tempo- = interactions
rary variations transform to c Halftone
spatial variations between = structure Material
editions and copies within a ee
printing run. Unevenness in
$5: ate FREQUENCY—>
prints arising from variations <— WAVELENGTH
in paper properties due to
less than perfect formation Figure 14. Sources of noise at different scales.
of the paper is a macro scale
noise.
The human visual system does not perceive macro scale variations of density
and gloss very sensitively if they do not occur within a sheet or if comparisons are not
made between two copies. Color variations are visible and easily detected if the colors
are so-called memory colors such as popular trade marks.
The fine structure of materials and interactions between them contributes to
micro scale noise and invisible noise. Examples of the fine structure of materials include
silver grains, toner particles, and the particle structure of paper surfaces. The latter is
the source of roughness and porosity of the paper that interacts in imaging with toner
particles and printing ink. These interactions lead to variations in the amount and pene-
tration of the coloring materials at the paper surface. When illuminated with light, the
variations transform into optical noise. At the higher end of the wavelength range, micro
scale variations are well perceived. Visually, micro scale noise masks the image signal,
i.e., prevents perception of image information. Although invisible noise is too fine to be
visually apparent, it does influence the mean value perceived. As an example, the
human eye does not distinguish the variation of gloss caused by inherent paper rough-
ness. Roughness does influence the mean level of gloss perceived because it is a spa-
tial average.

209
 CHAPTER 7

M@References iy
1. Kortum, G., Reflectance Spectroscopy. Springer-Verlag. Berlin 1969. p. 366.
2. Bryntse, G., A Method for the Analysis of Ink Mottle Using Polarized Light
Reflection. Royal Institute of Technologydepartment of Paper Technology.
Dissertation 1981. p. 88.
. Yule, J.A.C., Howe, DJ, Altman, J.H., TAPPI Journal 50(7):337(1967).

. Yule, J.A.C., Nielsen, W.J., TAPPI Journal 50(7):337(1967).

. Oittinen, P, Saarelma, H., TAPP! Journal 65(1):47(1982).
SD}. Gustavson,
EN
Sqy
py S., Dot Gain in color Halftones. Dissertation. Linképing University. Dept.
of Electrical Engineering. Linképing Studies in Science and Technology. No. 492,
1997.
7. Leekley, T.M., Tyler, R.F., Hultman, J.D., The Effect of Surface Reflection on Color.
TAGA Proceedings 1978, Technical Association of the Graphic Arts, NY 1978
pp. 291-3117.
8. Yule, J.A.C., Theory of Color Reproduction. John Wiley & Sons, New York.
9. Hunter, R.S., Harold, R.W., (ed.), Measurement of Appearance. 2. ed, New York,
Wiley 1987. ISBN 0-471-83006-2. p. 411.
10. Micale, F., Am. Ink Maker 62(9):20,22,40,41(1984)9.
11. Oittinen, P, The Surface Structure of Coated Paper an the Formation of Gloss. The
Role of Fundamental Research in Paper Making. (Ed. J. Brander), Mechanical
Engineering Publications Ltd. London 2:635(1983).
12. Tollenaar, D., Ernst, R, Optical Density and Ink Layer Thickness. Advances in
Printing Science and Technology (Proc. Intern. Conf. Printing Res. Inst.) (IARIGAI,
ed. W.H. Banks). 2:214(1962).
13. Ruckdeschel, F.R., Hauser, O.G., Applied Optics 17(21):3376(1978).
14. Oittinen, PR, Autio, H., Saarelma, H., J. Imaging Science and Technology
36(5):496(1992).
15. Leskela, M., Simulation of Particle Packing for Modelling the Light Scattering
Characteristics of Paper. PhD Thesis. Helsinki University of Technology. Laboratory
of Paper Technology Report A8. Otaniemi 1997.
16. Kubelka, P,, Munk, F,, Zeitschrift fur technische Physik 12:593(1931).
17. Lim, PY.W., Steinlage, R.C., Gantner, T.E., Application of Fuzzy Mathematics to the
Analysis of Human Perception of Brightnes of Office Papers. IS&T’s Tenth

210
 Optical imaging in printing

International Congress on Advances in Non-lmpact Printing Technologies,

Springfield Va 1994. pp. 509-515.
18. Bristow, J.A., Brightness and Whiteness — Definition and Measurement. Advances
in Printing Science and Technology. (Ed. W.H. Banks). Pentech Press, London
1990, pp. 193-217.
19. Barten, PG.J., Evaluation of the Effect of Noise on Subjective Image Quality.
Proceedings of SPIE. Human Vision, Visual Processing and Digital Display II,
1453(1991), 2-15.
20. Dainty, J.C., Lehmbeck, D.R., Triplett, R., J. Imaging Tech. 11(5):131(1985).
21. Hamerly, J.R., Springer, R.M., J. Opt. Soc. Am. 71(3):285(1981).
22. Oittinen, P, Saarelma, H., Kuvatekninen /aatu (An Imaging Approach to Quality).
Otatieto, Espoo 1992. (In Finnish).
23. Saarelma, H., Oittinen, P, Graphic Arts in Finland 20(2):8(1991).
24. Shaw, R., Burns, PD., Dainty, J.C., Particulate Model for Halftone Noise in
Electrophotography. | Theory. |! Experimental Verification SPIE Image Quality
310:137,143(1981).
25. Saarelma, H., Oittinen, P, Paper and Print Noise as Limiting Factors of Information
Capacity. Products of Papermaking (Ed. C.F. Baker). vol 1. Pira International.
Leatherhead UK. 1993. pp. 351-365.

211
BOHR ———
Paper in printing

1 Printing apetir.: <ccsc.sctaseeeccrcecaareaecgemetts: tactnes eects eectens emma et arene 213

2 Web: transport tints cea veosees teh sou ce Set, cadens terete be en, seen eee USmeeeata rote eee 216
3 SHEGUTANSPOMl =. hace cCidectaves coders tarcse fe tenceneene reas nate centres aa taa aes Sameer 221
4 FIMISU NO cac cece retey: c ceec oe Soe tca ava attaeie dace arian: na eee ee an ee eee 223
5 RUM aM itytecerrercscrerccces recent ce teucease seme cat cee cane eee anteeatte te ea aeeeiee eee neta 223
Bet We DIDO AKS st otetese anak Bast arece alata tedaerah Ord eo Gere se a Reece em a
52> SONGOtTEC G:PFODIG MS Cewccars eiccsey, sexes cateee cette cee ec etela tee ak eee te ere 22%,
Bids PPESIINUUING Seo cces scacedetsortes cakindots deaucitad cunt sbuetedsele excan unease a etn eee yaaa ec 229
Bhs MeLS VIIIGh eee cee eetcatscdexesne as ty Dotrprera eea dave terion va kame ys telet ae en oe e 231
6 PRINtADUINY Saree ructsceetste teveneeuaueabrenscncnettaeecesadtcisne tetas nee ace ceeee nae amas 231
6.1 PUIG! DIGS setscenke seoes sce erce cisssstescectacsvededone vides eta eet te cathe ee aeoe ne tc ele nee ter 231
Gi2> -<PrintabilitVwor-SOlid: PUINTS® ix cscuswceeceutocscletecotusre’ cea tsceee cence meme octet ee eee ee 233
6.32 SPrintabslityotimalllOnes-a.,: cccoscccvcc aucseu rete os et arretescde aeeeke tenet c eis gee Zar
7 Paperias carrier Ofinformation. <ccc.cccusceune-geceus<cctocstcescecacescvesvecseesteess veseees 239
Teh Mahe GONG Fella esata, Ace ale seach as feed ihc A Ue a As Pad apc en ye occ 239
(eZ SAAIMCOLMARIONHGADACILY sezesece ie:ciaac.tatneee cea cersrcansaeesecececstevnet ewatiee eet ton cee ee 241
fase = OpCal IAT OL Mato MCaDACILY: OF DADE ae sama sete cesssec 243
AP COLO FOR VAtION CApaCtVOl DADGl. ca. aeecee use argsecates te eee eee 247
8 Testing the printing potential Of paper ................ccccccececscecececeecscersceuenserenens 249
PROT GIOIICES Stes eds reeticircs er vance arth cos eters ecient ee eo ee 253

readpe
 CHAPTER 8

ePaper inprintin =z
Paper enters a printing press as a raw material and exits as physically finished or
semifinished printed matter and a carrier of information. Paper therefore undergoes
transformation in two different ways, as a physical material and as an information
surface.
Paper is definitely the most important raw material in printing — printed communi-
cation is almost synonymous with paper. Various other substrates including board,
metallic and plastic films, textiles, and semiconductor materials are also printed. As fin-
ished products, the other substrates fulfill functions besides information carrier that may
be far more Critical.
The performance of paper in printing depends on its runnability, printability, and
information capacity. Other important properties are those that determine product
usability and post consumer properties. Runnability relates to the factors that determine
how the paper runs in the press. “Good” runnability is a prerequisite for efficiency in
printing production. Its significance increases with the scale of printing. Printability fac-
tors concern the quality of paper in relation to ink and printing. Discussion of paper
using terms that allow its characterization as information carrier makes it possible to
compare the quality achieved in media such as prints on paper, TV broadcasting, or
computer display.
This chapter focuses on the principles of the operations, devices, and equipment
used to transport and finish paper in the press. It also defines major runnability, print-
ability, and information concepts and briefly discusses them. The chapter begins with an
overview of printing paper grades.

1 Printing paper
The term paper grade as used here denotes papers that share the most relevant
criteria. The main criteria used to classify printing papers include

- fiber furnish of the base paper

- coating

- type of surface finish

- end use.

213
CHAPTER 8

Table 1 identifies the main generic grades of printing papers.

Table 1. Generic printing paper grades.

Paper grade Description

MF (machine finish) news Consist mainly of mechanical fiber (= wood containing) used
MF special in newspapers, directories, and paperback books
SC (Super calendered) Uncoated wood containing gravure and heat-set offset paper used
in magazines and catalogs
ULWC (ultra light weight coated) Coated, possibly wood containing gravure and heat-set offset paper,
matte or super calandered used mainly in catalogs
MFC (machine finish coated) Coated (pigmented) wood containing offset paper, soft calendered in-
line with paper making, coating weights below those in LWC papers
used in magazines and advertising
LWC (light weight coated) Coated wood containing gravure and heat-set offset paper, super cal-
endered off- or on-line used in magazines and catalogs
LWC matte As above, matte calendering
MWC (medium weight coated) Double coated wood containing offset paper, coating weight above
MWC matte the level in LWC used in special interest magazines, advertising, art
books, and ink-jet printing
HWC (high weight coated) Double or triple coated wood free offset paper, coating weights above
HWC matte the level in MWC grades
Uncoated fine paper Uncoated wood free offset paper, used mainly in book printing,
copying, and electronic printing
Ultra light weight wood free paper | Uncoated or coated wood free “bible paper” used in offset printed
books and directories

The classification is clearly a hybrid since it includes manufacturing and composi-

tional aspects. This has historical reasons. In the past, paper grades were far less
numerous than today allowing identification by only a few criteria. From a printing stand-
point, a classification based on paper properties or specified application is more rele-
vant than one based on paper manufacturing variables. The generic paper grades
defined in the table give a good indication of the level of many paper properties such as
basis weight, brightness, and roughness without defining them explicitly.
Basis weight in printing papers is approximately 20-200 gm, i.e., its range cov-
ers an order of magnitude. Fiber based substrates of higher basis weights are paper-
board or board. The range of brightness covers approximately 50 percentage units, and
the range of roughness is approximately an order of magnitude. Figure 1 gives a sche-
matic positioning of the grades in a map defined by brightness and roughness.
Mechanical, chemical, and recycled fibers constitute the main types of furnish as
discussed in other books of this series. These types have various subclasses. Mechani-
cal fibers provide opacity due to their fine content and printability because of their com-
pressibility and deformability. Chemical fibers are the source of paper strength and
brightness of the chemical fibers. Strength is why very low basis weight papers consist
primarily of chemical fibers, i.e., are wood free. Surface sizing improves the surface

214
 Paper in printing

Uncoated
fine
papers

wood free weight WF

papers
HWC -
D WF
Wood containing book papers
Lu
> MWC
FE
MFC
O LWC LWC
or gravure Offset
faa)

SC
Peete SC offset MF special

MF news
ies
>
ROUGHNESS

Figure 1. Brightness and roughness of various paper grades.

strength and printability. Surface strength is necessary in offset printing due to the
presence of dampening water. Printability aspects determine why gravure papers are
generally mechanical fiber based and why they are smoother than offset grades of the
same generic grade. This is evident in Fig. 1.
Pigments such as Clay, talc, and calcium carbonate and surface finishing further
enhance the printability of fibrous sheets as other books in the series explain. In
uncoated papers, mineral pigments are dispersed in the fiber furnish as a filler. In
coated papers, a discrete pigment layer is applied on the base paper and bound to it by
polymeric materials. Coating provides more opportunities to control the surface proper-
ties, the typological profile, and the porosity and is more effective than filling. These
opportunities require separate manufacturing steps in-line with paper making as in the
case of MFC papers and possibly also LWC papers or off-line. With an increase in the
coating weight from MFC and LWC papers to MWC and HWC papers, brightness and
smoothness improve. The choice of raw materials can control porosity. MWC papers are
produced by applying two separate coatings (“double coated”) on the base paper, and
HWC papers are made by applying two or more coatings.
Surface finishing involves calendering the paper under pressure in nips formed by
rolling cylinders. This may occur in-line with paper making using machine finishing with
hard rolls made from cast iron or with soft polymer covered rolls. The latter distribute the
calendering pressure more evenly creating a more uniform pore structure and surface
finish. Heat and moisture may be applied to promote flow in the calendar nip and

215
CHAPTER 8

improve smoothness. Rolls filled with fibrous material can be employed to increase
gloss in an off-line supercalendar. Calendering modifies the surface properties through
mechanical and thermal mechanisms. Matte calendering reduces roughness and
porosity without improving smoothness or destroying the matte finish. This statement
assumes that smoothness is the relative proportion of plane-like or flat surface and
roughness related to deviations from the mean depth. In other words, smoothness and
roughness depict the surface profile as separate concepts.
Table 1 suggests that the number of gravure paper grades is smaller than the
number of offset grades. This is due to the mass production nature of gravure. In addi-
tion, the properties of gravure papers fall in a narrow range as Fig. 1 shows. The typical
basis weight range of gravure papers is approximately 40—100 gm.
Paper used in black and white copying and laser printing is traditionally uncoated
wood free paper of high brightness, although recycled copy and laser papers have their
own applications. The basis weights are usually ca. 65-100 gm*. The demands for
trouble-free running and archiving in office environments mean that the paper proper-
ties require narrow tolerances. This will be less important in the future, but more strin-
gent requirements for printability will be necessary. In electrophotography, the
requirements are the result of higher resolutions, color and duplex printing, and higher
printing speeds. This change has led to need of coated electrophotographic paper
grades with basis weights above those of uncoated grades for use primarily in digital
color printing.
High quality ink-jet paper grades are mostly still specially coated grades opti-
mized to function on specific printing devices. Photoprinting requirements emphasize
this approach. In the office sector, developmental efforts have the goal of producing uni-
versally functioning ink-jet papers that are not necessarily coated. Wide format ink-jet
for signage and display advertising uses coated ink-jet papers.

2 Web transport
In web fed printing illustrated by Fig. 2, paper proceeds through the press using web
tension.

3
é 5
1 2 <>

1 Web feed, braking 4 Folding

2 Infeed device, adjustment of web tension 5 Slitting, cutting
3 Printing units, registering 6 Delivery

Figure 2. Diagram of paper transportation in web fed printing.

216
 Paper in printing

Web tension in printing counteracts friction forces that occur in nips and from con-
tact with rollers and turning bars. Uniform web tension is also necessary for stability and
color register in the press. Color register means the accuracy, and color misregister
means the lack of accuracy to overprint the inks spatially. Color registering require-
ments may be +10 um at the greatest.
Issues related to stability include machine direction web vibration — flutter — and
cross directional sliding. Figure 3 illustrates how the influence of a disturbance acting in
the cross direction has less effect on web tension when the tension is higher. The rea-
son is due to the addition of two perpendicular vectors — one represents web tension,
and the other represents a disturbance. Paper properties exhibit distinct profiles across
the web. During web travel, these act as CD disturbances that a higher web tension
helps to compensate.

Lu Lu
O O
Ze FE
QO co Q a
2 = pect © = Result
D D x im>
Oo ra :
MD WEB TENSION MD WEB TENSION

Figure 3. A schematic illustration of the influence of a CD disturbance on two levels of web tension with
QQ, > A.

With an increase in
web tension, the risk of web
breaks increases. This is é
because the likelihood of an io
occurrence of spots in the fT re
paper that have lower a ie
strength than the web ten- G eS
sion increases. The control- oc t
lability and risk of web break = =
require a delicate balance as a
Fig. 4 shows. =
Web tension should Rl
be constant between the ecerenclor
printing units of a press so
that the colors register. Dur- _ Figure 4. Influence of web tension on color register and web break
ing running, web tension is frequency.
the primary adjustment for
color register. With other factors equal, constant web tension means that web elonga-
tion is also constant. Figure 5 illustrates the adverse situation. In the illustration, the
paper travels from left to right. If the conditions are such that the colors register when

21s
CHAPTER 8

elongation of the paper is ¢, a change in elongation causes misregister. If elongation

increases by A, then printing in the second nip is ahead of printing in the first nip by A.

Elongation Elongation
€ er

A |
Figure 5. Influence of web elongation on color misregister.

Figure 6 shows a typical MD web tension profile in heat-set offset printing. The
situation also applies to other printing methods. Web tension occurs at two steps: the
reel stand and the in-feed device. The tension decreases slightly when the web travels
from a printing unit to the next due to friction. The decrease is almost negligible for color
register. Drying causes shrinkage manifested as an increase in web tension. Cooling
and moisturizing have an adverse influence. Cutting of the web relieves web tension
totally.

eid Cooling

Folder

TENSION
WEB
1 2 3 4
PRINTING UNITS

Figure 6. Web tension profile in a heat-set press.

One can consider web tension at a given position on the press as a signal that
varies with time. Besides having a mean value, web tension has deterministic and ran-
dom variations. Figure 7 shows that it may also change continually with time. All rotating
elements of the press that touch the web such as guiding rolls cause deterministic
variatons. Their period is equal to the period of rotation. The condition of the bearings

218
 Paper in printing

determines the amplitude of variation. As wear occurs, the amplitude increases. In off-
set lithography however the most common periodic disturbauce in web tension arises
from the gap in the printing and blanket cylinders where the plates and blankets are fas-
tened. The pulses from the stagger bar can be eliminated by forming the plates and
blankets as continuous cylinder sleeves. Random variations arise from uncontrolled or
unknown sources. A reel change typically causes a peak in web tension. A trend may
originate from variations of paper properties in a reel from the surface to the core. Vari-
ations may have their origins in papermaking or reel winding. Winding should produce
even tension profiles across the radial and CD directions.

Peak
Random variation

TENSION
WEB

Period

TIME

Figure 7. Web tension signal.

Web tension causes elongation in the paper. As a first approximation, Hooke’s

law of elastic behavior applies:
web tension
Elongation
g
€ = ———————_
elastic modulus
(1)

The assumption of time independent elastic behavior does not hold if the web
tension varies rapidly. Instead the paper behaves time dependently, i.e., viscoelastically.
Reel changes are critical in considering web breaks. Reel change is the most fre-
quent single cause of web breaks. Large reels are therefore advantageous. The length
of paper, /, in a reel when omitting the influence of the core results from the following:

ihe Cee (2)

where R_ is reel radius
h paper thickness.

ZAG
CHAPTER 8

The expression is the first term of the mathematical series that depicts the length.
In practice, the influence of the core is negligible.
Through its influence on thickness, basis weight also influences the length of
paper in a reel. Assuming that the specific weight of paper (p) is independent of basis
weight (g), i.e. p is not a function of g, the relative increase in web length with a
decrease in basis weight is inversely proportional to basis weight because

In web fed printing, paper feeds to the press from a reel stand by rotation as
Fig. 8 shows. Except very small presses, the reels attach to the reel stand by chucks fit-
ted at the two ends of the reel core. When two reels can mount at the reel stand simulta-
neously, reel change can occur automatically. Figure 9 illustrates this for stacked reels
(left) and a reel star (right). The press can then run at full speed during reel change. This
is advantageous and necessary because a reel change in a one-web printing press
occurs every 10-20 minutes. Stopping the press would result in productivity losses. In
multi web presses, reel change frequency rises in proportion to the number of webs.

Reel stand (separate from the Reel stand (integrated with

press) for one paper reel the press) for one paper reel

aReels on top of one another Reel star

em
OQ
Figure 8. Different types of reel stands.

220
 Paper in printing

Reel changes with a reel star use dynamic conditions with the reels rotating at
press speed when making the change. The web becoming empty and the new web are
instantaneously pasted together with tape so that web is being fed to the press from two
reels momentarily. The web of the empty reel is then cut off. With stacked reels, the

ft
reels are stationary during reel changes. This is possible by using a festoon between
the reel stand and the printing units. Paper wound around rollers whose positions rela-
tive to one another can be regulated forms the festoon. The length of paper held by two
rollers is too times their distance. Web feeds to the press during reel changes from the
web store.

Figure 9. Automatic reel change using a reel star (left) and stacked reels (right).

Reel stands also have brakes for the following functions:

- acceleration and deceleration of the paper reel

- adjustment of web tension

- stopping reel rotation in emergencies such as web breaks.

Because paper reels may weigh several metric tons and the need for a fast press
response, the technical demands on the brakes are high.
Other elements that handle paper transport through the press include

- in-feed units consisting of roller assemblies

- idle rolls and bars that guide the direction of the paper transport

- driving nips. These are used for adjusting web tension as actuators in web
tension control systems.

3 Sheet transport
The principles that apply to transfer paper in sheet-fed presses differ considerably from
the principles in web fed presses. Figure 10 is a diagram of paper in a sheet-fed offset

ods
CHAPTER 8

unit. Sheets are transported as separate objects. Sheet passage from an in-feed pile
through a press to an out-feed pile occurs by the following:

- movement of mechanical or pneumatical elements that attach to sheets one at

a time
- transfer by rollers holding the sheets by mechanical or pheumatical forces or
to which they adhere by friction.
These comments suggest that sheet fed printing requires some degree of stiff-
ness in the substrate. Because thickness influences stiffness and basis weight influ-
ences thickness, the basis weight of paper printed as sheets should be above a level of
ca. 60 gm~*. The main use for paper in sheet form is sheet fed offset, packaging flexo-
graphy, and electronic printing.

1 Lifting of sheets from sheet pile using suction

2 Transfer to feeding table, registering
3 Transfer with swinging grippers to impression cylinder
4 Transfer of sheet to printing nip, holding the sheet with grippers
5 Transfer from nip with gripper chain
6 Delivery to pile

Figure 10. Automatic reel change using a reel star (left) and stacked reels (right). A schematic illustration
of paper in sheet fed offset printing.

In mechanical print-
ing, sheets are lifted from a Air suction gripper
paper pile pneumatically
with suction grippers. Trans-
portation in the press is by
Mechanical gripper
mechanical grippers. Figure
i Ra co han
11 illustrates the grippers.
They hold a sheet either ~~
from an edge or from the
Figure 11. Grippers.
middle in the case of air suc-
tion grippers. Copiers and
digital printers usually transfer paper by friction without grippers.
After printing, sheets progress to a delivery pile commonly on a rotating belt.

22a
 Paper in printing

4 Finishing
Finishing operations convert printed paper into printed matter. In sheet fed offset
printing especially, printed paper is commonly varnished in-line with printing after the
printing units. In varnishing, rollers apply a thin transparent layer of lacquer. Its
purposes are to even the topology of the surface to improve gloss and to protect the
printed layer from smearing and enhance strength and scuff resistance.
The “surface finishing’, what varnishing is, is followed by finishing of the “body” of
printed matter. In a strict sense, finishing refers to the latter operations such as cutting,
folding, binding, and casing. The following discussion briefly covers the underlying
mechanisms and the principles of technical implementation.
The mechanism of cutting involves breaking of bonds within a substrate. In
fibrous materials, a distinction exists between the internal bond of fibers and the
strength of interfiber bonds. Fibers commonly have a higher internal strength than the
bonds. The localized pressure exerted in cutting must exceed the strength of the fibers
to produce a sharply cut edge.
In slitting, the cut occurs in the MD. Die stamping produces pieces with shapes
other than rectangular for use in packaging applications. For perforation, the cut is spa-
tially periodic. This allows folding the pages along the perforation or tearing during later
use because the sheet has weakened sufficiently. Creasing improves foldability of board
and book covers. The operation compresses the material so that a permanent deforma-
tion takes place along a line at which folding subsequently occurs. The thickness of the
material and thus its stiffness decreases along this line.
Folding generates a permanent deformation along the fold. It should not cause
any internal breakage in the structure of the substrate.
By binding, folded pages collected in the proper page numbering order are joined
into one entity called the block. Different types of binding include stapling, adhesive
binding, and thread binding. Gluing of the cover to the back of the block or to the first
pages is casing. The former type of casing finds use in pocket books that have “soft cov-
ers.” The operation can occur in-line with printing. “Hard covered” books are cased off-
line by gluing to the first pages.
A fundamental requirement in adhesive binding is thermodynamic wetting of the
substrate by the adhesive. Penetration of the adhesive into the substrate enhances the
strength of the bond because the surface area of contact increases. Chemical similarity
of the adhesive and the substrate further improves the adhesive strength by promoting
molecular diffusion of the adhesive into the substrate.

5 Runnability
Transportation of the paper through the press and finishing place many requirements on
the runnability of the paper. In addition, the printing nip and drying operations have their
own requirements. Printing press handling and printing exert different types of stresses
including mechanical, thermal, electrical, and water related. “Good” runnability means
high production efficiency. A measure that can be used in a given press set-up is the
number of copies of acceptable quality produced in unit time. Runnability is often

223
CHAPTER 8

considered as reverse factors of efficiency — runnability problems. This implies that

runnability is not important in printing except in problem situations. From a paper point
of view, the definition of runnability is the following:

strain caused by the process > resistance to strain

This definition uses the term “strain” in general sense.

In printing and finishing processes, paper is subject to the strain indicated in

Table 2.
Table 2. Strains exerted on paper in printing.

Mechanical strain Web tension

Bending
Gripper and friction pull
Nip pressure
Ink splitting tension
Physico-chemical strain Drying heat
Moistening water
UV, EB radiation
Static electricity
Chemical strain Ink solvents
Water additives in dampening

As the table shows, paper

> experiences strain in its x, y,
= and z directions with x and y
rua}
= |
the coordinates in the plane
=e
FZ of the paper and z indicating
Fa
=) the thickness direction. The
or Electro-
Le photography strains often produce
O
> changes in the properties of
kK

oq the paper that then cause

=e various problems. Figure 12
O
= illustrates qualitatively how
& |\ {revo)
oO critical different printing meth-
COLORANT VISCOSITY ods are for runnability. Run-
nability is clearly the result of
Figure 12. Criticality of runnability in different printing methods. interactions between the
paper and the process.

Runnability problems have three categories:

1. A sudden disturbance

can cause a break in production. The disturbance is discrete by nature

such as a web break or a feed or delivery problem.

224
 Paper in printing

2. Acumulative disturbance

can lead to the need to interrupt the production at certain intervals. Dusting
and linting are typical examples of disturbances caused by accumulation.
3. A persistent disturbance

can cause a slowing of production or a deterioration of print quality.

Because dynamic strains on paper depend on speed, all disturbances generally
decrease with a decrease in speed. The control of the process is generally better at low
speeds also. One reason for this is that the time constants of these events are inversely
proportional to the speed — they are longer at low speeds. Typical speed-dependent dis-
turbances include register errors, folding, blistering, drying problems, and paper stack-
ing problems.

5.1 Web breaks

For paper transportation through the press in web fed printing, the paper must withstand
the web tension without web breaks and maintain its stability in the cross direction. Low
friction deteriorates stability.
The average web tension on a printing press depends on the printing method and
type of paper. It is approximatel 100 — 700 N/m. This is considerably less than the web
tension in rewinding (aproximately 1 000 N/m) or the laboratory-tested tensile strength
of paper (< approximately 2 000 N/m). Web tension is necessary in web printing to take
the paper through the printing press in a controlled manner. Note that the growing use
of color printing has resulted in increasing web tension. The typical web break fre-
quency is from 0.2 to 1 — 3 breaks per 100 rolls. This corresponds to approximately one
break per 200-104 km of paper. (This estimate may be low.) This indicates that web
breaks are rare. They are therefore difficult to predict by laboratory or pilot tests. With
improved control over the printing process, the web break frequency has constantly
shown a decreasing trend. A critical moment with high risk of a web break is the roll
change because it usually results in increased web tension.
The following is a model of web breaks. A web break occurs if the local tension of
the web exceeds the strength of the paper web as the following expression indicates:

pe j)Jeanesat las (4)

0 S

where P_ is the probability of a web break

t tension
Ss strength
p,(t) the probability distribution of web tension
p;(s) the strength distribution determining the web break.

220
CHAPTER 8

Figure 13 illustrates
the expression. In the + Web tension Strength
expression, the integral rela- distribution distribution

tive to the web tension from

a certain strength level to
infinity expresses mathemat- pi(t) p.(S)
ically the condition where
tension is greater than Ze s,| ‘
; Or-
strength. The integral across =
the strength sums all the x
cases of the probability that = [\e p.(s) when t= s
meet this condition. op) i >
Because a web break = a
is a discrete event, the prob- bs
ability, P, is typically Poisson- Jpt) p.(s) at
distributed’.
.
STRENGTH (s) AND TENSION (1)
Figure 13. Conditions for a web break to occur.

P(n) =e n'/m (5)

where P(n) is the probability that there will be n breaks in a population of the
size m (number of rolls)
n the average number of web breaks in a population of the size m
(1 = mp where p is the unit probability of a web break).

The strength distribution determining formation of a web break is unknown. The

strength of a wide web is not necessarily the same as the strength of a narrow web
although laboratory measured strength from small strips is the same.
The mechanism of web breaks follows three steps. A finite time is always neces-
sary for a web to break. In the first step, a hole occurs in the web or the hole already
exists in the web. This is initiation. Figure 14 shows that a hole does not support ten-
sion. The tension must distribute to the area surrounding the hole. As a result, the level
of web tension rises in the vicinity of the hole. If the surroundings cannot carry the addi-
tional load, the rupture begins to widen in the crack propagation step. If the line propa-
gates over the entire width of the web, the web breaks for the termination step.
Several paper properties* have a relationship to web break frequency. The most
significant properties depend on the level of the properties. For example, if the tensile
strength is sufficiently high, further increases in tensile strength do not have any influ-
ence on web breaks. Tearing strength or fracture toughness — tensile strength of a strip
of paper that has a tear — may correlate with the number of web breaks. The rheological
properties of paper also influence web breaks. A low elastic modulus and viscoelastic

226
 Paper in printing

behavior are favorable. They cause the neighborhood over which the load around a hole
is distributed to increase. The increase in load per unit length remains lower, and the
risk of rupture propagation is smaller.
Testing of the web
break propensity of paper on Increase in
the laboratory scale is diffi- web tension
4 around hole
cult because breaks are
such rare events. If the con-
ditions of testing are far
eee eee
more severe than in practical
printing, the critical step may
shift and the result be mis-
leading. The situation is
SS
analogous to linting which is
Hole
exemplified in Fig. 17. PROFILE
TENSION

5.2 Sheet feeding problems <+——_—_ __ Web width SSS

Feed and delivery problems _ Figure 14. Mechanism of web breaks.
are situations where the
path of the paper through a printer unit, printing press, or finishing machine experiences
an obstruction. The following are reasons for such disturbances:
Dimensional changes

- Geometric reasons: the deviation of the sheets from the plane in excess of the
tolerances allowed by the printing press, defects in the paper, and defects due
to climatic effects such as curling

- Structural reasons: insufficient stiffness to allow the sheet to pass as an object

supported under a limited number of points

- Affinity between surfaces: the tendency of sheets to adhere to each other or to

the surfaces of the printing or finishing device as a result of friction or static
electricity.
Curling occurs if the sheet deviates from the plane because the area of one side
of the sheet is larger than that of the other. One specifies the direction of curling accord-
ing to the axis along which the curling occurs (MD, CD, or diagonal) and its direction (top
side or under side as Fig. 15 shows. The curling radius indicates the amount of curling.

Curling Cockling

va
—_—» ——>
/ on.

Figure 15. Curling and cockling.

eel
CHAPTER 8

The basic reason for curling is anisotropy of the sheet. The following items can
cause this:

- fiber orientation and differences in orientation between different parts of the

paper at the fiber level because fibers swell more in thickness than in the
length direction resulting in differential strains on the two sides
- variations in fines retention between different parts of the sheet

- one-side surface treatment.

Figure 16 illustrates
curling due to fiber orienta- Top side Bottom side
tion. This assumes the fiber
orientation is more severe
on the under side (wire side,
bottom side) of the web and
that the paper sheet is level
when the relative humidity is
at a certain level such as
50%. At lower humidity, the
j RH < 50 %
dimensions of the top side of ES Top side
the sheet are larger than Bottom side
those of the wire side caus-
ing the paper to curl toward 9%,
the wire side. This is due to se ail Top side
the CD shrinkage of fibers Pollometde
(the shrinkage is also Figure 16. Curling.
greater in CD than in MD
because the axis of the fibers is predominantly in the machine direction). At higher
humidity, fibers swell in CD causing the wire-side dimensions to increase. Because of
the fiber orientation, dimensional changes are greater in the CD than in the MD of the
sheet.
Dimensional stability or the sensitivity of sheet dimensions to humidity is a paper
property that influences sheet feeding and color register. Dimensional changes of paper
also cause problems in finishing especially where two different types of paper are
printed in the same press and finished together.
The friction between two surfaces? depends on the following phenomena:
- real contact area between sliding surfaces

- strength of bonds between the surfaces at the points of contact

- way in which the material on the contact surfaces and in their vicinity under-
goes shearing and breaking during the sliding action.

No general information is available on the relative importance of the three phenom-

ena for paper. Because paper surfaces are rough, the area of contact depends on the load.
The interfacial force between the paper surfaces are van der Waals dispersion forces.

228
 Paper in printing

The electric charges? accumulating on the surfaces are static electricity. With a
rising electric voltage to the level of approximately 3 kV, an accumulation of dust first
occurs. Then problems occur with the sheet movement. When the voltage level reaches
12-13 kV, sheet feeding undergoes obstruction.
Several mechanisms can lead to electric charges. When two objects touch, an
exchange of electrons or other charged particles occurs at the point of contact. If the
objects are insulated, the charges do not leak away. The contact between objects may be
due to pressing or sliding. The temperature gradient between the objects causes transfer
of electrons from the hotter surface to the colder one. Rubbing of material together gener-
ally results in triboelectric charging when the surfaces differ in electrical poential.
The electrical behavior of paper like other materials depends on its capacitive
and resistive properties. The product of capacitance, C, and resistance, R, is the time
constant of electronic phenomena. It describes the time scale at which the electric
charge discharges from the surface of the paper. Larger time constants mean slower
discharge. Capacitance controls the accumulation of electric charges. The dielectric
constant is the geometrically scaled capacitance:

e=C/t (6)

where t_ is thickness
€ dielectric constant.
The resistance consists of surface resistance, R, and volume resistance, R,. The
former refers to resistivity in the plane of the paper and the latter to resistivity in the z
direction.
The accumulation of static electricity has a number of harmful effects including
running problems, adhesion of sheets to each other, and register problems. There may
be positive and negative charges on the surface of the paper occurring side by side.
Since dust consists of electrically charged particles, the accumulation of electricity
usually accompanies the accumulation of dust. If the electric charge of the ink and the
paper has the same polarity, impaired ink transfer and misting of ink may occur.
To reduce these harmful effects, ionization of the air or metal conductors are used.

5.3 Linting
Linting®® is the phenomenon involving separation of fiber or pigment particles from the
surface of the paper and accumulation of these particles on the printing plate and in ink
and water systems. Linting is a problem because it requires regular cleaning of the
printing press. This results in lost production. Linting also impairs the quality of the
printed image. Linting is a problem particularly with uncoated, unsized, mechanical
offset papers (newsprint and SC offset). Problems related to linting and dusting also
occur in electrophotography and flexo printing. In the former, dust is attracted to the
photoconductor by charges, and hot nip fusing attracts the hot paper surface on the
fusing roll. In flexo printing when the ink loses solvent by evaporation on the printing
plate, the surface of the paper undergoes a splitting resistance that may exceed the
surface strength of the paper. In offset printing, this problem of linting is particularly

229
CHAPTER 8

acute when printing medium screen tones. Their lint accumulation capacity is large
because of maximized perimeter between the nonprinting and printing dots. Lint
accumulates most at the interface between nonprinting and printing areas.
Linting results from three mechanisms:
(i) Loose dust on the surface of the paper such as slitting dust transfers from the-
surface and adheres on the rubber blanket of the offset press in the first print-
ing unit — the phenomenon is also called dusting.

(ii) Because of its splitting resistance, the ink pulls out individual fibers and fiber
bundles from the surface of the paper that accumulate along the edges of
printing areas (halftone dots, characters). (The effect is greatest in the first
printing units.)

(iii) The dampening water results in weaker fiber bonds and breaking of fibers’.
(The effect is more pronounced towards the end of the printing sequence.).

A first-degree kinetic model can describe the accumulation of lint on the rubber
blanket. Solution of this gives the accumulation curve the form of a saturating exponen-
tial curve. The accumulating amount of lint is m, and the loosening from the paper and
the accumulation per unit of time is ky. The coefficient of transfer further from the rubber
blanket to the ink and water feeding systems is k,. The first-degree model of accumula-
tion is as follows:

dm
ai = ky—k,m (7)

By solving and setting as the initial condition =0 and m = 0, the following

results:

m= Kory Sie) (8)

ky

The amount of lint that has accumulated increases towards a limit value predicted
by the ratio between the parameters describing the loosening dust, kg, and the lint
transferred from the blanket, k,.
The factors contributing to the linting problem all depend on
- the splitting resistance of the ink such as nip geometry and web feed angle
from the printing nip

- strength of the paper surface in the printing press as influenced by the volume
and type of dampening water
- adherence of lint to the rubber blanket that depends on adhesion, volume of
dampening water, and its alcohol concentration.

230
 Paper in printing

Because of its cumu-

lative nature, linting is diffi- F
cult to measure under
laboratory conditions. By Production Laboratory
testing
subjecting the surface of the
paper to a significantly
greater force than is the Security margin
case in printing, measurable >
DISTRIBUTION
amounts of dust are
released. This distorts the >
dusting phenomenon as STRESS
Fig. 17 shows and the mech- __ Figure 17. Testing of linting in laboratory conditions and the strain
anisms may be different from caused in production situations.
those in production printing.
When considering runnability, printing nips require that the paper surface endure
the forces that arise as the ink films split in transfer to the paper even under the influ-
ence of dampening water in offset printing. In addition, the dimensional stability tested
under the influence of water and web tension must be sufficiently high to allow color reg-
istering in MD and CD.

5.4 Drying
The runnability requirements during the drying step of printing differ according to the
type of drying. In drying by evaporation of a liquid phase from the ink, the paper surface
should withstand the heat and the inevitable loss in moisture without deterioration of its
properties. Because the interfiber bonds in paper are hydrogen bonds, evaporation of
moisture from the paper may lead to bond breaks. This will cause surface roughening.
The phenomenon is fiber rise or fiber puff. In densely coated papers, the coating
obstructs water evaporation from the paper. As a result, water may exit forcefully and
cause abrupt local breakage of the surface. This is blistering. In drying by oxidation and
polymerization, the paper should not interact chemically with drying and impede the
process. Radiation should also not cause deterioration in paper properties such as
debonding or optical changes.
Loss of strength of paper or embrittlement in drying may cause partial breaking of
the structure when folded. This is cracking. Wood containing papers are more apt to
crack than wood free papers because of the higher flexibility of chemical fibers than
mechanical fibers. If the paper retains ink solvent, this may cause smearing in finishing
and mutual adhesion of the sheets in storage before finishing. Lowering printing speed
allows more drying time at the expense of production efficiency.

6 Printability
6.1 Principles
Printability of paper is the combination of paper related factors that contribute to the
achievement of a desired quality level. A corresponding definition can exist for printing

231
CHAPTER 8

ink, but the term printability does not have wide use in the context of ink. Printability of
paper is equivalent to the quality potential of paper in printing The definition makes it
evident that printability is not the same as printed quality because other factors
influence printed quality. These include pre-press, printing press, and printing factors.
Figure 18 illustrates schematically how different printing methods compare with the
overall criticality of printability.
In choosing a paper
grade for a particular printed A
product, printability defines
certain criteria. In general
importance, it ranks second
to runnability. A paper grade
has little chance of exhibiting
its good printability proper-
ties if it does not run in the
process. Electro-
photography
Printability parame- OF
CRITICALITY
PRINTABILITY
ters are measured as opti- >
COLORANT VISCOSITY
cal, colorimetric, and
mechanical print properties. —_Figure 18. Criticality of printability.
This means that paper prop-
erties influence printability but are not printability properties themselves. Because print-
ing is definitely an interactive process, printability is the result of the following
interactions:

- paper and the printing ink?

- paper and the printing press

- light and print in visualization of the print for viewing or measurements.

The reason one cannot predict printability by paper properties is that several
mechanisms typically govern the interactions. Without printing, it is almost impossible to
arrange measuring situations where the paper properties would similarly influence the
result as they do in printing unless the test situation were some type of printing. A con-
sequence of the interactive nature of printability is that measured values also depend on
the other components, the ink, and the press. Printability values have unambiguous
meaning only provided the test conditions are specified in detail.
Sections in Chapters 5 and 6 discussed interactions between paper and ink and
paper and press. Chapter 7 focused on the principles of light and print interactions.
Table 3 lists the most commonly used printability factors grouped according to
whether they are determined from solid areas or halftone areas in a measuring situa-
tion. All the factors related to a solid area also have significance for halftone areas, but
the reverse is not true.

232
 Paper in printing

Table 3. Printability factors.

Printability factor Definition

Solid area:
Ink requirement Quantity of ink required to reach a given level of density
Color range Range of color that can be reproduced
Print through Optical density of print measured from the reverse side
Gloss Relative reflectance in the specular angle
Gloss contrast Gloss difference between paper and ink
Trapping Transfer of inks in overprinting
Unevenness Uncontrolled variation (=noise) of density, gloss, and color
Fiber puff Nonuniformity of printed surface
Solvent retention in print Quantity or proportion of solvent retained in printed layer after drying
Touch-proofness, set-off Smearing of print on a contacting surface under pressure
Rub-proofness, rub-off Smearing of ink on a contacting surface under shear
Halftone areas:
Tone reproduction curve Relation between halftone percentage and print density
Halftone contrast Density difference between solid area and a dark halftone area
Unevenness Noise in halftone areas
Missing dots Number of unprinted halftone cells/unit area (gravure)
Halftone dot properties Uncontrolled variation of haiftone dot size and shape, sharpness of edge,
raggedness of edges

6.2 Printability of solid prints

Ink requirement
Ink requirement — the amount of ink needed for a given level of optical density on the
paper or the level of density achieved with a given amount of ink — is the most basic and
the most important printability factor. A small number is advantageous. Penetration of
ink into the paper? and variation in the ink layer thickness reduce the optical efficiency of
a given amount of ink on the paper. The reduction of optical efficiency also causes a
smaller color range.
The influence of penetration is due to the fact that the ink that has penetrated into
the paper becomes figuratively part of the light scattering paper. Light scattering directs
light intensity away from the ink and reduces the light absorption. The variation of ink
film thickness reduces density as noted in an earlier chapter. Uneven contact between
the ink layer and the paper surface is a reason for variation of ink film thickness. Due to
the nip conditions, gravure printing is far more demanding than offset or flexographic
printing on the smoothness and compressibility of the paper surface.

233
CHAPTER 8

Laboratory conditions allow determination of ink requirement curves by combin-

ing gravimetric and densitometric measurements. Ink mileage is the corresponding term
used in production conditions for comparative purposes.

Print through
Even if no ink penetrated into the paper, the print would be visible to some degree from
the reverse side provided paper opacity were not 100%. Print through depicts the
magnitude of visibility in density units. The contribution of the transparency of paper to
print through is show through. The contribution of the ink consists of that arising from
pigment penetration and that from oil penetration. The former acts by reducing the
effective thickness of paper and the latter by reducing its light scattering.
Print through is a deterministic noise source on the reverse side. The spatial dis-
tribution corresponds to the distribution of printed area of the sheet. Print through is
measured as density on the reverse side. Values above 0.1 density units are unaccept-
able from a subjective viewpoint.
The term strike through describes a situation where some ink has penetrated
through the paper. It may be measured as the number or coverage of points that are vis-
ible at the reverse side.

Gloss

Surface gloss as explained in Section 7.2 influences print density and color. High levels
of density and a wide color range are not possible without a glossy printed surface.
Gloss is also an epitome of high quality although it is not preferred in all situations.
Gloss is a measure of surface smoothness. It equates to the proportion of small-
scale, smooth surface area, A, ina sample”. Roughness and the corresponding
smoothness have origin in more than one source. The sources such as base paper
(“coarse”) and coating structures (“fine”) have different characteristic scales and mean
depth amplitudes. If the sources of lack of smoothness are statistically independent, the
proportion of smooth surface is a multiplicative combination of the proportions of
smooth areas (Aj, Ao, ...) of the different scales — macro scale and micro scale. This
means that gloss, G, is as follows:

G = A,A, (9)

The expression suggests that small scale roughness is sufficient to disrupt gloss.
Matte papers purposely use this.
Achievement of high printed gloss requires that the paper be smooth or that the
ink layer can fill rough profile of the paper and produce a smooth top surface. In the
absence of ink penetration into the paper, this is possible if the roughness depth is
smaller than the ink layer thickness. Figure 19 illustrates different cases.

234
 Paper in printing

+ um Ink film
thickness

1 Limited by gloss of
the ink layer

ea eee | 2. High gloss contrast

0 ee ee ee —ee ee eee | . .
ONG RG GEA ENN ee | 3 | Fairly high gloss contrast

aN (ON
LY, 4 | Lagerly devoid of gloss

eae Ni 5 | Nonglossy

Figure 19. Illustration of filling of paper roughness by ink.

The difference in gloss between print, G,,,n;, and paper, G,,,)¢,, 1s the gloss contrast:

Gloss contrast AG = G,,ing— G Gpaper (10)

Highest values of gloss contrast occur where the paper is not glossy but the
roughness is small scale. Matte coated papers fall in this category.
Gloss measuring devices are glossmeters. They allow measurements of specular
gloss at more than one combination of fixed angles. The angle of 75° relative to the sur-
face normal is common for papers and prints. A smaller angle (45°) is practical for very
glossy printed surfaces, and a larger angle (85°) is practical for matte surfaces. The
scaled value is called sheen. Note that specular gloss is not the only property that is
measurable for characterization of surface reflection behavior. For paper and prints, the
others are secondary.

Trapping
Section 4.2 in Chapter 5 discussed the definition and significance of trapping from the
viewpoint of color.
In multi color offset printing, ink layers are transferred on top of underprinted ink
layers that are not dried. This overprinting situation is the most demanding of all meth-
ods. The paper should allow the setting of underprinted ink layers so they are in a semi-

235
CHAPTER 8

solid state at the moment a new ink layer contacts the surface. Simultaneously, paper
should imbibe the dampening solution so that its presence will not hinder ink transfer.
Setting in offset occurs by separation of oil from the ink using the capillary suction
of the paper. This requires that the paper has a sufficiently large pore volume and a sat-
isfactorily high capillary suction that is best generated by a fine pore structure. Small
sized pores fulfill the latter requirement. The surface chemical properties that influence
absorption must allow water and oil absorption at approximately the same rate. This
requirement can be met with the knowledge that emulsified dampening water in the ink
and oil amounts are similar.
Trapping is commonly measured densitometrically. The measurement is made
from two areas. One represents an overprint of the color pair of interest (D;¢>), and the
other represents single color areas. The filter is that normally used to measure density
of the second ink. Calculation of trapping percentage uses the following:

Trap% = ———x 100 (11)

Trapping is ideal when the percentage is 100.

Uneveness

For prints, the terms “noise” and “uneveness’” are interchangeable. If not specified in
more detail, noise refers to noisy print density or diffuse reflection. Gloss noise refers to
variation of specular reflection and color noise to variation of colorimetric reflection.
As Section 6 in Chapter 7 discussed, a given noise source gives noise with distin-
guishable characteristics. Mottling (density or gloss) refers to low contrast noise on the
scale of approximately 1-10 mm, i.e., a frequency band of 0.1...1 mm”'. The term
mottle is used to mean noise caused by paper formation; the relationship is not neces-
sarily straightforward’? '2. Sometimes mottling is a general term for noise’?.Compared
with mottling, granularity occurs in a higher frequency band and has higher contrast.
The visual impression of granularity is graininess. High contrast also characterizes
speckle. In speckled prints, one can distinguish spots not covered by ink. The term back
trap mottle indicates that overprinting is the source of mottle noise.
Types of noise that
have distinct frequency A
bands are halftone noise
and wire mark and felt mark
noise. Such noise also has ov o —
distinct textural characteris- hoo
tics’. In halftoning using a
fixed lattice, prints are noisy
at that frequency (1/102 um” SS
') Wire marking noise in FREQUENCY, u
prints originates in the wire- Figure 20. Power spectrum of noise.

236
 Paper in printing

mark of the paper and occurs at the same frequency (1/10° um’!).
Figures 13 and 14 of Chapter 7 illustrated noise as a function of frequency. A for-
mal representation involves plotting the power spectrum of noise on the vertical axis.
The power spectrum expresses the distribution of signal variance over frequency as
Fig. 20 shows. The integral of the power spectrum is the variance. Distinct noise bands
occur as peaks in the power spectrum”. The curve is commonly continuous without
any peaks as a result of several contributing factors.
Calculations of uneveness commonly use two-dimensional digital images of
prints. Imaging is accomplished through microscopes with computer linked video cam-
eras by systems called image analysis systems.

Drying properties
Fiber puff is an earlier mentioned drying defect. It involves roughening of the paper and
print surface with eventual influence on gloss and gloss noise. The surface roughness
texture generated by fiber rise may also be visible. The type of noise is surface
nonuniformity.
In printing where ink drying is by solvent evaporation, the solvent must have effi-
cient removal from the layer for hygienic work and environmental reasons. A low
retained solvent concentration also favorably influences the surface intergrity of the print
measured as touch-proofness and rub-proofness. The relationship is not functional. Sol-
vent retention depends on many factors of which the solvent retention tendency of the
paper is only one. Others include the drying conditions and the inks.
Chromatographic techniques?? are available to determine residual solvent contents.

6.3 Printability of halftones

Halftones generate different levels of darkness and color in printing processes that are
capable of only bi-level output, i.e., one bit/pixel with or without ink. Such processes
include lithography, flexography, area modulated gravure and bi-level
electrophotography, and many ink-jet and thermal ink transfer methods. In printing with
multi level output, i.e., several bits/pixel, the relative coverage of the paper or halftone
percentage is often ultimately what is visible in the print. This is true even if initially the
modulated variable may be the amount of the coloring material.
The printability of halftones consists of two issues: tone reproduction and dot
properties. The former is a macro scale and the latter a micro scale feature. The critical
aspects of the tone reproduction curve are the following:
- dot gain and filling in of medium and dark halftone areas as in Fig. 7 in
Chapter 7
- nonprinting of small halftone dots.
Because filling-in is a consequence of dot gain, both aspects can be character-
ized jointly using halftone contrast. Halftone contrast, K, can be absolute contrast

Kays
a = Pioo- P75 (12)

237
CHAPTER 8

or relative contrast

D099 — P75
Koa = D
100

where Dyop is density of a solid area

D745 density of a 75% area.

Higher halftone con-

trast gives wider tonal area A
available for storing informa-
tion in dark tones of the
image. Instead of the 75% Controlled by
dot area in the equation dot gain
above, any dark tone per-
centage can be the refer- Controlled by
ence. Dot gain causes the halftone
principle
halftone contrast to CONTRAST
HALFTONE
decrease. With an increase
in the amount of ink in the SOLID DENSITY, D495
print, halftone contrast first
Figure 21. Halftone contrast vs. solid density.
increases as Fig. 21 shows
consistent with the Murray-Davies model. More ink signifies more dot gain anda
decrease in halftone contrast. The level of solid density that gives maximum halftone
contrast — normal color intensity (NCI level) ’©:’” — represents optimum density for spe-
cific paper and printing conditions.
Paper properties with an influence on dot gain include surface topography, poros-
ity, and light diffusion properties. The influence of porosity relates to the random posi-
tioning of pores at the surface. Ink is also pulled sideways. Topography acts as a
constraining factor.
Halftone contrast concerns the halftone densities averaged over a statistically
representative number of dots.
The “halftone print properties” defined in Table 3 concern dot characteristics
within dots. Figure 22 illustrates some characteristics.

C) C) | Sharp dot

DENSITY
Unsharp |
dot

X X X X

Figure 22. Illustration of (from the left) missing dots, raggedness of edge, edge noise, and lack of sharpness.

238
 Paper in printing

Dot size and shape can specify ideal dots. Real imaged dots can differ from the
ideal in many ways. A total absence of dots is the printability factor of missing dots com-
mon in gravure. Dots may also be partially missing. The shape of the dots may be sys-
tematically altered or dot distorted. The reason may lie in the directionality of substrate
properties or nonideal rolling in nips. Irregularity of the position of the dot edge is rag-
gedness. It occurs in all printing methods, but it is more common on rough papers. Edge
noise refers to scattering of particles around the edge and is a characteristic of particle
processes such as electrophotography. In ink-jet, splatter and satellite drops also show
up as particle scatter. Dot sharpness is of most concern in high quality printing where
the dots come closest to the ideal shape. It is associated with the slope of the edge. The
edges of ideal dots are infinitely steeply sloped.
Nonideal dot properties are a source of noise on their respective scale. Quantifi-
cation uses image analysis techniques.

7 Paper as carrier of information

7.1 General
Paper serves as a carrier of optical information in the communication process. In the
early days of computer technology, paper was the medium for saving information as bit-
size holes on punch tape and cards. Since then, there have been attempts to use paper
as an optically readable memory in computers. This idea uses the low cost of paper.
Compared with magnetic and optical memory devices, the information storage capacity
is at least one order of magnitude smaller. In bar code and other machine reading
applications, paper serves as a partly machine-readable memory medium. In the so-
called glyph technique, machine-readable information is appended to visually readable
text and image information. The glyph writing system forms a field with a uniform tone.
The information it contains is not visible provided the glyph pattern is at a higher
frequency than what the human eye can distinguish. Despite the examples the most
common use of paper is as a carrier for visual information.
One can define the concept of information in various ways. Among the different
definitions, the spatial (or temporal for time-divisible signals) bit information represents a
technological approach to information according to Shannon’s communication theory ’8.
Spatial information is bit information at the x, y level. Visual spatial information refers to
the capability of the human visual system to distinguish tone and color levels. The tones
are separated on the basis of luminance and the color levels with the aid of colorimetri-
cally filtered spectral information. The concept of spatial self-information comes from the
probability-based definition of information: Information (intrinsic information) is the neg-
ative base two logarithm, —log> p;, of the probability, p;, of an event level.
For spatial information, events include darkness and color levels in a certain
image. The information capacity is the maximized value of information calculated over
all events.

239
 CHAPTER 8

[n= aslSrtoesr (14)

|

The value is maximized in relation to the probability distribution, p;. The maximum
value occurs when all events are equally probable, i.e., the distribution is even. If the
number of possible tone and color levels is n, the probability of a certain level occurring
in the case of equal probability is n'! and the information capacity is as follows:

I(bit) = logyn (15)

The information capacity is the base two logarithm of separable darkness or color
levels. The following section examines the approach to calculating the number of dark-
ness and color levels.
Another basic concept of statistical information theory is the mutual information of
two sets of events such as input and output of a process and the transfer of information
averaged over all event levels. Mutual information is a measure of how much informa-
tion transfers via some channel or process from the event, 7, to the observation,7.The
formula for calculating it is as follows:

I = log, oe (16)
J

The formula incorporates the princi-

ple by which the amount of mutual informa-
tion increases as the probability grows that i Channel/ J
the event, 7, when transferred through a Process
channel or a process produces event,j,on
_ the output side. Figure 23 shows a sche-
Figure 23. The input event levels of the channel or
matic diagram. The mutual information process denoted by i and the output events by j.
grows when the j value is obtained system-
atically as a result of the i value and not arbitrarily as a result of disturbances. Printing is
an example of a channel or process.
The average of mutual information calculated over all event levels is the following:

(17)

The value of information transfer maximized relative to the distribution, p;, of input
event levels, 7, determines the capacity of the information channel. The channel capacity
can as a maximum equal the information capacity of the output set of events. This
means that the channel cannot pass more information than what the output can repeat.

240
 Paper in printing

A discussion that examines paper as a carrier of information can use the con-
cepts of spatial information defined above. Such evaluations can be applied

- for comparing different modes of communication such as television or com-

puter-based communication and human vision with printed communication

- for optimizing the parameters in printing’ ’9-°7 such as screen structure and
screen ruling.

From the viewpoint of information, the levels of the various parameters are opti-
mal when the information transferred is at its maximum. In addition, the information
capacity is a quality factor that has use as a criterion for quality development 22°.
7.2 Information capacity
The microscale capacity of single-color information describes how many tones
expressed in bits as the base two logarithm of the number of tones the paper or printed
image can reproduce. Calculation of the information capacity can have reflectance units
or density units. The latter relate more closely with visual spatial information. For
reflectance units, the following expression gives the information capacity:

2 R (u) — Ry in)
1G)
R eS vets82 aG (ul) (18)

For density units, the expression becomes

—log,,R._. (u) — (-log ae (u“))

Ip(u) = w log 1-¢ ea
ae amine }
Op(u)

where u is spatial frequency

R relative reflectance
—log oR density
s(u) noise power at frequency u.

The expression within the outer brackets of this equation represents the signal-to-
noise ratio:
SNRp oa (Rinax = Rmin/OR
SNRp = (Dia Di! OD
Dmax = 10810 nin
D nin = —log oR max

where D Max and D,,,,, are the printable minimum and maximum densities
R min and R,,in their corresponding reflectances.

241
CHAPTER 8

Using a common con-

vention, SVR + 1 indicates the ie Signal
number of distinguishable
tone levels as Fig. 24 shows.
In the figure, the signal
decreases from 10 to zero as
the frequency grows. Multiply- SNR = 10
ing the logarithmic expres-
sion with uv? scales the bits per
unit of area.
In the example shown SPATIAL FREQUENCY, u
in the figure, the information
Figure 24. Example of signal-to-noise ratio with the noise assumed
Capacity decreases as a independent of the spatial frequency and equal to 1 unit.
function of spatial frequency
as Fig. 25 shows. This is true
of most cases.
Determination of the
relation of noise (07) {>
-
expressed in density units or
reflectance units uses the
variance conversion formula:
2
op = (53) OR (20) CAPACITY
INFORMATION
Inserting the differen-
tial gives the following: SPATIAL FREQUENCY
Figure 25. Typical frequency dependence of information capacity.

2
os ee 3; (21)
The examination of information capacity in density units emphasises the informa-
tion in the middle area of the reflectance range while dampening the importance of
information with low and high reflectance. This means one must be aware whether a
measurement will be in reflectance or density units as a minimum requirement. Image
analysis often disregards the density conversion and thereby measures reflectances.
The previous expressions of micro scale information capacity can be converted
into expressions of macro scale information capacity. The term macro scale means zero
spatial frequency or the scale obtained by integrating over all frequencies.

I = 1 Dae min
piu) = log, Soar es (22)

242
 Paper in printing

7.3 Optical information capacity of paper

Inserting in the expressions magnitudes derived from paper properties allows calculation of
a limit of the capacity of paper to carry information. The following text discusses these items.
Dynamic area
When examining the optical information capacity of paper, R,,,,,,(u) and R,,,;,(u) are the
maximum and minimum reflectance as determined by the optical properties of the
paper. The difference R,,,,,(u) — Rypin(u) or the difference between their logarithmitic
values is the dynamic range. When printing, information can be stored in the reflectance
or density area delimited by the dynamic range.
At macro level, reflectance representing the brightness value of paper determines
Rmax- A single-color printed image cannot be brighter than that excluding the possibility
that the printing ink would be white or fluorescent. Surface reflectance distribution con-
trols the minimum reflectance R,,,;,. This thinking uses the assumption that the only
reflection from the surface is the surface reflection if there is a hypothetical layer of ink
on the paper that totally absorbs light. Figure 26 shows this. If the hypothetical ink layer
is ideal — infinitely thin, the surface structure of the paper determines the surface reflec-
tance behavior. If the reflectance is diffuse, the surface produces a reflectance of 4%.
Thus R,,,;, = 0.04 using a measuring geometry of 45/0°.

Illumination Detection IIlumination Detection

Surface reflection
Surface reflection Sis é
indicatrix
indicatrix

min

Paper Hypothetical print on paper

Figure 26. Reflectance R,,,;,, restricted by surface reflectance. Left: paper surface, right: print (ideal) on paper.

With a greater paper gloss, a smaller proportion of incident light reflects in the
measuring direction (measuring geometry usually 45/0°). This gives a smaller R,,,;,, and
greater information capacity as Fig. 27 shows.

IIlumination Detection Illumination Detection

Surface reflection Surface reflection

indicatrix indicatrix

min min

Paper Hypothetical print on paper

Figure 27. Paper with greater gloss than in Fig. 26, with surface relectance less directed in the 0° direction.

243
CHAPTER 8 |

From the above, it is

obvious that the information Information
capacity
capacity of paper depends "Glossy" surface
on illumination and viewing
angles except for matte
paper. This also applies to "Matte" surface
displays. With a more
directed surface reflectance, 45
the information capacity ANGLE, °
depends more on illumina-
tion and viewing angles. This _ Figure 28. Dependence of information capacity on viewing angle
shows a compromise with illumination angle 45°.
between the angle stability
of information capacity and the magnitude of the capacity as Fig. 28 shows.
The relative reflectance area, 0-1, in a digital presentation represents reflectance
levels of the number 2" where n is the number of bits used per pixel — typically 8-12.
When the brightness of the paper in relative reflectance units is R,,,,, which is < 1 and
the minimum reflectance R,,,;,, is > 0, reproduction of reflectance levels is possible in the
range Ray Rin, |-€., the number of reflectance levels is the following:
n
he Cee ne) 2 (23)

from which

= log s[(Rinax Ars Ree) = logo [(Rnax 7 Ria) +n (24)

because th) ik. <= I:

the information capacity is
2" levels
smaller as logs (Rmay —
Rin) iS a Negative
value, than in the ideal case a
where it is n. Using realistic
reflectance values, one can R max Ae
conclude that the reducing
effect is at its biggest — the Figure 29. Reflectance range as bits.
brightness level of
newsprint — approximately 1 bit. Half the reflectance levels are lost due to the nonideal
brightness level of the paper.

Optical dot gain of paper

When examining the information capacity as a function of frequency, the dot gain
means that the density and reflectance area reproduced by the paper decrease with a
decrease in the size of a detail. The dot spread function — two-dimensional
presentation — or line spread function — one-dimensional presentation — describes the

244
 Paper in printing

optical dot gain at the x,y level. In the spatial frequency plane, the Fourier transfer of
these — modulation transfer function (MTF) — describes them. Figure 30 shows the
darkness distribution of edges, an ideal image, a spread edge and the slope (derivative)
of the edge. The line spread function is scaled slope.

Ideal edge Real edge Derivative of edge

“x= 0 ~ =O x=0

Figure 30. Darkness distribution of ideal edge on paper and edge spread function.

The single-parameter model of the M7F of the line spread function is as follows:

MTF = 1/[(1+(2nuw)’] (25)

where u_ is frequency
Ww average dot spread or the average diffusion distance of light
in paper.

If the structure of paper were isotropic in all directions (which it is not in reality),
the average diffusion distance could be predicted with the brightness of paper, R,,,,,, the
light scattering coefficient, s, and the specific weight of paper, p, as the following equa-
tion shows:

Scaling with specific weight gives the diffusion distance the dimension um.
Strong light scattering effect reduces the dot spreading. The expression results by
applying the expressions using the Kubelka-Munk theory and assuming light scatters
isotropically in the paper. Because the components in the paper web orient in the plane
of the paper, the assumption is only an approximation.
Because of the dot spreading, the available reflectance range at frequency u is
smaller than the dynamic range of the zero frequency with the factor MTF (MTF = 0-1).
In the spatial plane, the reflection profile caused by dot spreading is created as
the convolution (*) of the idealized reflection term and the spread function (cf. Section 4
in Chapter 2). In one-dimensional form this is

RoulX) = Rin )*A(x) (27)

245
CHAPTER 8

According to the definition, the one-dimensional convolution is the following:

co

Roul®) = J Rie hax, dx, (28)

—co

In the Fourier plane, the convolution is the product R,,,, (u) = Rin (u) MTF.
Placing this into the expression of information capacity gives the following:

(R io —
[= ulog,| |+ (29)
max KF in

The smallest reproducible pixel size results by inserting the above MTF model in
the expression for information capacity and by defining that the frequency at which the
signal-to-noise ratio is one determines the pixel size. The smallest reproducible pixel
size is 1/(2u) where:

l EZ 4 Ro ie
~ Iw Op(u) ee

for which the following approximation results for pixel size:

P ixelisize are
= Ung Tw
Colney:
pA
max
3
min Cy
Smaller optical dot spread and greater brightness and gloss give the smallest
reproducible pixel size and the highest achievable resolution.
The calculation yields values of approximately 10-30 um for the smallest pixel
size. The frequency-dependent information capacity reaches its maximum oe at the
frequency ee ene ae to the smallest reproducible pixel size: ur logy2= u>.This is
because the u? term rather than the logs (1 + SNR) term controls the value of the
expression. A smaller value gives greater information capacity. Details smaller than this
value are indistinguishable on paper. By choosing pixel size for printing smaller than this
value, individual screen dots remain below the visibility level.

Noise

Refering to Fig. 29 when coding information for printing, the step between two tone
levels is as follows:
step: = 1/2"

246
 Paper in printing

This equals 0.0039 — on the reflection scale 0-1 — when n = 8.

For the steps to be distinguishable, the optical noise of reflection from paper
should be smaller than the step. If the small-scale variation of paper expressed as ao
value is greater, the reflectance levels are not separable statistically. The number of real
steps decreases. A smaller variation means printing does not fully use the optical poten-
tial in this sense. Calculation of the number of real steps comes from the previous
expressions as the signal-to-noise ratio, SNR. Note here that SNR = 100/y, when v is
the variation coefficient of the reflectance variation (= op/R). If the variation coefficient is
0.01 or 1% —a realistic value for paper, approximately 100 tones or 6.6 bits can be sep-
arated from each other.

7.4 Color information capacity of paper

Digital coding of color information uses n bits for each component — RGB or Lab. The
number of color levels, m, is then:

m = 2"2"2" i)
As in single-color information, the information capacity of digital coding is
logym = 3n. The color space — color range of digital coding — is a cube. Its information
capacity is the two base logarithm of the volume of the cube shown in Fig. 31. Repro-
duced on paper, the color space is two distorted cones or “pear-shaped”.

AL*

Ideal b*
oo ———
process

Halftone

Vy

Figure 31. Color space of halftone printing and ideal color space assuming the paper is achromatic.

247
CHAPTER 8

Neglecting the influence of noise, the color information capacity is the logs value
of the volume of the color space.
Various factors influence the color range. In the Lab coordinate system, the differ-
ence in luminance (dynamic area relative to luminance) between the luminance of the
paper and the luminance of a solid print influences the volume as Fig. 32 shows.
In the color space, this
corresponds to single-color Paper

Another factor influencing a ponte Full tone and solid

the volume is the base area Sar raaton overprint areas
of the cones. In simple == Sond arene
terms, the saturation of solid = aa and black
0 A ac
print single-color (cyan,
magenta, yellow) and over- CHROMA
print (red, green, blue) areas
determine this. The color Figure 32. Formation of the color space of a printed image in a two-
information is approximately dimensional presentation restricted by chroma, c*, and luminance, L*.

the sum of achromatic infor-

mation and chromatic information. A more detailed examination considers distortions of
the color space that also depend on the paper. Distortions arise because the papers are
not achromatic. This affects the shape of the color space.
Paper influences chromatic infor-
mation optically and by interactions of Surface Internal
paper and printing ink. The optical effect reflection reflection
consists of surface reflectance and the
effects of the paper’s own color. This
assumes that the reflectance caused by
the printed image consists of a surface
reflection component and a reflection Figure 33. Surface reflectance and reflectance from
component emanating from inside the inside the printed image.
printed image. Figure 33 shows this.
The white color of the surface reflectance reduces the chromaticity. If the printing
ink does not fill the roughness volume of the paper, the surface reflectance behavior
depends on the surface reflectance behavior of the paper.
Of the interactions between paper and ink, the following influence the color and
thereby the information capacity:

- ink penetration (effect on the spectral absorption of light and gloss)

- slow ink set-off with resultant poor transfer, i.e., trapping in overprinting of inks

- crystallization of inks containing soluble color substances.

Figure 34 illustrates the influence of trapping.

248
 Paper in printing

b*

AG

magenta

cyan blue: blue; < 100 % trapping of magenta

< 100 % trapping
of magenta

Figure 34. Effect of trapping on color with magenta printed on top of cyan.

Imperfect trapping acts by diminishing the range of reproducible colors.

8 Testing the printing potential of paper

The criteria for testing the printing potential of paper are as follows:

- functional criteria that depict the use of paper as a medium for communication

- criteria related to production technology using the behavior of paper in printing

and finishing processes — runnability and printability

- criteria related to the product and uses.

Figure 35 shows that testing of the printing potential of paper can use cost and
scale terms:

- laboratory-scale testing

- pilot-scale testing

- production-scale testing.

The benefit of a test

does not depend only on the
costs. A large expenditure
Graphic
does not automatically pro- production
duce additional benefits test

because testing require- COST Pilot printing

test
ments vary widely.
Lab. printing F/ End user's
Laboratory-scale End user's
test field test
lab. test
printing tests use controlled
conditions with presses that SCALE
are smaller and simpler than
Figure 35. Cost and scale of printing potential tests.
commercial printing presses.
In pilot-scale printing tests, the printing conditions are controlled and repeatable like the

249
CHAPTER 8

laboratory, but the printing equipment may consist of commercial-size or smaller

machinery with functions corresponding to those of commercial printing equipment.
Because commercial printing equipment has many different units in a functional unit,
the presses are often very individualistic. The results from one environment are there-
fore not necessarily generally applicable.
Electronic printing as descussed in Section 1.1 of Chapter 6 has low end, middle
range, and high end segments. The low end printers are personal devices. Their low
cost allows easy testing corresponding to laboratory printing tests. In the middle range,
the cost level is approximately one-tenth the cost of the high end category. Middle range
equipment is very suitable exclusively for pilot testing. The very high end equipment is
good for printing tests side-by-side with normal production operations.
Table 4 summarizes specific features associated with testing printing papers at
different scales.

|
Table 4. Alternative printing tests.

Laboratory Production
Printing method Discontinuous* Continuous Continuous
Paper use Specimen ** Sheet/web Sheet/web
Test stripping 100% surfaces roto-gra- | Freely selectable Freely selectable or deter-
vure and flexo: mined by printing work
halftone areas
Number of inks/toners | 1—4; 1- or 2-sided 1-4, 1- or 2-sided 1-4-8, 1- or 2-sided |
Drying No drying/hot air/ UV/IR | Various alternatives Standard solutions
Suitability Basic printability*** Printability, quality, basic | Printability, quality, run-
runnability, information nability, information
mation capacity, toler- capacity
ance ranges
Print run per test point Tolerance range refers to
and press width restrict | the sensitive of the print-
testing of runnability phe- | ing result to changes in
nomena printing conditions; toler-
ance ranges vary for dif-
ferent paper grades |
excludes testing of time-dependent and cumulative phenomena
excludes testing of phenomena related to web tension
*** quality properties ofsolid printed areas

Discontinuous printing refers to printing of one sheet or specimen at a time.

As noted above, the production equipment used in the graphics industry consists
of individualistic units. The inks used are often specific to each press. In electronic print-
ers, the toner or ink is an integral part of the device. This is even more so than in ordi-
nary presses. There are also differences in the transfer and path of the paper between
different devices within a segment of electronic printing devices. For this reason, the
possibilities for universally applicable testing are poorer with electronic printers than
with printing presses. Lack of stability may also limit the use of low end printers.
Various methods can test the printing potential depending on different principles
and variables applied in the testing situation. In principle, Table 5 shows that testing of

250
 Paper in printing

printing papers can use five different methods. The most suitable testing method
depends on the objectives of the test and the available equipment. The test can also be
a combination of alternatives.

Table 5. Printing test methods.

]
Testing methods Suitability
—

Printing conditions as variable

Laboratory printing Suitable for obtaining basic information within the framework
of available parameters
Pilot printing Suitable for preliminary review for developing a paper grade, for
verification whether objectives have been reached, and for exam-
ining the sensitivity of the printing potential to changes in printing
conditions such as tolerance to dampening water supply volume
pene: printing Costs and established printing routines are a restriction
|
Printing under standard conditions
Laboratory printing Suitable for monitoring the level of a desired property

Pilot printing Difficult to repeat from one print run to another

Production printing For the objective of production printing, what applies to pilot
printing applies here also
Optimized printing of each sample
Laboratory printing Optimum values are determined for the most important printing
variables such as ink-to-water supply ratio for each paper sam-
ple tested (The reproduction of dark screen tones has been
used as a Criterion in determining optimum values.)
Pilot printing Laborious as a two-stage process (1st stage: searching the
optimum, 2nd stage: printing) (A better alternative is to select
the most important printing conditions and, if necessary, deter-
mine the optimum points from the printed samples.)
Production printing Same as for pilot printing The possible effects of random fac-
tors can be compensated by adjusting the print run and by
using parallel points.)
Printing under the optimum conditions
of the reference sample
Laboratory printing Unsuitable for this purpose

Pilot printing Suitable if the samples to be tested are sufficiently similar com-
pared to the reference sample or if a comparison with the refer-
ence sample is particularly important

Suitable if the printing is alongside production operations

Production printing
Printing “under the printer’s optimum
conditions”
Lahoratory printing Unsuitable

Pilot printing Does not fit naturally with pilot printing

| Production printing “Easiest” method but senstitive to the effects of random factors

251
CHAPTER 8

In printing tests with low end electronic printers, the user does not have any
immediate way to influence printing conditions. Testing each time therefore uses stan-
dard conditions. When making printing tests, one must always verify the standard condi-
tions separately. Experience has shown that differences in the amounts of toner or ink in
the cartridge or aging of the toner can cause variations in print quality. Middle range and
high end printers offer better possibilities for control than low end printers. The control
features typically concern the ink supply volumes and drying and setting temperatures.
Consideration of the measured results must always include reliability in accordance with
the principles of statistical mathematics.
Despite the principle of the testing program, one must remember that all printing
presses and devices differ. The testing program must therefore ensure representative
results. A correct method is to test variations in the equipment first and then select the
testing program so it meets the requirements for achieving representative results. One
way to avoid this is to analyze the results in relation to the test results obtained simulta-
neously using a known reference paper.
Table 6 shows the relevant factors for testing the printing potential of paper by
printing method.

Table 6. Relevant factors in testing the printing potential of paper.

Printing potential Newspaper | Heatset| Sheet-fed Gravure; Electro-
offset photo-graphy
Web break frequency X X
Dusting X Xx xX
Set-off x Xx
Wrinkling X
Register error sensitivity x Xx Xx x
Darkness level (ink requirement) oT ea ae x x X
Nonuniformity(compact and X x x X X oe
screen areas, Screen dots)
Halftone contrast and x X Xx Xx Xx
dot gain x X X
Color space x x xX X
Rub-off resistance x X x x Xx Xx
Print-through
Trapping Xx Xx X
Gloss, nonuniform gloss Xx X X X Xx
Sharpness Xx X X x
Waviness X Xx X
Fiber puffing Xx Xx
Blistering
Cracking Xx
Dimensional stability, bending Xx X
Curling X X Xx
Friction problems Xx X
Missing screen dots Xx XX
Solvent retention X Xx dl
Although the list is not absolute, it reflects the expectactions with the require-
ments of the methods.

252
 Paper in printing

BReferences
1. Page, D.H., Seth, R.S., TAPPI J. 65(8):92(1992).
2. Niskanen, K., Strength and Fracture of Paper (Review Paper). Fundamentals of
Papermaking. vol. 1 (Ed. C.F. Baker) Pira International, Leatherhead, UK.
pp. 641-725.

3. Tabor, D., Interaction between Surfaces: Adhesion and Friction. in Surface Physics
and Materials vol 2, (Ed. J.M. Blakely), Academic Press, New York, 1975.
4. Tobazeon, R., Electrical Phenomena of Dielectric Materials. in . In Handbook of
Electrostatic Processes (Ed. J-S Chang, A.J. Kelly, J.M Crowley). Marcel Dekker.
New York. 1995. pp. 51-82
5. Mangin, PJ., A Review of Offset Linting Evaluation. TAGA 1987 Proceedings.
Technical Association of the Graphic Arts, Rochester, NY 1987. pp. 397-442.
6. Mangin, P, Silvy, J., De Grace, J.H., Offset Linting Studies: Part II. Further
contribution to Linting Theories. 1990 International Printing and Graphic Arts
Conference. TAPPI/CPPA. Vancouver Canada 1990. pp. 109-119.
7. Aspler, J.S., Nordic Pulp and Paper Research Journal 8(1):68(1993).
8. Lyne, M.B., On the Interaction of Liquids with Paper under Dynamic Conditions.
Oxford Fundamental Research Symposium., 1993, Pira International, Leatherhead.
pp. 885-9117.
9. Oittinen, P, Saarelma, H., Graafinen materiaalitekniikka. (Materials Technology in
Graphic Arts Production). Otakustantamo, no. 493, Otaniemi 1987. p. 221.
10. Oittinen, R, The Surface Structure of Coated Paper an the Formation of Gloss. The
Role of Fundamental Research in Paper Making. (Ed. J. Brander) Mechanical
Engineering Publications Ltd. London 2:635(1983).
11. Kajanto, |., Effect of Formation on Print Unevenness with Uncoated Woodfree
Papers. PhD Thesis, Helsinki University of Technology 1991. p. 187.
12. Ness, C., G6ttsching, L., Formation of Paper and Mottling of Solid Prints. Advances
in Printing Science and Technology. (Ed. J.A. Bristow), Wiley Chichester
23:429(1997).
13. Johansson, P-A., Print Mottle Evaluation by Band-Pass Image Analysis. Advances
in Printing Science and Technology (Ed. W.H. Banks) Pentech Press, London
22:403(1994).
14. Nguyen, N.G., Jordan, B.D., Paperi ja Puu 71(8):933(1989).

253
CHAPTER 8

15. Shaw, T., Phot. Sci. Eng. 6:281(1962).

16. Schirmer, K-H., Tollenaar, D., Ernst, RA.H., Optical Densities of Halftone Prints.
IARIGAI 1971 (Ed. WH. Banks). pp. 109-1217.
17. Schirmer,K-H., Renzer, W., Normal color Intensity of Prints. Advances in Printing
Science and Technology (Ed. W.H. Banks). Pergamon Press, London 6:151(1971).
18. Shannon, C.E., The Bell System Technical Journal XXVIII(3):379(1 948).
19. Hradezky, R., Objective Qualitatsbeurteilung der Druckprodukten Moglichkeiten zur
Analytischen Behandlung von Reproduktions- und Druckprozessen mit Hilfe der
Informationstheorie. M.Sc. Thesis, TH Darmstadt 1977.
20. Kriss, M., O’Toole, J., Kinard Information Capacity as a Measure of Image Structure
Quality of the Photographic Image. SPSE Conference Proceedings (ed. R. Shaw),
1976, pp. 122-133.
21. Kuntz, W., Das Papir 23(10A):771(1969).
22. Metz, H.J., Ruchti, S., Seidel, K., J. Phot. Sci 26:229(1978).
23. Oittinen, P, Paperi ja Puu 65(1):19(1983).
24. Oittinen, P, Saarelma, H., J. Opt. Soc. Amer. 76:897(1986).
25. Oittinen, P, Saarelma, H., J. Imaging Tech. 14:15(1988).
26. Oittinen, P, Saarelma, H., SID ’89 International Symposium. Digest of Technical
Papers 20:94(1989).
27. Oittinen, P, Saarelma, H., J. Opt. Soc. Am. A3(7):897(1986).
28. Chareza, C., Greve, T., Gdttsching, L., Das Papir 41(A10):V59(1987).
29. Chareza, C., Greve, T., Gdttsching, L., Das Papir 42(2):57(1988).
30. Oittinen, P, Saarelma, H., TAPPI Journal 74(8):197(1991).
Saarelma, H., Oittinen, P, The Quality Potential of Paper in Printing. Advances in
Printing Science and Technology, (ed. W.H. Banks), Pentech Press, London
18:40(1986).
32. Scheuter, K.R., Hradezky, R., Print Quality and Information Theory. Advances in
Printing Science and Technology (ed. W.H. Banks), Pentech Press, London
213(1977).
33. Scheuter, K.R., Hradezky, R., Objective Print Quality Evaluation — Based on
Information Theory. TAGA Proceedings, Technical Association of the Graphic Arts
Industries, NY 1978.

34, Oittinen, P, Characterization of Paper as an Information Carrier. Products of

Papermaking. Transactions of the Tenth Fundamental Research Symposium. vol 2.
Pira International, Leatherhead. pp. 311-349.
35. Oittinen, P, Saarelma, H., Autio, H., J. Imaging Sci and Tech. 36(5):496(1992).

254
 Paper in printing

36. Wolf, K., Beitrag zur Systemtheorie der Druckverfahres. Dr Eng Thesis. TH
Darmstaat 1970.
37. Lin, P, Hudyma, E.., Modeling of Web Tension Response and its Application to
Newspaper Printing. TAGA 1991 Proceedings. Technical Association of the Graphic
Arts. Rochester NY 1991, pp. 588-606.
38. Linna, H., Moilanen, P, Measuring Methods to Control Web Handling during
Printing. 1994 International Printing and Graphic Arts Conference. TAPPI/CPPA.
pp. 243-250.
39. Leskinen, A.M., Influence of the Paper on the Evaporative Drying of Inks. paper in
“Advances in Printing Science and Technology,” vol, 17, Proceedings of the 17th
International Conference of Printing Research Institutes, [ARIGAI. Saltsjobaden
Sweden, June 1983. pp. 381-398.
40. Aurela, B., Raisanen, T., Journal of High Resolution Chromatography 16:422(1993).

255
BCHATR
Printing inks

1 Hk Ti th@sPrOSS es iss, opts cts eres enerae ots we yoteaceereiy scan ae aaa ate cae Zor
2 GOMPOSIUON sac. sas sven dess success <omadonnearsces maine ete ser seece ana cae ene 263
Z sh wigep IOI NIUSiewes net sehen see ciet ai, chad arene Ua tinea ee pea orice emer oa ete eee ee 265
250 BINGOS Oicuseteetttirecce ba aststiesicvct vn pa doedveosevessah relccraoeermee eke aaa at ene oe ee ae ae 268
ZO SOLVE INES Mares ae ate Trask cn Fete rae vps sage cl Mok dy cao eee a STIG nee 269
DA alias SACLCIITIN Sixers cst ou rane conetees uo osc kta ans ae one ee en 270
3 Onticalinropertiese cseersceecesctencpaoaaenn ead acu eae tee eae aataee tote eta 272
4 Rheological properties s-e2-..-. 005+ ot vascrcuveaoee cavGunes teacecoudecen lence corer eee 274
A alee DetinitONSrandDenaviOk ..sssierattv ccs vee ates ete tere Benes erecta ea, come eee ee ee 274
A= AMGaSUrinG MeOlOgiCal PFODONIGS sereccrccsccacectas ees Meee ore rene ane 278
BS OAC Kacett ere ete See RPMI Se AI ar entide SEN RID an ga I 87 2 aaa ae eee 280
4) Surface and colloid chemical properties ......................cccccceceeeececeeeseceeeueuees 281
6 Drying characteristics: of printing INK ..ccscecs.cacess.. +scwayeeceectoecess coesegeesereeeees 282
Gel MeOlVIMOTIZAUION: A eecee seer cor gerecue rst cun as ocd eile cxcess oun ase ore ee 283
fadcl(c)
hclf]bsop een hoa eer eee RON TRL Rcer ea Rawr Pre R Oto 8) MA Mei Me hh ele UT: 286

256
CHAPTER 9

Priigniis =
This chapter covers printing inks’ and their use in mechanical printing. The methods for
electronic printing discussed their use of coloring materials in Chapter 6.
Ink film thicknesses in printing can be less than 1 um, and printed spatial details
can be less than 10 um. The ink per unit area of paper in four color printing is approxi-
mately 1 gm. The mass of ink is less than 5% of the mass of paper. On the paper, an ink
film absorbs at most more than 99% of incident light of a given wavelength band. As these
figures suggest, the requirements on inks and their use in printing are very stringent.
In industrial statistics, printing ink manufacturing is a component of the chemical
industry. Compared with paper, the unit price of ink is approximately tenfold. Its usage is
only one-hundredth. The cost of ink to printers is thus 10% of the cost of paper. Ink tech-
nology shares many features with paint technology, although some major differences do
exist. A major difference is that the color of a painted surface comes from one paint, but
the color of a printed surface uses three or more inks.
The first section of this chapter explains principles related to the use of inks in the
press such as ink transportation and drying. An overview of ink composition follows this.
The remaining sections discuss ink properties.

1 Ink in the press

Figure 1 shows that the main steps concerning printing ink in the press are the
following:
- transfer of ink from a transportation container to a press container, to the ink-
ing unit, and to the printing areas of the printing plate as a film of desired thick-
ness
- transfer to the paper in the printing nip

- drying of the printed layer.

Transfer to and Information Transfer

Transfer metering in Transfer as ink to paper Printed
to press inking unit as a to plate distribution and Paper
container continuous film on plate drying

Paes
Latent
Paeson

Figure 1. Ink in the press.

20/
CHAPTER 9

Transfer of ink to the

press container from the 1 2 3
transportation container Ink film Ink film
uses pumping or a manual mf . ; — ink
method in small-scale pro- Ink \ /
4
duction. One transportation oe
container can feed several
web presses. 1 Ink flow through a gap formed by a rotating rollerII
and a stationary blade (bottom of a duct) (offset),
Transport of ink in the 2 Ink flow on a rotating cylinder which revolves in an
inking units uses the three ink duct (gravure, flexo) and
principles of Fig. 2. First, ink 3 Pumping of ink to a roller surface through a blade
can move by its own flow. assembly (newspaper offset, flexo).
The use of flow as a Figure 2. Principles of transporting ink by flow.
feeding mechanism depends
on how high the resistance to flow becomes. Flow resistance is equivalent to the prod-
uct of viscosity and shear rate as mentioned in Section 4.1 of Chapter 5 in the context of
filling of halftone cells. In simple terms, shear rate is the ratio of the speed of flow and
the thickness of the flowing layer. Power consumption in flow is proportional to the prod-
uct of flow resistance and shear rate and therefore depends on viscosity:
2
Power consumption of flow = viscosity x (shear rate) (1)

The main use of flow is when the

inks have low viscosity because of the
indicated influence of viscosity on flow h
resistance and power consumption.
The second technique of ink trans-
port is splitting as occurs in roller nips h h
combined with rotation on the roll sur- a :
faces. In splitting, the total ink layer thick-
ness divides in approximately the middle. Figure 3. Definition of splitting.
Using the symbols of Fig. 3, ink film thick-
ness on both of the rolls after splitting is as follows:
; h, +h,
8) ») 7 (2) hs

Splitting in consecutive nips h, h, he

transports ink in the direction of
decreasing thickness. Because ink is
taken off from the press by transfer to Bs Plate
paper, thickness decreases from the
infeed to the plate. This mechanism is ho No he
useful for viscous inks. Heat-set and Paper
sheet-fed offset inks belong to this cat- . ee
Figure 4. Ink transportation by splitting.

258
 Printing inks

egory. Figure 4 illustrates how the ink travels along the inking unit using a simplified ink-
ing unit.
Writing a mass balance equation for each nip gives

Zieh ectalt, (3)

2h, = hy +h, (4)

and

hy = hy
5. (5)
The amount of ink transferred to the paper is h3. Combining the previous three
equations shows that /3 relates to the input thickness, /p, by

un
hoa5-3—. (6)
Repeated splitting therefore causes the ink film thickness to decrease as the ink travels
from the ink pan toward the printing plate. If no ink transferred to the paper, all the rollers
would obtain the same film thickness. Numerous rollers in the ink train give thinner ink
film on the printing areas of the plate compared with the amount of ink fed to the inking
unit. One can generalize the situation of Fig. 4 to any number of rollers, n. The rule
relating ink transfer to the paper and ink feed becomes

ho
Loaner = Gea (7)

The reduction of ink film thickness of the previous equation is an advantageous

property in control. It means that the accuracy required in feeding ink need not be as
high as the accuracy desired on the plate. If there is an uncontrolled variation in ink feed
by Aho, the variation is attenuated in the inking unit to become Ahp/(n+1) on the paper.
Real inking units are more complex than the configuration of Fig. 4 especially in offset
presses.
The complexity of inking units relates to the fluidity of ink. Inking units are the sim-
plest in gravure and flexo as Fig. 5 schematically illustrates.
The simplest inking unit is one in which the printing plate itself directly regulates
the quantity of ink fed to the printing areas of the plate. This type of inking unit is possi-
ble in gravure where the printing cylinder rotates in the pan and lifts ink by rotation. The
doctor blade in Fig. 6 wipes any ink that lifts in the nonprinting areas. A thin lubricating
layer of solvent between the blade and the cylinder is necessary to prevent mechanical
contact and wear. This requires careful control of the hydrodynamical conditions
between the cylinder and doctor blade.

209
CHAPTER 9

ep) Heat-set and

0 sheet fed offset

a
Le

oc
Ww
= —
: Ss
faa)

= Newspaper
2 offset

\ jf

‘ ; Anilox flexo
\ vi and offset
Gravure
a

INK VISCOSITY

Figure 5. Various inking units.

One step more complicated is the inking unit where the ink transfers to the print-
ing cylinder with a separate application roller. The application roller lifts ink from a pan
as in flexo and possibly also in gravure printing, or the ink transfers to the roller by
pumping. The latter is common in newspaper offset and also in flexography.
Flexo presses use
rollers with a continuous cell Lc \

structure on the surface not Ink in cells —

unlike a gravure printing cyl- Thin lubricating layer -
inder would if it printed a prevents blade from
scratching cylinder
given, constant tone area.
The rollers in flexo are anilox
rollers. Ink flows in the cells
of the anilox roller and
between the cells. A doctor Figure 6. Doctoring in gravure.
blade adjusts the quantity.
The line frequency of the cell structure is 3-4 times higher than the screen frequency in
printing so the cell structure on the anilox roller does not interfere with halftone dot for-
mation. Cell shape is another variable of anilox rollers.

260
 Printing inks

Anilox inking units also find use in newspaper offset printing. A characteristic of
anilox inking units in offset is that the total number of rollers is greater than in flexo but
smaller than in convential offset inking units. The former characteristic is primarily due
to the complicating influence of dampening water and the more viscous offset inks.
Anilox ink application and doctoring allows more precise control of ink quantities than
feeding with smooth rollers. This allows a smaller number of rollers. It also eliminates
the need for zonal adjustment of ink feed discussed below.
The use of anilox®
in heat-set and sheet-fed Columns
presses is not possible aa \ Ye
\

Printing plate
because of the slow flow of GE PZ)
the inks in the cells. The inks Z
are more viscous than news ae CZ
inks. The inking units in In
heat-set and ae offset Sones
units are multi roller trains
with multiple paths of ink
flow and additional rollers for
smoothing the ink layer.
In conventional roller 2 £5 /e(8) Printing area
train offset inking units,
columnar or zonal adjust- Figure 7. Columnar ink consumption.
ment of ink feed is used to
take care of the variation in consumption as Fig. 7 shows. Without adjustment, infed ink
in zones of small consumption accumulates on the rollers causing smearing and losses
in detail reproduction. To verify the accumulation tendency, one assumes for the sake of
simplicity no printing area on the plate in a given column as Fig. 8 shows.

h, h, 0 |

ie he 0
Paper

Figure 8. Ink accumulation in a roller train.

This means that no ink should transfer to the plate in that column:

261
CHAPTER 9

hyot+h, = 2h, (9)

from which

h, = ho. (10)

Ink film thickness on the form roller that feeds ink to the printing plate becomes
the same as the ink film thickness originally fed into the roller train. The situation does
not change when the number of rollers increases from that of Fig. 8. This means that ink
accumulates at those columns of the form rollers characterized by low ink consumption.
To prevent this, zone adjustment of i) regulates ink feed. An offset inking unit is “key-
less” when there is no CD ink feed adjustment. Keyless inking units configured to con-
tain only a few rollers are an approach to avoid CD adjustment. Their development is an
important activity in press construction.
Evenness of inking means how constant the output to the paper remains over
time when the input is constant. Due to deficiencies in the construction of inking units,
some variation normally occurs. This causes periodic variation in printed density and
represents a performance limit of a given type of inking unit.
In offset printing, the
application of water requires
a separate feeding system.
Figure 9 illustrates the two
Ink frome
principles of feeding water to rollers
the printing plate. Water feed
uses a continuous film as in Water frome
contact methods (top) or < Contact water
droplets as in noncontact i feeding unit
methods. Droplets can be |
applied directly on ink form
rollers or on a water form
roller.
In contact water feed- Aaa
==
01010 Olas
ing units, water is normally : \

applied to the plate in the Water feed

direction of rotation at every Noncontact water
revolution before the ink. feeding unit
This means that water is
sandwiched between plate
and ink film. This leads to
mutual interactions dis- Figure 9. Ink application by the contact and noncontact principles.
cussed in Section 9.5.

262
 Printing inks

2 Composition
A variety of ink types fali in the category used for printing?%.The main components of
inks in mechanical printing methods are the pigment, the binder, and the carrier phase.
Ink making operations consist primarily of chemical raw material conversion and
surface, chemical, and mechanical processing.
Ink manufacturing consists of three steps:

- dissolution of the binder in the carrier phase

- dispersion of pigment in the vehicle

- addition of additives.

The first step uses mixers or dissolvers. The operation may use an elevated tem-
perature. Dispersion of pigment particles in a sufficiently fine state in the continuous
phase — milling or grinding — requires higher shear and impact forces than mixers gen-
erate. The specialized devices used are mills. Additives are mixed in inks using disper-
sion devices or by mixers after dispersion.
An indication of the range of ink systems is the fact that printing ink can be any
type of a multi component system. Table 1 gives definitions of the alternatives.

Table 1. States of multi component systems.

SU cae Characterization
Solution - Consists of molecular particles of the size <10-9m
- Homogeneous
Hydrosol (solubilized - Soluble surface-active substance forming an interface beween liquid phase
colloid) and partly dissolved solid phase
- Particle size of solid phase 10°? - 10-’ m
- Particles visible in electron microscope |
Emulsion (colloidal) - Liquid phase surrounded by surface-active substance as small liquid particles
in a continuous liquid phase, heterogeneous
- Particle size 10-7 - 10° m
- Particles visible in light microscope
- Interface properties have great importance
Dispersion (colloidal) | - Solid phase surrounded by surface active substance in a continuous liquid,
heterogeneous
- For other characteristics, see emulsion
|Solid dispersion - As dispersion, but solid phase in another solid phase
Suspension (coarse - Solid phase in a continuous liquid phase
suspension) - Particle size > 10° m, surface properties of minor importance
- Particles visible in light microscope
Aerosols c Liquid in fine form in gas phase
- Particle size in colloidal range

263
CHAPTER 9

Considering system structure, solutions and hydrosols are homogeneous sys-

tems. The other types are heterogeneous. The pigment component is usually in the sys-
tem as a dispersion.
The binder component is one of the following:

- solution in carrier phase

solubilized colloid

- emulsion

- dispersion

- continuous phase

- dissolved or solubilized in an emulsion.

The pigment in printing inks is mostly dispersed into a liquid phase that consists
of a carrier phase and the binder dissolved in it. The solid toners used in electrophoto-
graphy and other particle-based processes are solid dispersions. For reference, in
microcapsule inks used in self-copy papers and in microcapsule copying, the pigment is
dispersed into the liquid phase that resides in the particles. The outer shell is a polymer.
In ink-jet, if the color is a dye rather than a pigment, it is dissolved into a carrier phase.
The binder in printing inks can be dissolved or emulsified into the carrier phase to
various degrees, or the binder can form a continuous phase. In aqueous inks binders
are emulsions or solubilized colloids. In sheet-fed offset inks, the monomer of the binder
or its prepolymerized intermediate form is a continuous phase. In solid toners, the
binder is a continuous phase.
In general terms, the main function of inks in printing — spatially generating con-
trast and color on the paper — determines the composition of printing inks. The choice of
raw materials and their amounts depend on the requirements of the printing method and
the product including ecological requirements. The requirements are expressed as the
following three categories of ink properties:

- functional

- technical production

- technical product.

Functional properties refer to those that have significance for the primary function
of inks in printed products. Because the primary function is to provide contrast and
color, optical properties are functional properties.
For an ink to function in the press, it must meet certain requirements that depend
on the printing method. Functioning in the press is related mainly to rheological proper-
ties, surface chemical properties, and drying properties.
An acceptable ink must be user and environmentally friendly in the sense that the
prints should have sufficient durability and be toxicologically and ecologically accept-
able. Table 2 compiles types of materials used in inks with their proportions and func-
tions.

264
 Printing inks

Table 2. Printing ink composition.

Component Function Description

Pigment 10%-30% Light absorption to | Insoluble fine particles
Black: carbon black give color dispersed in the continuous
Cyan: phtalocyanines phase (= vehicle) consisting of
Magenta: azo pigments and salts carrier phase and binder
Yellow: azo pigments
Binder 10%-30% Bind the pigment Amorphous polymeric materials
Pitch (newspaper printing) particles to the paper | called resins or monomers of
Phenolics (all types of inks) and give gloss these or oxidizing vegetable oils
Alkyds (offset)
Hydrocarbon resins (news inks)
Drying oils (sheet fed offset)
Acrylates (UV and aqueous)
Nitrocellulose (flexo)
Carrier phase 0%—/0% Provide the neces- | Solvents (boiling point < 100°C)
Mineral oils (offset) sary fluidity and oils (boiling point > 100°C)
Vegetable oils (offset)
Toluene, xylene (gravure)
Water (flexo)
alcohols, esters, ketones (flexo) a

The proportions of the constituents vary widely depending on the type of ink. The
main types are liquid inks, aqueous inks and oil based or paste inks. Flexographic and
gravure inks are liquid or aqueous inks, offset inks are paste inks. Within a given type
such as inks for a specific printing method, the margins are tighter.

2.1 Pigments
Several factors influence the color of prints starting with the atomic structure of
pigments* and extending to compatibility of the ink and the substrate. The influencing
factors are the following:

- atomic level: atom groups in pigments resulting in absorption of light

- molecular level: molecules containing groups that absorb light

- pigment crystal level: crystal structure

- pigment particle level: size and shape

- printing ink: compatibility of pigment and the vehicle dispersion

- printed image: compatibility of printing ink and paper.

At the atomic level®, distinction is made between organic colored groups and inor-
ganic ones in color formation. The color of organic substances and pigments derives
from atom configurations that selectively absorb light. There are chromophores such as
Neen C.= 0) 6 = — NV = N- aN =.0,— CH = CH =

265
CHAPTER 9

and their color strengthening auxochromophores such as

—SO3, -NH,, -OH, —NO,, -CH;, —Cl, —Br
Inorganic colored compounds are the salts of transition elements Sc, Ti, V, Cr,
Mn, Fe, Co, Ni, and Cu and compounds that contain an element representing two
degrees of oxidation such as Fe (Il) Fe (III) cyanides.
For organic color compounds, there is a connection between the structure of the
molecule containing chromophores and auxochromophores and the wavelength of light
absorption. Molecules with a great mass and closed structure absorb radiation with a
long wavelength (red), and straight-chain molecules with a small mass absorb radiation
with a short wavelength. At the molecular level, other structural parameters than mole-
cule size influence the color such as
- the degree of cyclicality: monocyclicality vs. polycyclicality

- type of chain: straight chain vs. open chain vs. closed structure.

Monocyclical, open chain structures absorb shorter wavelengths than polycycli-

cal closed structures. Considering molecular structure, pigments can have the following
Classifications:
- straight-chain (yellow, red, magenta)

- polycyclical (blue, cyan, green)

- open chain: salts of acid pigments (red, magenta, cyan)

- open chain: salts of basic pigments (red, magenta)

- inorganic pigments (white, yellow, blue, cyan, black)

- carbon black (black).

Figure 10 shows the

relation between color and Polycyclical
structure in organic pigments. closer er uctule
N = aromatic ring
The most common yellow pig-
C = aromatic ring
ment is benzidine yellow. WwW

Magenta pigments a = Mono- orpolycyclical

rubines, naphtols, and 5 open structure
rodamines. Cyan pigment is ee N=N Monocyclical
phtalocyanine blue, and black tn C=O straight chain
pigment is carbon black. coc
C=O

cyan magenta yellow

red green blue

COLOR ABSORPTION

Figure 10. Relation between molecule structure and color.

266
 Printing inks

Table 3 lists pigments commonly used in process printing inks.

Table 3. Typi cal pigments of four-color inks.

Type Chemical structure Color --

Azopigments Monoazo:-N=N- diazo: R- N=N-R'-N=N-R" Yellow
Magenta
For example: Red
acetoasetarrylamides, arylidic yellow, arylamidic yellows, diarylamide
yellows, diaryl yellows = benzidine yellow naphtols
Polycyclical C= Nand | Blue
pigments C = structures composed of aromatic ring units Green
Cyan
For example:
phtalocyanines C3,H\4CuNg
Inorganic pig- | Lead chromates: lead yellow
ments Cadmium sulphides: cadmium yellow

lron ferro cyanines

XFe**[Fe**(CN)¢] ce H>O; X: NH4, Na, K
For example (X = NH,):
milor blue, iron blue, bronze blue, Prussian blue, Chinese blue

Treated calcinated kaolins:

ultramarine Blue
Salts of acid Salts of azo pigments (rubines) Magenta
pigments [R-N=N-RX-],Y** X: COO, Cl, SO3, OH, Y: Ba, Ca, Na [
Salts of basic =| R- X*Y- Magenta
pigments R=X+Y-; X:NH(C>Hs), N(CyHs)*, _Y: Cl, Br, CuFe(CN)¢ Red
For example rodamines: PMTA, victoria blue Blue |
Carbon black Contains Black
(channel and C90%-99% organic substances 1%-10% (O, H, S, hydrocarbons)
furnace black)
manufacture:
by burning oil (furnace soot) or natural gas (channel soot)
Blue tone,
brown tone

Soot (lamp Contains: Black

soot, vegetable C91%-99.7%
soot) moisture 0%—-2.4%
extractives 0.1%-0.9%
ash 0.01%-0.3%

Manufacture: Gray
| by burning tar or crude oil distillation residues

267
CHAPTER 9

Soluble dyes in printing inks adjust the tone of the ink to make it different from the
tone of the pigment. In addition, they can adjust the density of the color and provide flu-
orescence. In ink-jet inks, dyes may give the ink the actual color intended. For reasons
of image quality and archiving durability, pigment use is increasing.

2.2 Binders
As a general rule, the binder content in inks is approximately the same as the pigment
content. High quality inks typically contain more binder. Inks can use a wide range of
binders. This is partly due to the fact that there are several sources including plants
(drying oils), wood processing (pitch, resins), crude oil distillation (hydrocarbon resins),
and chemical synthesis from which binders may originate.
In addition, the requirements set on binders by the carrier phase of the ink and
the printing substrate cover a wide range. The binder and the carrier phase must be
mutually compatible. This means that the binder has to be soluble or dispersable in the
carrier phase in a controlled manner. The binder and the printing substrate should also
be mutually compatible. The binder should adhere sufficiently and bind the pigment on
the substrate. From a physical viewpoint, the total quality level of printed matter is
largely a result of the quality of the paper used. Logically, paper therefore influences the
choice of binder. Surface chemical principles predict that similar chemical substances
adhere best.
Printing ink binders are oils or resins. Oil used as a binder typically dries through
oxidation in air in contrast to nondrying oils used as a carrier base. A mixture of oil (dry-
ing and nondrying ) and resin dissolved into it is a varnish.
Resins can be natural resins and semisynthetic or synthetic resins. Because
binders are polymeric materials, their structure consists of the following characteristics:

- chain length (molecular weight)

- chain branching and stiffness

- double bonds in the chain and secondary forces between molecule groups
(ion forces, hydrogen bonds, dipole interaction, and van der Waals’ forces)
- reactive groups occurring in the chain.

All these characteristics influence the glass transition temperature, melting point,
solubility, adsorption behavior, and bond strength of the binder. Printing ink resins have
the following features:

- initial monomer is bulky: average molecular weight of approximately 300

- resin molecular weight averages 1000-3000

- binder often has a polar and a nonpolar end.

The low molecular weight promotes solubility and low viscosity. A small share of a
higher molecular weight fraction, approximately 5000, prevents the resin from penetrat-

268
 Printing inks

ing into paper. The existence of two types of ends aids adsorption with different kinds of
surfaces.
Glass transition temperature and film formation temperature adsorption proper-
ties, bond strength and drying potential are functional binder properties.
The minimum film formation temperature is the temperature at which a viscoelas-
tic material spread as a uniform layer produces a continuous film as a result of perma-
nent or viscous flow. The minimum film formation temperature is typically the same
magnitude or slightly higher that the glass transition temperature. The latter is the tem-
perature at which the behavior of a polymeric material changes from that of an elastic
solid material to that of a viscoelastic solid material. Below the glass transition tempera-
ture, the structural changes occur in the end groups of molecular chains. Above it, the
degrees of freedom in molecular movement are greater with structural changes occur-
ring in different parts of the chains.
The melting point is the temperature at which the behavior changes to that of a
viscoelastic liquid. The glass transition temperature depends on the stiffness of polymer
chains that is mainly influenced by the frequency of cross bonds and the proportion of
side chains.
The drying potential of oils drying by oxidation depends on the existence of dou-
ble bonds between carbon atoms as Section 9.6 discusses. Radiation induced drying
resides in the capability of monomers to radicalize.
In oil- and solvent-based printing inks, the polymeric binders generally dissolve in
the carrier base. The solubility depends on the structure of the polymer and specific
interactions between the polymer and the solvent. Structural factors are the stiffness of
the polymer chains, three-dimensionality, and crystallinity. The most important interac-
tive factors are electrostatic interactions and hydrogen bonding.
Aqueous printing inks®” find use in flexographic printing of newspapers and
packaging materials. In rotogravure printing, aqueous printing inks represent a growth
area. Environmental issues serve as a catalyst for development. In rotogravure printing,
currently available printing equipment and requirements for quality standards restrict the
use of aqueous inks. The range of binders suitable for use in aqueous inks is much nar-
rower than the range of binders soluble in solvent- and oil-based printing inks. For this
reason, binders are also in the form of hydrosols and emulsions when used in an aque-
ous system. Another alternative is ink in which the binder (containing alkali-soluble acid
groups) has been dissolved in water by raising the pH of the water with volatile alkali
such as NH,OH. The alkali evaporates from the ink on paper with the solubility of the
binder deteriorating rapidly and the binder precipitating on the surface of the fibers.

2.3 Solvents
According to the carrier phase, inks fall into the following groups:

- liquid inks (gravure, flexo)

- oil based or paste inks (offset, letterpress, silk screen)

- aqueous inks (flexo, silk screen).

269
CHAPTER 9

In offset, the combination of the binder and the oil is a varnish. The general term
used to denote the combination is “vehicle.” The carrier phase becomes superfluous
after transferring the ink to the paper. It undergoes removal by evaporation (gravure,
flexo, heat-set offset), conversion into solid form by oxidation (sheet-fed offset), radiation
polymerization (sheet-fed offset), or absorption into the paper (newspaper offset).
The ability to dissolve binders is the most significant functional property of bind-
ers. Removability from the print or conversion to a solid are equally important process
requirements. Lack of odor and environmental acceptability requirements concern the
product and the production.
Solvents used in printing inks are hydrocarbons. High boiling aliphatic hydrocar-
bon distillates are mineral oils. Fractions reaching their boiling point in the range
230°C-320°C with a boiling point interval of 20°C-—30°C within a fraction find use in
heat-set printing inks. Fractions with a higher boiling point find use in newsprint and
sheet-fed offset printing inks. Liquid toners contain pure aliphatic hydrocarbons.
Ring-like aliphatic hydrocarbons result from hydrogenation of aromatic hydrocar-
bons, i.e., by adding hydrogen. Their dissolution capacity is comparable with that of aro-
matic solvents, but they lack the typical smell and toxicity of aromatic solvents. The
aromatic hydrocarbons, toluene (C6H5CH3) and xylene (C¢6H4(CH3)>), find use as sol-
vents in rotogravure printing. Their use in this application takes advantage of their good
dissolution capacity and volatility. For work safety reasons, some countries restrict the
use of toluene and xylene in printing inks. Mineral oils also typically contain aromatic
substances to increase the dissolution power.
Methylated ethy! alcohol finds use in rotogravure and flexo printing of packaging
materials. Propanol finds use in aqueous rotogravure and flexo inks because of its com-
plete solubility in water.
Cyclohexanol and methyl cyclohexanol are special cases for use in silk screen
printing. In the glycol category, ethylene glycol, OHCH, — CH>OH, has application
because of its hydrophilic properties. It has use in water-diluted and moisture setting inks.
The latter are special inks. The same applies to other two- or three-valency alcohols.
Solvents belonging to the group of glycols are solvents for various binders such
as tall oil-based resins and nitrocellulose. Applications consequently include rotogra-
vure and flexo inks. Examples are ethylene glycol and monomethy! ether.
Ketone solvents typically have good dissolving capacity. Methylethy! ketone
(MEK) finds use in rotogravure and flexo printing of packaging materials. In the acetate
category, ethyl acetate has use due to its ability to dissolve nitrocellulose.

2.4 Additives

Besides the main components, inks contain various additives that may have a profound
influence on the functioning of the inks in printing. The additives are typically surface
chemically active, i.e., they tend to migrate to and accumulate at surfaces such as
pigment surfaces (on the micro scale) and on ink layer surfaces (on the macro scale).
This means that additives have an effect even in amounts as small as a few percent.

270
 Printing inks

Table 4 gives a summary of commonly used additives.

Table 4. Additives in printing inks.

eae dyes
Additives
ali Purpose and Use
- Modify the shade provided by the pigments
- Used in all types of inks
Filler pigments: white mineral pigments Reduce the color strength of the ink
- Add “body” in all types of inks
|Wetting, emulsifying, and dispersion agents - Wet and improve dispersability and stability of pig-
ments (and emulsion type binders) in inks
Used in all types of inks
Gelling agents: Al compounds Increase the “body” of ink by networking after ink
transfer to the paper in an effort to reduce smearing
and spreading of ink
Use in oil-based inks
Waxes: polyethylenes, polytetrafluoroethyl- Reduce surface energy and prevent adhesion of wet
enes,paraffine waxes print to other surfaces
Prevent physical contact
- Reduce ink tack
- Used in heat set and sheet fed inks
Diluents and cosolvents: alcohols Reduce ink viscosity without influencing the binder
Used in liquid inks
Defoamers: silicone compounds Prevent foaming
- Used in liquid inks
Plasticizers: esters with high molecular weights Improve flexibility of printed ink layers by having a
dissolving effect on binders
- Used in high quality inks
Drying catalysts: organic Co and Mn containing - Initiate and speed chemical drying
compounds in oxydizing inks, aromatic ketones - Used in sheet-fed offset inks
in UV polymerizing inks
Drying inhibitors: reactive compounds |- Prevent chemical drying in can
[- Used in sheet-fed inks

Additives find use in modifying original ink properties as dyes, fillers, wetting,
emulsifying, dispersion and gelling agents, diluents, cosolvents and defoamers do.
Some influence ink drying and properties of the dry print. These are drying inhibitors,
drying catalysts, waxes, and plasticizes. Besides additives that influence drying from
within the ink, the application of additives on top of a printed layer as a separate step
can influence the smearing of printed layers in the press. This type of additive includes
starch powder used in sheetfed offset. Its influence is similar to that of waxes in inks
shown by Fig. 11. These function by reducing surface energy and by providing a physi-
cal barrier.

271
CHAPTER 9

Wax or starch powder

O-O OQ-O—9
Cane paces
Figure 11. Physical influence of powder additives.

In heat-set offset, silicone solution is applied after drying as a thin layer to reduce
the surface energy of the printed layer.

3 Optical properties
Four-color printing uses printing inks that are transparent, i.e., the inks selectively
absorb specific bands of wavelength of light and allow transmission of unabsorbed light
as Fig. 12 shows. This is not fully true; inks usually scatter some light. Lack of
transparency is opaqueness. Paints are opaque. The purpose is to hide the background
color. If an ink were totally opaque, underprinted ink layers would not influence color of
the printing. At the other extreme — the goal in process color printing, the color of
overprinted ink layers does not depend on the degree of overlap because the inks are
transparent. This was discussed in Chapter 7.

Incident Incident
light Reflection light Reflection

Transparent Partly opaque

1 Surface reflection
2 Reflection reduced by light absorption
3 Scatter

Figure 12. Definition of transparency.

ele
 Printing inks

Spectral absorption of inks has two aspects:

- selectivity of wavelength

- the magnitude of desired absorption.

The selectivity of spectral absorption can be measured from prints spectrophoto-

metrically. The results are commonly given as CIE Lab or CIE Lch coordinate values.
Standards and recommendations are available for the color of ink sets. Table 5 lists the
CIE Lab color values of the European standard inks.

Table 5. Colorimetric values of standardized inks (Euro standard).

L*

Cyan 93.75 —24.05 —48.55 04.15

Magenta 47.60 76.65 —4.20 76.75
90.00 12.77 92.56 93.45
The range of colors obtained on a given paper by using a specific set of inks is
called the color range. Figure 13 shows an example presented in the plane defined by
the chromatic CIE coordinates a* and b”*.
The magnitude of desired light absorption is commonly measured densitometri-
cally through a narrow band filter that transmits the color of light that the ink should
absorb (red for cyan inks, green for magenta, and blue for yellow). A higher density
reading means the ink is a more efficient light absorber. The quantity of ink necessary to
produce a specific level of density on a reference paper is ink require-ment defined in
Section 6.2 in Chapter 8 and illustrated in Fig. 14. Ink mileage is conceptually the oppo-
site of ink requirement.

b* A

Yellow
Reference level

Red

Green Magenta
DENSITY Ink requirement

Cyan >
Blue
QUANTITY OF INK, g/m?

Figure 13. Color range of inks and their overprints Figure 14. Definition of ink requirement.
presented in the a*b* plane.

273
CHAPTER 9

The relationship between the amount of ink and density is linear at small quanti-
ties of ink, becomes curvilinear with an increase in the ink amount, and approaches a
saturation level. The linear portion is consistent with the prediction of the Lambert-
Beer’s law. The nonlinearity is due to the influence of surface reflection. A higher gloss
of the ink layer equates to a higher saturation level of density.
The above comments suggest that ink gloss is also a relevant ink property. This is
true since print gloss depends on ink gloss, paper gloss and ink and substrate interac-
tions.

4 Rheological properties
4.1 Definitions and behavior
Rheological properties are
fundamental to deformation
and flow of materials.
They’9 can predict —
together with surface
chemical properties — the
transferrability of ink in the
press and interactions with
the paper for ink spreading SHEAR
STRESS:
¢
and penetration. Figure 15 SHEAR RATE, D
shows the basic rheological
property of fluid materials — >
viscosity.
Viscosity is ameasure
of the resistance (measured x a >
as shear stress, Pa) to flow
that arises when a material F
is made to flow at a given Sh Bs Se
relative speed (measured as
shear rate, gs"): Figure 15. Defining Newtonian viscosity.

5 (11)
Printing inks used in different methods have viscosities that fall in the approxi-
mate ranges of Table 6.

274
 Printing inks

Table 6. The order of magnitude of viscosity in different types of inks.

Ink type de Viscosity (room temperature)

Newspaper offset 2-10 Pas
Heat-set offset 5-15 Pas
Sheet fed offset 10-50 Pas
|Gravure 5-50 Pas
Flexo 5-50 Pas
Letterpress 0.1-10 Pas

The above data are very approximate because the viscous behavior of offset inks
and letterpress inks at the higher end of the scale is shear rate dependent as Fig. 16
shows. Ink viscosity decreases with the rate of shear. The behavior is shear thinning or
pseudoplasticity. To be precise, this term implies that viscosity with an increase in shear
rate reaches a constant level. Shear thinning does not.
Oil-based inks may require a given shear stress before any flow occurs. This is
the yield value, and it is an elasticity expression. The magnitude of the threshold stress
is small. The term apparent yield value depicts the intercept of the tangent of a flow
curve at a finite shear rate indicated in Fig. 16. An ink is plastic when its flow after yield-
ing is Newtonian i.e. the flow curve is a straight line. Inks that have high real or apparent
yield values are difficult to pump. This may impede transfer from ink pans to rollers. A
high yield value is advantageous for halftone dot sharpness.

A A
oD
Ww
or
kK
o
z IE
x S
Apparent o
yield value S

Real eee
yield value a ae
SHEAR RATE SHEAR RATE

Figure 16. Shear thinning (pseudoplastic) viscous behavior.

Figure 17 shows that oil-based inks are also thixotropic. Thixotropy refers to flow
curve behavior that depends on the history of shearing. If a sample has initially been at
rest, shearing causes a decrease in viscosity (upward curve). The opposite occurs if the
sample is in a sheared state (downward curve) at the start of the test.

275
CHAPTER 9

A A

”
2
or
fp
D
KE O
) < O
oc BS we
ia Thixotropy a
eT
a)

> =
SHEAR RATE SHEAR RATE

Figure 17. Thixotropy.

Inks are also viscoelastic. This means they have an instantaneous deformation
component in the flow. Pigment structures may develop elasticity at low speeds, and
binder molecules may develop elasticity at high speeds. Elasticity influences halftone
dot formation and ink splitting. Moderate elasticity appears to be advantageous.
Table 7 summarizes the importance of the rheological properties of printing inks
in the printing process.

Table 7. Impact of the rheological behavior of printing ink for the printing process.

Property Impact
Viscosity - Flow of ink in pipes, lifting by rollers from containers, flow through press aed
- Transfer to paper, tackiness, penetration and diffusion on paper
Shear thinning - Fine-tuning of the behavior of printing ink: transfer vs. set-off, tackiness |
Thixotropy - As above
Viscoelasticity - Behavior in connection with rapid change of shear: tackiness, transfer to paper
- Behavior at low shearing rate

Solvents and oils used in printing inks are liquids with small molecules whose
rheological behavior is Newtonian. If the molecular weight of a polymeric binder is
small, its content is low, and its solubility into the solvent is good, the rheological behav-
ior of the medium will be Newtonian. Otherwise, the medium will exhibit shear thinning
behavior. Its behavior may also show significant elasticity. If the printing ink contains
gels, the flow is typically plastic and thixotropic.
Adding a pigment to the medium increases the shear thinning behavior primarily
so that the viscosity level measured at low shearing rates increases more than at high
shearing rates. If the proportion of pigment is low, the medium retains its typical flow
curve. Weak floc structures in the pigment particles increase the thixotropy of the print-

276
 Printing inks

ing ink. The flow curve may also display a real yield value as a result of the interaction
between pigment particles. In the area of the yield value, there may be a certain elastic-
ity in the flow.
Ink tack is an experimental quantity defined as the resistance of a thin ink film to
Splitting (cf. Section 7.3.2). It is relevant for lithographic inks. Figure 18 shows that ink
films can split by shear or by tension.

Tension

Shear

Figure 18. Splitting by shear (left) and tension (right).

At the exit side of nips,

both types of force occur. At A
the initial stages of splitting Filament elongation:
close to the nip center, flow tension controlled
curve behavior — viscosity — Caviation
primarily controls the resis-
tance to splitting as Fig. 19 Two possible
~ Curve shapes
indicates. As splitting RESISTANCE
progresses, an ink layer can
no longer flow and fill the Shear
controlled
space between the two >
receding nip surfaces. Inter- TIME
nal rupture or air entrain- Middle of Splitting
ment — cavitation — occurs. ne point
As splitting progresses fur- ee
ther, ink around air bubbles ian MN MIA
he
forms into filaments that
Or
es inal)
elongate before they finally
split. Liquid inks split largely Figure 19. Splitting phenomena.
controlled by shear. Tension
forces are likely to play a role in oil-based inks due to higher viscosities.
Roller devices commonly measure ink tack. The tack reading obtained corre-
sponds to resistance integrated over time, i.e., it is the area of the resistance curves
shown in Fig. 19. With an increase in the speed of printing, ink tack usually increases.
As a first approximation, tack depends on the product of viscosity and speed. Tack also
depends on viscoelasticity — more elastic inks have lower tack values. Because of the

mark
CHAPTER 9

shear thinning behavior of inks, measurement of the viscosity should use a shear rate
that is representative of the speed conditions in the nip.
When a given amount
of ink rotates in a roller a
device, its tack value increa-
ses with time due to evapora-
“Slow”
tion of volatile solvents and setting
oils as Fig. 20 shows. The
slope of the relationship —
tack vs. time — can predict the TACK
setting and drying behevior of “Fast”
sheet fed and heat-set inks, soning
respectively. At some point.
the slope tends to become
negative, and the tack value >
: : ‘ TIME
reaches its maximum. A high
value is disadvantageous Figure 20. Changes of tack with time.
because it indicates high
stickiness of the drying film. The downward part of the curve represents decreasing ink
cohesion in the test.

4.2 Measuring rheological properties

Rotational viscosimeters and rheometers are the common means to determine viscous
and viscoelastic ink properties. Capillary rheometers are suitable for measurement of
ink-jet inks, and viscosity cups are suitable for routine control. Table 8 compares the
features of different viscosimeters.

Table 8. Features of different viscosimeters.

Capillary viscosimeter High-pressure capillary viscosimeters: D = 10-1 08 =

- Poor suitability for measurement of thixotropic samples (Shearing rate in
measuring nip not constant)
- Suitable for measuring rotogravure, flexo and newspaper offset inks
- Correction terms must be marked at high speeds
Viscoelasticity measurement with die swell and jet thrust methods possible
Rotation viscosimeter; Shearing rate range: 107 1407 g1
conical viscosimeter - Suitable for measuring different types of printing ink because the shearing
rate in the nip is constant
- Secondary flows and breaking of samples at the free surface restrict the
maximum rate
- Viscoelasticity measurement as normal force or vibration dua
Rotational viscosimeter; Shearing rate range: 104 ey 08 st
coaxial cylinder viscosimeters |- Suitable for measuring rotogravure and flexo inks; closed measuring space
prevents sample from evaporating
Viscosity cups Shearing rate is determined on the basis of viscosity; not constant in the cup
- Suitable for determining the viscosity level of Newtonian liquids (used for
controlling the viscosity of rotogravure and flexo inks in the industry)

278
 Printing inks

Measurement of viscoelasticity can be a low-amplitude vibration measurement

and a normal force measurement under a uniform shearing rate.
Vibration measurement can be done in the basic geometry of viscosity determi-
nation betweeen two planes. One plane moves in a sinusoidal manner. The change of
deformation is the following:

é = e,sin wt (12)

where e, = amplitude of vibration

The rate of deformation change, ie. the shearing rate is as follows:

é = @ e,cos Wt (13)

If the sample behaves linearly, the shearing stress comes from the sum of the vis-
cous and elastic component. In vibration-based measurement, viscosity is dynamic vis-
cosity, and the shear modulus is dynamic shear module. The viscous stress lags 90°
behind the strain. The elastic stress is in phase with the strain. As a result, the dynamic
complex viscosity is as follows:

ula aems
ul (14)
where 1’ is dynamix viscosity
G' dynamic shear modulus.

The expression for shearing stress (t) applying these definitions is then:

Cec, @|n) COS ote sin or (15)

In the case of a purely viscous sample, G' = 0.

The loss tangent is calculated from dynamic viscosity and dynamic shear module
as follows:

16
tan@ =
= 1® (16)

where © is phase angle.

For a purely viscous sample, @ is 90°. With increasing elasticity, the value
decreases. At the frequency range of 107'-102 Hz, the phase angle of news inks typi-
cally is 80-90°, and that of sheet fed and heat-set inks is 50—80°.

2/9
CHAPTER 9

4.3 Tack
in contrast to basic rheological properties, tack is not a pure material characteristic. The
measured value depends on the measuring conditions. The tack value also depends on
factors related to the design of the measuring device such as bearing friction.
Tack can be measured in the geometry formed by two planes or two rolls. In the
former case, it is possible to record the splitting event as a function of time from the
beginning to the end of the splitting. The latter measurement gives a value influenced by
all the phases leading to the splitting event.
The roll geometry is more suitable for simulating the conditions in printing. It is
therefore the more common of the two methods. An important factor is the speed at the
point where the surfaces diverge. Based on the peripheral speed of the rolls and the
layer thickness, the shearing rate in the nip is u/h when uw is the peripheral speed and h
the thickness of the layer. Because of deformation in the nip region, the actual shearing
rates are 1-2 decades lower than the theoretically calculated rates. In the nip, the speed
at which the surfaces diverge is approximately wR, when R is the radius of the roll and
t is the time measured from the center point of the nip. The acceleration of the diverging
movement results from the expression u-/R, and the corresponding shearing rate
comes from the expression u-t/(Rh). Using realistic values for printing speed and roll
radius, the diverging acceleration is approximately 500 ms*—50 g (g acceleration of
gravity) — and the shearing rate is ee

The roll tack device consists of three rolls as Fig. 21 shows:

- ink distributor roll

- driven rubber-covered roll

- freely rotating roll (rider roll).

Force Riding roller

measurement

Shearing foller

Figure 21. Basic principle of tack measuring device.

280
 Printing inks

Because of the splitting resistance of the printing ink, the rider roll is pulled in the
rotating direction of the driven roll. A balance occurs between the splitting resistance
and the force created by the spring fastened to the rider roll or the strain of the strain slip
(indicates the roll’s position). This provides a measure of tack.
The tack measured with the aid of the tack measuring device is formed in an inte-
grated manner across the contact zone between the printing ink and the roll surfaces. In
this way, the tack value depends on

- deformation in the nip region (affects length of nip zone and the magnitude of
the shearing rate)

- the flow of ink in the laminar zone in the center part of the nip (flow resistance
proportional to the product of viscosity and speed)

- the splitting phenomenon on the outgoing side of the nip

- the friction in bearings.

Because of factors related to equipment design, the values obtained with different
tack measurement devices may differ in the direction of change caused by differences in
the most important measuring variables such as layer thickness and speed. Since the
printing ink is viscous, an increase in layer thickness results in a decrease in splitting
resistance, and an increase in speed causes an increase in the splitting resistance. As a
first approximation, tack depends on the product of viscosity and speed. Tack also
depends on viscoelasticity — more elastic inks have lower tack values. Because of the
shear thinning behavior of inks, measurement of the viscosity should use a shear rate
that is representative of the speed conditions in the printing process. If the changes in
measured tack are contrary to these, the changes are due to equipment-specific factors
overriding the effects of the sample itself.
When a given amount of ink rotates in a roller device, its tack value increases with
time due to evaporation of volatile solvents and oils as Fig. 21 shows. The slope of the
relationship — tack vs. time — can predict the setting and drying behavior of sheet-fed
and heat-set, respectively. At some point the slope changes, becomes negative, and the
tack value reaches its maximum. A high value is disadvantageous because it indicates
high stickiness of the drying film. The downward part of the curve represents decreasing
ink cohesion in the test.

5 Surface and colloid chemical properties

When considering an ink, the phenomena in the form roller and printing nips primarily
have the character of surface phenomena. In these nips, the ink transfers to the plate
and the paper. In offset printing, the ink also interacts with the dampening water. The
interactive behavior of ink in relation to water is called lithographic behavior. Apart from
macroscopic surface chemical properties, the surface tension (or energy), interfacial
tension (or energy), and ink and water interactions are also influenced by colloidal
properties and rheological properties. Colloidal properties depict the surface energy
conditions on the micro scale, i.e., on the scale of pigment particles.

281
CHAPTER 9

Colloidal properties also influence the stability of inks in the press with changes in
temperature and chemical environment. Without cooling, the temperature increases
with the time of running, and the chemical environment may change due to mutual dis-
solution of constituents from the ink, paper, and dampening water. Lack of stability may
be manifested as agglomeration of pigment particles in the ink with losses in optical effi-
ciency and accumulation on the press rollers.
The surface tension of liquid inks can be measured directly. Measurement of the
surface tension of oil based inks is more complex and uses the contact angle principle.
Techniques to determine interfacial tension against water also exist. Standard surface
chemical methods are static, but printing is highly dynamic. Surfaces form in millisec-
onds. This has led to development of dynamic test methods.
Emulsification tests commonly measure lithographic ink behavior. The test mixes
ink and water in given conditions and measures the water up-take of the ink gravimetri-
cally. From the curves of water up-take vs. time, one can read the speed and saturation
level of emulsification as Fig. 22 shows.

1 "Slow" emulsification, no stable level

3 2 "Rapid" emulsification, "medium" stable level
3 "Rapid" emulsification, "high" stable level

%
INK,
IN
CONTENT
WATER
TIME
Figure 22. Mixing of water in ink in a test situation.

Mixing of water in ink in printing is necessary to eliminate water from those areas
of the plate where it is harmful, i.e., from text areas and from the top of halftone dots. It
is necessary between the dots. Fast emulsification is desirable because the printing
conditions after start-up are not stable until the saturation level of water pick-up occurs.
The optimum magnitude of emulsification depends on the type of water feeding unit.
This is because of differences in the efficiency of water use. A higher efficiency means a
lower water consumption.

6 Drying characteristics of printing ink

In the ink drying process, two phases can occur: the ink setting phase and the actual
drying phase. As a result of the setting phase, the printing ink forms a touch-proof layer.

282
 Printing inks

The drying phase results in a rub-resistant layer shown in Fig. 23. The phases can use
the same or different mechanisms. For example, the mechanism is the same when the
printing ink contains a solvent that requires evaporation during the drying phase
(rotogravure and heat-set inks) or when the printing ink is polymerized after each press
unit (UV inks). Conventional sheet-fed offset inks involve two separate mechanisms.
The setting uses filtering and evaporation, and the actual drying uses polymerization by
oxidation. Newspaper printing inks do not have any drying stage.

Rub-off test Set-off test

Figure 23. Active force in testing touch and rub resistance

The following discussion only concerns chemical drying. The most common
chemical drying mechanism is polymerization of the binder. Precipitation of the binder —
a physico-chemical mechanism — can occur in aqueous inks.

6.1 Polymerization
A polymerization reaction occurs in printing inks with UV or IR radiation or due to the
oxygen in the atmosphere around a thin ink layer. In the future, electronic radiation will
probably also find use as a source of reaction energy. In all these cases, the
polymerization is radical polymerization. The polymerization of inks dried with the aid of
UV and IR radiation is analogous to the photo-polymerization used in printing plates.
The polymerization process has three stages:

- initiation

- propagation (chain reaction)

- termination.

The initiation stage produces radicals by the following:

homolytic decomposition

hv 5 p*
ene (17)

separation of hydrogen

Tee REL TH +R (18)

283
CHAPTER 9

transfer of electrons

1+M wT + M or | +M* (19)

where J is initiator
s radical
M monomer.

In polymerization by oxidation, the initiation mechanism can be the last alterna-

tive. In polymerization by UV radiation, all initiation reactions are feasible.
In the chain reaction, the radicals react with the monomer units that then become
radicals and react further with other monomers and with any pre-polymerized resin
polymer (UV inks):

Le ial
a
M,*+M,>M,M,* (20)
M,*+P—>MP*

The reaction chain ends when two radicals react with each other:

(M1) (Mo), eM) Die (21)

The initiator used with oxidizing inks is a drying agent usually consisting of metal-
lic salts of organic acids dissolved in oil or in a solvent (liquid drying agent). Suitable
metals are those with two degrees of oxidation such as cobalt and manganese. The ini-
tiation can be as follows:

++
Co + ROOH — Co" + RO* + OH
cs ++ (22)
SOH = Co = RO “HO

The metal content in printing ink is 0.05%—2% by weight. In the drying agent, it is
20% at most. The salts are typically octoates, resinates, naphtenates, and linoleates.
The surface tension of the drying agent determines whether the agent will migrate to the
surface of the printing ink. Depending on this migration, the main effect of the drying
agent is directed to the surface of the printing ink or inside it. Organic metallic salts can
be added to the ink as a paste. Drying agents other than those already mentioned avail-
able as pastes are borates and inorganic peroxides.
The binders used in oxidizing printing inks are natural oils and alkyds. Their reactiv-
ity uses carbon double bonds. In the propagation stage, binder radicals react among
themselves to form more radicals for further reaction. The result is a polymeric structure.

284
 Printing inks

The catalysts in UV inks are typically aromatic ketones. Homolytic decomposition

involves the following transition:

O OR, O OR,

II | | | (23)
IANO ANONfee Nd OOS OICIie

Separation of hydrogen involves the following:

R,C = = + R,CHOH >R,C- OH + *R,COH (24)

The transfer of an electron occurs as follows:

Ph=C =O+R,N>R,N* +Ph-C=O0*. (25)

Monomers in UV inks are commonly esters of acrylic acids. A general formula of

a monofunctional monomer in which R represents esterification and the rest of the mol-
ecule is acrylic acid is as follows:

KR—O>C—
CH = Ci,

I (26)
O

Difunctional monomers contain two carbonyl groups, and trifunctional monomers

have three. Acrylic chemistry has an important role in printing ink chemistry. Aqueous
inks are another area of application. They are therefore the focus of ink development
from two directions — radiation curing and aqueous printing.

285
 Printing inks

BReferencees a
. Printing Ink Manual. Ed. D.E. Bisset, C. Goodacre, H.A. Idle, R.H. Leach, C.H.
=~

Williams. 3. ed. Northwood Books, London 1979. 488 p.

2. Fadner, T.A., Bain, L.J., Worldwide Status and Progress of Lithographic Keyless
Inking Technologies. TAGA 1989 Proceedings. Technical Association of the Graphic
Arts. Rochester NY 1989. pp. 545-568
Todd, R.WE., Printing Inks. Pira Internationa. Leatherhead UK 1994. p. 427.

& Chemical Technology in Printing and Imaging (Ed. J.A.G. Drake). Royal Society of
KR
Chemistry. Special Publication No. 133. Cambridge 1993. p. 188.
5. McLaren, K., The Colour Science of Dyes and Pigments. Adam Hilger Ltd, Bristol
and Boston, 1986. p. 209.
6. Surface Ohenomena and Fine Particles in Water-Based Coatings and Printing
Technology. Ed. M.K. Sharma and FJ. Micale. Plenum Press, New York 1989.
p. 331.
7. Surface Phenomena and Additives in Water-Based Coatings and Printing
Technology. Ed. M.K. Sharma. Plenum Press, New York 1991. p. 310.
8. Oittinen, P, Fundamental Rheological Properties of Printing Inks and their Influence
on Ink Behaviour in a Printing Nip. Acta Polytechnica Scandinavica Chem include.
Metallurgy Series No. 131. Helsinki 1976. p. 192.
9. Lyne, M.B., The Importance of Extensional Viscosity in the Impression of Ink into
Paper During Printing. in Advances in Printing Science and Technology (Ed. W.H.
Banks), vol 20. Pentech Press, London 1990. pp. 236-248
10. Aspler, J.S., Sui, O., Zang, Y-H., Bonding between Vegetable Oil Inks and Cellulose.
TAGA 1995 Proceedings. Technical Association of the Graphic Arts. Rochester NY,
894-914.

286
 Conversion factors

MiConversion facto rsS=__

To convert numerical values found in this book in the RECOMMENDED FOR\M, divide
by the indicated number to obtain the values in CUSTOMARY UNITS. This table is an
excerpt from TIS 0800-01 “Units of measurement and conversion factors.” The complete
document containing additional conversion factors and references to appropriate TAPPI
Test Methods is available at no charge from TAPPI, Technology Park/Atlanta, P. O. Box
105113, Atlanta GA 30348-5113 (Telephone: +1 770 209-7303, 1-800-332-8686 in the
United States, or 1-800-446-9431 in Canada ).

To convert values expressedin |... To obtain values expressed In

(Property |RECOMMENDED FORM | Divide BY |CUSTOMARY UNITS a
Area square centimeters [cm?] 6.4516 square inches [in?]
square meters [m2] 0.0929030 | square feet [ft2]
i square meters [m2] 0.8361274 | square yards [yd?]
Energy joules [J] 1.35582 | foot pounds-force [ft Ibf]
joules [J] 9.80665 meter kilogams-force [m e kgf]
millijoules [mJ] 0.0980665 | centimeter grams-force [cm e gf]
kilojoules [kJ] 1.05506 British thermal units, Int. [Btu]
megajoules [MJ] 2.68452 horsepower hours [hp e h]
megajoules [MJ] 3.600 kilowatt hours [kW e h or kWh]
kilojoules [kJ] 4.1868 kilocalories, Int. Table [kcal]
joules [J] 1 meter newtons [m e N]
| joules [J] 9.80665 meter kilogams-force [m e kgf]
Force per | newtons per meter [N/m] 9.80665 grams-force per millimeter [gf/mm]
unit length | Kiignewtons per meter [kN/m] 0.1751268 | pounds-force per inch [Ibf/in]
Frequency | hertz [Hz] 1 cycles per second [s-1]
Length nanometers [nm] 0.1 angstroms [A°]
micrometers [um] 1 microns
millimeters [mm] 0.0254 mils [mil or 0.001 in]
millimeters [mm] 20.4 inches [in]
meters [m] 0.3048 feet [ft]
L kilometers [km] | 1.609 miles [mi]
Mass grams [g] 28.3495 ounces [02]
kilograms [kg] 0.453592 | pounds [Ib]
metric tons (tonne) [t] (= 1000 kg) |0.907185 | tons (= 2000 Ib)

287
 To convert values expressed i |Toobtain values expressed In
Properly |RECOMMENDED FORM "ive bY |CUSTOMARY UNITS
Mass per | grams per square meter [g/m?] 3.7597 pounds per ream, 17 x 22 - 500
plated grams per square meter [g/m?] 1.4801 pounds per ream, 25 x 38 - 500
grams per square meter [g/m?] 1.4061 pounds per ream, 25 x 40 - 500

grams per square meter [g/m?] 4.8824 pounds per 1000 square feet [Ib/1000 ft?]
grams per square meter [g/m?] 1.6275 pounds per 3000 square feet [Ib/3000 ft?]
grams per square meter [g/m?] 1.6275 | pounds per ream, 24 x 36 - 500

Pressure, | kilopascals [kPa] 6.89477 | pounds-force per square inch [Ibf/in? or

stress, psi]
Nes Pascals [Pa] 47.8803 | pounds-force per square foot [Ibf/ft?]
kilopascals [kPa] 2.98898 feet of water (39.2°F) [ft H»0]
kilopascals [kPa] 0.24884 inches of water (60°F) [in H»0]
kilopascals [kPa] 3.38638 inches of mercury (32°F) [in Hg]
kilopascals [kPa] 3.37685 inches of mercury (60°F) [in Hg]
kilopascals [kPa] 0.133322 | millimeters of mercury (0°C) [mm Hg]
megapascals [Mpa] 0.101325 | atmospheres [atm]
Pascals [Pa] 98.0665 grams-force per Square centimeter
[af/cm?]
Pascals [Pa] 1 newtons per square meter [N/m?]
kilopascals [kPa] 100 bars [bar]
Speed meters per second [m/s] 0.30480 feet per second [ft/s]
millimeters per second [mm/s] 5.080 feet per minute [ft/min or fpm]
Viscosity Pascal seconds [Pa e s] 0.1 poise [P]
millipascal seconds [mPa e s] 1 | centipoise [cP]
Volume, milliliters [mL] 29.5735 ounces [02]
fluid liters [L] 3.785412 | gallons [gal]
[voume cubic centimeters [cm] ~| 16.38706 cubic inches [in?]
cubic meters [m°] 0.0283169 | cubic feet [ft*}
cubic meters [m°] 0.764555 | cubic yards [yd°}
cubic millimeters [mm°] 1 microliters [ML]
cubic centimeters [cm*] 1 microliters [ML]
cubic decimeters [dm] 1 liters [L]
Aeiiemeters [m°] 0.001 liters [L]

288
Miindex ES RO
business communication .............cccceeeee
a
a a ROEM gtr TTR a ie OER are 98 DUSINGSS FOU sentir sect ceemerteeee ecae ut
20) dO) eS Re Re Sere ee ate er ee 98 D)VIG Seca te ets catty. out ter eee 38
MUSOUILELCON dG tates pa eee eee eae 237
BUS ONILIEMSRV geet tet ciate ele197 C
AUSOLDUOM a reser anne a ttca a194, 196, 273 calibration
POLO Naeetondan tesearsten caves, coer cue costh? 52,56 CO Oleweac:e-antate: 44 47, 96, 105-106, 109
HOOT oe cok 68, 80-82, 89-90, 104-105, GOLSHY ith acc cr ceei gen eee eee eee 202
194,196,265-266, 273 CAMG lide ears 46, 51-52, 55-56, 69, 86, 115
Ue De De hearer me Menten esate 204-205 CAINE deonter ea careaceete aces te tea ae 22, 86
ACHNOMIAU Cues, Se Sarge, 77, 104, 247-248 carrier phase .............. 182, 263-264, 269-270
BCIMOU AC DADE seta cieaann antares: 205, 247-248 GOS ING echaiectie cemtan teeny oer ae ere oe eas
achromatic reproduction.............ccce cece. 104 CaunOGE AVstU Dees eee ees 152559;.53
GCHMOAM dO GSarrstessc tree aaiancncne 32, 49, 70 CAV ILA Olen entessnsucl oeaee eee ak eetone atl
additive color formation ..........0.cccceeee 81, 86 CC ie aecase es on See eee Ie nae 164
adhesion.....147-149, 152, 181, 184, 229-231 COMMON MOLeS SION ese e series sane ees 161
LEK een tency ah een tree ce 147-149, 152 COM FANZOC ORIENT sce ae een enone 120
ROIG Ieee teeta
nist er eee 181, 184 CRAM Clas creek aareeesccais earns LE eee22:(0)
NU aeareeaea ee cake eon aetna eae tata 229-231 GHanMGli CANA Cty es san. nsec cere eee es 59, 240
AMAOCNCOD Vigra seseredaeattesece eee ee ae 178 GANG COE. wack wee erates ene ean oo
IIL XGUONG etn ace eke Pe sts ee a ee 260 ChalieisC COUT cena cae nner ae are serene tee 55
ADPanelU WGI VANUO tis, aces cscxcen ee eager 215 character probability coding................... 53-54
area probability COGING..........::::s2cs0:-.c0s2s 93-54 Charge tOcmMaSS TatlOesuer) .cecec ne eter 183
GINO IMals ene fo eect ere ee 77, 248
B CHLOMALIGeeeicsan cases 79, 202, 248, 273
archi
(0AR tA hecas cee coe ne eee eee eee 59 GIGO TGs OAT Ciemergers ene nee 136, 144
DaSiS WEIN sea nealanccd 214, 216, 220, 222 6 |Ee eon oe aro Aa PRO ee ee, 76-77, 273
DUCla WOIALCSave cen aseigtnaneea tartrate ares aati 144 CIES CHa nate! eksaciccd setae in cee DAVY Pali)
DIR OCI Staaten wr emspeakal re264, 268-270, 284 GIES UM Se Ane! Senin, Manan e iin cea, eee 76-77
PUM Oeawiseras eereae teeters: 21, 124, 166, 223 GlOG CIN creas cares sea teenie tee en ane nae 187
1 eae eae 29-30, 38-41, 43-44, 53, 55, 72, GIN ateersee 70-71, 74, 86, 101, 106, 109
189, 237, 239, 244 CMY Korie Soir a saretct e cad acts, uci eee 78
DNL esseaerotnis. aeaacee vcmmrene ae dae teeta 176 COCK NMOS eaetant2, sheet tern de eutccta at 188, 227
a) ETRE Ti ae RE i ee oe RA ae race a2 COUGH a are 29, 39, 49, 53-55, 106, 112-113
blanket-to-blanket printing ............c.ceee 138 COGING/s neues 16, 35, 38-40, 44, 53-55, 113
DESECRURGiaete teine cere rc,esata 220,20) 246-247
DIU Cec: 36, 58, 68, 70, 72, 74-75, 78, 80-81, GOANON een tkee nis cere nee 186
83-84, 86, 89-90, 101, 104, 266 GONGSIONia. eect ea ast oa: 147-150, 278, 281
DARI ERUI Che orth. orte ecco Terence 98, 200, 206 COLI DAISGest ae n hones tees ate 181
OOK eitesi ss aeei 14-15, 20, 124-125, 223 CONANO(exe a ces. chee aman 124-125
DROAOSIEEU re ev coerce: 40, 164 COMCCU PKODUCHON eestnw cates tet tes: 164

289
COMOIMAINDTODGINGS sercceeeetee tet. aes 281-282 detail reproduction..............0. 33-34, 98, 205
COlOKeee 14, 18, 22-23, 28, 35-36, 44-45, 49, CEVELODMENE FEACTION sicsceceeececatrccannestesre
ornareSy
52, 55, 58-60, 62, 66-82, 84-91, 94, device independent color............ 105, 109-110
96, 98, 101-102, 104-110 (O[EVA ONG on: oe) SS ae ees ne a nae Bee 144
COIORCOMCCHON tet ae 45, 85, 88-91, 98 die'StamOniGes saat ae eee 223
COOMOdINUUeeeemeecten errr errr eres il GIGIECIICe tee ia ee 178, 180, 229, 253
COIOMHOISG eee eee eae eee 105, 236 diffusely reflecting Surface ........0.. cee 196
color range......74, 81, 91, 101, 105-106, 155, digital ....13-14, 19, 22, 24, 28, 30, 34, 38-39,
203, 233-234, 248, 273 42-44, 49, 52-53, 55-56, 62, 72,
COlOICCISICL Mrerrentnrere: 201, 217-218, 228 84, 86, 93-97, 116-117, 173, 175,
color separation ...... 22, 55, 74, 78-79, 85-90, 244, 247
106, 110 CiGitalCOp Vind: ts en a ceeetce ees 121,178
COLON TEMPCTALUTE s.ccecasescceasvcocsaees
ence60, 66-67 GIGital TAXWIG te csercces chet pera eee ees iZ2
colorimetric coordinates...................0 74, 202 digital image............ 28, 30, 38-39, 43-44, 49,
COIOLIMOTE Callens ane weet cosuse, cones ere eeceane 154 56, 62, 93
communication process......... 13-14, 119, 239 Gigital media sesscacesaietann esr eee 13-14, 19
COMMUTE MUON inne encie enerreece. 23-24 digital PHOUOOLaDDW as, 32, eee ee 185
computer to plate:.......... 23-24, 114, 142-144 digital printing ............... 24, 62, 120, 122-125,
computerto press..23-24, 114, 142-143, 169 (SIMIGTAt Zoned
COMPULEKAOD Vita erste eres: 23-24, 114, 169 CUNO cance we eames eee eee 130
CONLACH AN Clerc ate: 149, 169-170, 282 dimensional stability........... 184, 188, 228, 231
COntaCtS CLEC Mittens te ee ete nates 92 CikeCt IErMOG EADY occ. eeccacte neers 178
continuous tone images............. 22, 29-30, 85 GISPALC Ieee encece ements 121, 126-127
CONTONG Sere eee eeu eR aero 30, 38, 49 GISDEISION Hare eee 228, 263-265, 271
contrast .32, 34, 36, 38, 49, 56, 178, 188, 208, CISD ete os anes: 17, 19, 22, 32-33, 35,
235-238, 264, 268, 280 38, 43, 49, 55, 69-70, 81-86, 104,
COMMASTSEMSILVITY sare seccete enero. 36, 208 TIG=111, 1tomer —129
contrast sensitivity function ................. 36, 208 CISSEIMIN ANON ees heen.eee 13-14, 21
conventional printing.............. 56, 95, 143, 161 CISTMOUTCCD IME ecg eeesece eee 120
COSINGMrANSTOUM eee eauci se tee. enema 54 GOGUINGNT cee os 14, 20, 39, 43, 109, 113
ChACKIMNG cae ree tae ascents 231 EOD Gio ce, cue te eee 178, 185-186
(ORE ET (1 RRA a eae, BHR ap NT tc PV 223 dOUGalist ete 105, 200, 206, 210, 237-238,
GEO PING cresssccers matecceranrsattoen eatcat can ace"86 244-245
Goleta more eine chien aout 178, 185-187 Cotprotileimodelicr.<ac,ccre cree 108-109
CURIM a ee etree eee 188, 227-228 GOt-SNaMMNGSS = aacecy cece eee 239, 275
GUILINOtte cn arnt: vaca 130, 166-168, 218, 223 GOLStRUCTUTG: ene 91-93, 95-96, 105, 144
GWalieersu- canoe: 70, 74, 80-81, 89-91, 106, double DLOguCTION =. ececs eee 164-165
154-155, 203, 248-249, 266, 273 CRY OUSC hicras rarest tenance tena ene ee tee 149
CIEVINIG pees 139-140, 156-159, 161-163,
D 181, 187, 224-225, 231, 237, 252.
264, 268-269, 271-272, 281-284
BPALoSbeat CirePAA ft tie ark eee CE ERIN 188 CVO aheeeregeen 58, 70, 187-189, 191, 264
dampening water....138, 147, 155, 230-231, 236 dye diffusion thermal transfer ..........0..0.0... 188
Gata COMmMMteSSION tenses ene earn: 53 (YG SUDINNAUOUi sneer see 188
Cate tloW eee cee ee eee seme 38-41, 43 dynamic range............ 33-34, 36, 99, 243, 245
GenSitVieereet ene: 32-34, 36, 38-39, 52, 59, 70, GYMAMICVISCOSILY: 25 ceccnse eee ee eee 279
78,81, 87, 89, 93, 96, 99, 106-109,
197-202, 233-234

290
 Index

E G
EBICUMING Wares sele Lees out oo Beate oe 159 GAETECOIOMCINCG CtearcRascreseceuee
eek ete 78
COGS IIOISG hee tenet ecee tere deate 238-239 CLE Ligttce pits eae ee ae ant 56, 145
ClaSti@ GETONMALOM: crseeetw esate ence 149-150 GEOMEIG OPClatlOnS emcees ee es 44 46
electrical charge pattern... 178-179 GlaSSeSClOe Maree ere ee eee 91-92
ClEGtVOraSSISte enero ae See a Hol GlobalkOperalOnss eee aeee eee 44 46
CIECTLOGKDN Veen errs eee ee eee 178 GIOSS rer cree ear ee ee33, 196, 209,
ClECtOM Dealit IMAGING esterases tte ee 178 216, 223, 234-237, 243, 246, 248,
ClECIFONIC PAINTING! ssc: 14, 24, 39, 56, 94, 253, 274
116, 121, 124, 173-178, 190, 222, GIOSSSCONIAStE eee ie eee 239
290, 279 GIOSSTNOISE erence eee ee 207, 236-237
electrophotography ............... 90-96, 128, 173, OIA ATION sree ee eee mee ree 52, 57, 60, 91-92
178-181, 183, 188, 190, 211, 216, QhANMINGSS oysters ethno oes aa eee tee 236
230, 237, 239, 264 GAMULA TITS Re hehe sh eee eet 61, 236
BlSCIKOSEALIG IMCLIOUS a ye tee eee 178 GraphiCSeenaee 18, 20, 22, 30, 38-41, 54, 114,
GIMISSION: mecee eer 70-71, 73, 80, 82-83, 87 li P1350-250
CNMUISiiCatiOnier ct tare rtct acces 282 gravure.....1386-137, 140, 142, 144-147, 151-
OMOAVIN Glecersckatt ese: 15, 140-142, 144-146 152, 155, 157, 160-161, 163-166,
enhancement................. 34, 36, 44, 86, 89, 96, 168-169, 215-216, 233, 237, 239,
98-99, 101-102 259-260, 265, 269-270
CULODV erent ter creme mee nce Jo-04 Gravure Cylinders... 140, 144, 146
equivalent neutral CENSIY s.scces.sseee cscs 89 Glav-DalanC@ ste eee tent ee ee ee 36
EXC emer tre este rer even tetuorss cme 144-146 GiayacOMPOnel a eeneetee
atectee: 81,104, 155
OTCGtiee ore 36, 58, 68, 70, 72, 74-76, 78,
F
80-81, 83-84, 86, 88-89, 101, 104,
LACSiiill Caeeee eck creer 99, 122 188, 266
AX eee ea ee tei aes bes dine ese ena, AB 122 CHICK CCUGNCVn iil See prema. 41
FD) pee wena ter tee Arion nce ctivtcndn seen 42 GUtCNDelUiz se eter tee See
TULIMNGEIMeeecert stent rues esnaae 200-201, 237
AIEMO PEKAUONS sis teoreceete Heda, carte 44-45 H
AUTISM eerecne sees: 20, 22, 124, 126, 129-130, halftOnesahigle iectncssn shite eee 93
160-161, 166, 215-216, 223-224, NalitOnerGOmbldSher ese. =screeeencee 237-238
22/-228, 231, halitOnei OUSIZeerreeeree sete 93, 96, 106
first surface reflection ...................06 194, 202 halftone: frequency :....-+4.ceeeenee 40-41, 43
flexography ............... 135-136, 142, 161, 166, halftone image.............. 30, 38, 49, 91, 96, 136
169, 222, 237, 260 halftone percentage ..30, 92, 95, 200, 202, 237
TLOWAOSISLAICO cee eee 208, 281 haittOnin (ese 45, 85-86, 91, 93, 95-97,
MIOLESCONCO re ct eesti an Geen 268 116-117, 137, 186, 189, 237
foldindeee 124-125, 130, 157, 166-168 NardiGOpy een yee: 49, 55, 127-128, 191, 193
223,220 heat-set offset ........... 140, 162-164, 168, 218,
TOUMEAIN ate hace tec tet as eee 138 210212
FOUMECIFAnSIONN ee te: 46, 48-49, 109 MGT CEOMINUN Geer. aetee Sear aeuteruee 176
TACTARCOO IN teeter et rectice Sat nares: 53-54 HIGHT AGatlOl ee Marnartescrertaeies cen 57, 91
friction... 184, 217-218, 222, 225, 227-228, MOUWANECNV IN) tees cet ne, stececcrmcces
eres Ary ats eye
293, 280-281 NOL D MUSING het crenieacteene teen: 181, 230
EVOLASTAITIDIING etme ae natacataerstee
Me 135-136, 161
HISMee token tian te, tse oeucn eae 74, 101
hue .....67, 72-74, 84, 101-102, 202-203, 205

291
 Nieran gle ee ate ree 202-203, 205 J
FDERIIOCIG syste ca cee! ten oem 19 JOU aca eee 186-187
VID ORO XT en cae eee care oe eneer ncaa sta IQ AD SIND ele cet tee es eee eee 36, 76

| L
image ............ 15-16, 18, 22-24, 28-39, 43-62, laboratory SCale <5 etereee 154, 227
67-74, 79, 81-87, 91, 93-96, 98-99, landSCape OFGNtatiOn csc cence.-ees-2vsoseerneceees 165
101-102, 105, 109, 193-194 LaS@r: DFMMUNC = sats. ee ees 178, 180, 191, 216
image analysis ......... 37, 46, 237, 239, 242, 253 LAVAOUt sccvedees Sete Ave evet tact tek eee ete 114
image enhancement.....34, 36, 86, 96, 99, 102 LChivaunspaiete te aa Bee ee 202, 273
image restoration ................. 62, 85, 96, 102 LED uauseene momen eee! 180, 200, 216, 282
LTA CSCIONEG serine coms tieruceve cane-nteceeate 62, 116 letterpress .....15, 94, 111, 185-137, 143, 147
image signal......28, 32, 43, 49, 52, 55, 81, 98, 161, 269, 275
207, 209 light absorption......... 68, 85, 89-90, 194, 196,
image transform COdING ..............ceeeee 53-54 198, 204, 207, 233, 266, 273
imaging system....... 14, 33-34, 36-37, 46-47, lige CHWSON cstee woe cece e.194, 199, 205, 238
49, 51-52, 69 lGit QUANtas seers: enc ceiisnre ese eee 67
NNTPACH CHINN se. eee secre. 94, 161, 177, 190 light scattering .....84, 194, 210, 233-234, 245
ITD OSTEO erect os set nen stant, sone ects 125 LEG: PROSSURGE tree. : eee eat eee eee 150
information......... 13-14, 17-22, 24, 34,37-41, livte Spread TUMCTION cateke. xree-cesecuneee ss 48, 245
43, 47, 49, 53, 55, 86, 94, 96, 99, HALLSOss catcinta Sted e, 225, 227, 229-231, 253
102, 117, 239-248 ViCHUEEL TE)KSee ssc era, contac 265, 269, 277, 282
information capacity......37-38, 206, 211, 213, HUGE BOMGNS 25h otek cea en eee: 182, 270
239-248, 254 TEM OGKADITIC: PIALOSS cs.s-cacos tee oe eee 142
HORSTOMICAN TOK Se erates 2casseeees 213, 254 FOLGE SUR CCT UNEGos e sot ee 18, 20
IMPORMALONISUMUICTURC ec.ccucstrcdeccteee: 18-19, 22 LONG) GI AllVa7: gets ret ee 164
TULSS TATA
TSE(07), le eee a eee een 2045210 ROW ORIG! DFRIIELING 225. 4m 2 mah ee eee eee 174
HK ADTCROILES aces Ca. oye, (5.0 ae ee log luminance.......... 29-30, 32, 52, 69, 72, 76-77,
INK KEQUILEMONE =. .ccscosntuncoosrasecvane 233-234, 273 83-84, 104, 202, 205, 239, 248
MERSIN Garcia aeectcte Rouen ere 1829276 WEED ADOT: chic: oceania eee een eee 138
WK tramSte lhCQUAEION |...<0:.5.001.csansmaree eee 2
=" ink-jél.......... 56, 128-129, 135,141, 144, 173, M
176-179, 184-188, 216, 237, 239, macro scale....... 153, 193, 199, 208-209, 234,
264, 268, 278 237, 242, 270
HUTPAQUUN IDS Paes agers,
ccanes259-260, 262, 279 magenta ........ 70, 74, 80-81, 88-91, 104, 106,
HaKSi tee 15, 24, 47, 59, 79-81, 85, 89-91, 188, 203, 266, 273
94, 104-108, 186-188, 257-259, MeQN@EOCra DN, cic exc teat ee eee 178
263-265, 267-279, 281-286 THADBING trate kak S, fone he kha eee 44
input-output relationship....44, 47-48, 50, 106 MASKING EQuatiONS .........cccccseaceseeeee 106, 109
IMSCCUNG erence 149, 164, 168, 242-243 246 MD .....164, 166-167, 218, 223, 227-228 231
IWATA CTO ss tease or eee ee te a ley media ......... Boa lo, biawtet tomo nt2he2t
intensity ............ 32-33, 47-49, 51-52, 72, 92, WONTONYcuter Bees 24, 36, 39, 54-55, 72, 106,
105, 142, 193-194, 196, 198-199, 117, 179, 199, 209, 239
204-205, 233, 238, 254 MMeMOly COIS Ae ee SOpf25 1405209
interfacial tension...............008 149, 281-282 IGF UING se ae oc ccc a eee eee 86
internal reflection ............... 196, 198, 202, 207 MCTAIMOL Ve cmc cates teeaeen eeeree meee Tlaf4e29
internal surface reflection................. 194, 202 MEP 2uscotes seen, 8 eects eee aenearea na 174
INVISHDIESCale Me ceeeeere sree es. eae 208 MiGhOSCale yarns eee 153, 193, 208-209, 234,
LR CRYING ce icenratinrernnremmemetencee cetaee aes 156 237, 242, 270, 281

292
 Index

IVIIO TaN Olle e secd ae see ayer el157 output...28, 34, 38-41, 43-47, 50, 52, 55, 85—
TISSHNGUOA OLS aectecers cee ct cote 238-239 86, 96, 99, 105-106, 108-111,
modulation ...34, 38, 48, 52,59, 136, 141, 180 113-114, 129, 207, 240
modulation transfer function............ 34, 48, 59 oxidation ..157, 159, 231, 266, 268-270, 283-
MONO COMPONENT TONES 3.3. eeteens 183 284
MTF....34-35, 48, 50, 108-109, 206, 245-246
P
FUNCCicer med Pott cee Renee yee tt 62
ANUTCAIN FORMATOM: Wc ante ele: 240 page data 20, 23-24, 173, 175, 179, 181, 185-
NOs li aa eee ee 215 186
Page AeSCription CODE ...c.....cccceceeceeeeees 55, 174
N page description languagé..............c0ccc 114
NG ee cating. utero enen eee ee te 238 Dage MOUATIVES: :.etcaecta re mee 115, 142
[email protected]. 107-109, 201-202 PAGE POSITIVES) settee kp ee ee 142
MIDIDTESSUMNG ceacccorenat tees 135, 150, 152-153 PAGE: SDC oui Cohesion ted eee ene iheaey, AVES)
IVE) WIC ape oer ursaniriehuaaeocie 150 PAQiAUONex ce ketatcy a aa ee ae 20
noise .....28-29, 33-37, 45, 50, 52, 69, 72, 96, WADCIRONAUG Er: csecnac: scence ean meee 21 oN232
98, 105, 116, 183, 206-209, 211, Panel Paci chee: ae eee 194, 205, 234
234, 236-239, 241-242, 246-248 Paper reSiSTIVITY meee ceheeee te 183-184
IMVISIDIGESCAlG ein eee eee 208 DASSIVEAMAQES 0, att iess terap keene meee 32, 49
macro scale153, 193, 199, 208-209, 234, DONGE PLUG! AGG ss vn. casement vases 146
237, 242, 270 DENCCUMO easy. tan ined cerca aes 138
micro scale........ 153, 193, 208-209, 234, POTLORAMON emer needar etc enter eee tapas
237, 242, 270, 281 POUCH AVON Mrccotcctaeaee eet an Reena 233
PONCONTAGEIMeCHIOd sa eetre ses ee einen 128 DitaSesGhanGeyinkS etcetera 186
nonimpact ...94, 128, 136, 140-144, 173-174, DMOSPMOFESCENCE wu wenn semen eee83-84
176-179 DHOtOGONGUCIONA etre 179-181, 183
POU UAUTOLM Tiree ee nee een eee 13ofcor HNOLOOCADNICSPCC Ct ciency weccncee seer 61
NOZTG ee Rar ere hit sseha mts 178, 185-187 photography......16, 22, 46, 49, 55-60, 62, 70,
OieOiea 17 LO doo
0 photomechanical................. 56, 140-142, 146
object .49, 51, 66, 68, 71-74, 85-87, 110, 227 PINQUOTIS reaerate eau arcetinas heen men ozo elo
GLERi pot Ser Mette es eee dace DINOTON OVI Clete sansa cece seco nena ee 59, 144
OfSelie vans 43, 115, 122, 135, 137-138, 140, physical nature: Of COOLS accccaraeee ee 110
142-144, 146-149, 151-152, 155- DNV SIGAIStHLIGLUTG 5 eveececetatiortreeeses18, 20, 24
157, 160-166, 168-169, 215-216, physiological nature Of COlor............. ee 110
236, 281, 283 PISZOCIECTIECIMK=|OU, Sesctesaceecs cere eeeetees 186
offset lithography ....... 122, 135, 142, 147, 161 NIGMIEM Lae ae 145 boewor 102, 1905193,
COMES CTMOIALG fat exeersececteea
eaters 115, 142, 144 215, 230, 234, 263-266, 268, 270,
ON SEMOMIMCI Ce orth ee rottacereeest ee, A 276-277, 281-282
oi1157, 234, 236, 265, 268-270, 282, 284, 286 INCH Etccrrceccauie. thc eaten ne een 268
DASE MK See ee etn Sensuous pacasr nacre 282 plate making........ 24, 121, 123, 126, 140-144,
(0)fieO( OIL lrecabpa ter Annee eee hil 14, 21 146, 161, 173
OU AMUCMOCS tinea eiegeret Antara ca havencartie De DOMTLODEHATONS pan corerseer a ca 44-45
optical character recognition.............. acl ea point spread function ......45-48, 50, 108-109,
Optical enSity <soeesceS-cteeaneets 82) 1b0n21OIZ38 194, 200, 205
OPUCAl SIMA scare Beeeeet sescceacento 49, 52, 86, 121 polymerization............... 58=09, 159,231,270,
OOlIcAlSPreadinG: ves we-rteseees 194, 200, 206 283-284
BUiGalesy SiO Marae eaetandacsctnnccoameenrar 51 DOM CATMOLIGTILATIOMattes ct nhe ose teaches axial 165
AN erect a OUR ect Mian. ceckeeaaeneraas8 188 NOWEMS DECITUMicwwaaeeciet arene 208, 236-237

293
Die-DL6SSee enna 1H INY AVE Soy 02 (5S ripping........ 23-24, 39-42, 121, 173, 175-176
precipitatlOnes atesen cee eeentee eae eee 283 OMI tesa ta daeteen ete eee 1552216,.289
DKOSSOLMAse erecta
ee eee: 160, 164 rotary principle ate ee eee 134
OINUTICOUGN ee eeteea4 sees hee ees 234 roughness .151, 153-154, 195, 209, 214-216,
printability ......... 191, 213-216, 231-233, 237, 234-235, 237, 248
239, 249 rub-offitesin Ae Bek ae ee ae. 156
printed) COMMUNICAMON seer eteeece rn seer 241 UD-DROOMIESS eee pee eee oe159, 237
PRUING CAC acer nese ceeAS: Whee lisloy, Wee runnability .......... 183, 213, 223-224, 231-232
printing nip........ 105, 134-135, 137, 146, 150,
152, 155, 2245-230, 279):286 S
DRUG Pape leQradeStyercste tec swree 213 SAMPING TCONCNE: 4. hence eae eee 38
printing reproduction process ..22, 50, 98, 119 SATCHEL ceo ee ee ee Roe ee 161
DEMING Uilts.n cetaceans 155; 157, 161-163, 216 Saturation.......... 72-74, 77, 84, 101-102, 104,
DhIWAteICOMMIUNIGALIOM zd. ccsece: conver cee erence 14 198, 248, 274, 282
PDROCUGHIOIIeDIIMNIC meee ecscawoece: 1499231 SCANNE( chew nies eae eee 51853, 85
ROOM termes et scces ose scen sealesen ne T1852 168 scanning........ 36, 38, 50, 52, 83, 85, 121, 142,
DSCUCOPIAStiGumesseeccmcte: coarse armen 215 174-175, 179-180
psychological nature of Color... eee 110 SCANMING AOQUENCY.4: Ae. Boe ee 85
DUDINGATONcie asccne 14, 21-22, 113, 166, 170, scattering coefficient .........:...s:c-s000.--. 204, 245
185, 286 Screen freQUeNCy................0:002: 38, 93-97, 260
DUDISINING meet tence 13-14, 22, 24, 50, 62 Sei INFORMATION «sc. eee eee ee53
sensitivity ........ 36, 52, 56-61, 67, 69, 87, 144,
R 208, 228
faclavlOnCuUnlnGuDViU Ve: .ance
ener eeeeee ee157 SONSILOM EI sexi cots nice cae. c ete Cee 59
radiation sensitive .............. 140, 142, 144, 146 SEriaraphy mete h iia ee ee ee ee 139
raster image processor .................. 39,058 175 SOuING scenes. 16, 21, 156—157,.163, 169, 230;
FASLETESCAN corn eesti nav, erneenle 28, 39, 52, 54 2364252, 210, 2/6.2elecoa
COUGIING Reese eR Peer ane ee 24 SGM Ree eee en toe ee ae 18: 013d
redierewtes: 58, 68, 70, 72-75, 78, 80-81, 83-84, SMAGOW [ASK feshcer ata serene ee ne 84
86, 89, 104, 198, 203, 248, 266, 273 SHANDNESS ane ee 36, 96, 155, 183, 206,
RECS TANG ere cter seats. 161, 218, 220-221 238-239, 275
OhaGuvetiGeX were eae teen 195-196 SHOAN FAlGi cence ene. 208, 274-275, 278, 281
Felallve: COMMAS ea. pete eee 32, 238 Shean THINNING. ..ference eee 275-278, 281
GE MOISUUAZAUOM ectraaracres feats eet eee hee 157 SISO ca taceacs sete eae re eect eee 235
REDIACEMOL texnacer deta der, Meek eee 44,104 SHON OFAN. rweee eee eee ee 164
reproduction............ 14, 17, 22, 33-34, 37, 46, SNOW, thTOUGlic Meee seem ree ee es cece 234
49-50, 52, 58, 71-72, 74, 79-82, Sigiidheteaeees.: 24, 28-29, 32, 34, 36-37, 43,
85-86, 88-89, 91-92, 96, 98-99, 48-49, 52-56, 81-82, 84-86, 89-90,
101, 104-107, 109-111, 115-116, 98, 106, 109-110, 121, 207-209, 242
119, 203, 244 signal to noise ratio......29, 34, 36-37, 52, 207
reproduction of color.............. 71, 79, 110, 116 SUICSCTOCN eae 129, 1388-139, 269-270
resolution ....31-32, 34, 36, 40, 43, 52,55, 85, SINISE PrOCUCTION asec eee eee 124
114, 142-143, 155, 174-175, 179, SIZING. .c. 5s: esenae Seca acannon 46, 86, 214
182-186, 188, 206, 246, 255 SIUUIMG ere aeeerneceece ene 166, 168, 223, 230
fESONMING POWOlen. sateen te, cee 34, 59, 61 smoothness............... 152-153, 157, 184, 196,
restoration............ 62, 85-86, 96, 98, 101-102 215-216, 233-234
retrieValiccenatt ee eee eee 13-14, 119-120 SNR Sec sico ieee cacneck acute 29,242
RGB....... 70, 74, 78, 81, 84, 86, 101, 105-106, SOR CODVibg nieeten eon ee oe 127-128
109-110, 202 SOlid: area sere eee 152-153, 232, 238

294
SOMCUK=|Clometantereetes ct hatteast eee 135 Rha SIMISSI ONIN Gee eaakercete scree et 40
SOUGHT Mecas tere ea tcdeunt 106, 203, 207, 248 NEAMSIMMANGO nea eens rat ee 196, 202
solvent evaporation uu... 158-159, 237 LAU SDALCMGY Bx.8 unary cee ae, 234, 272
SOIVENEMCTEN OMe cements. Me, hen eam ee 237 aD DING eee ee154-155, 235-236, 248-249
SOUTCORN eee 1os01f 00100, LUs1o. 00) Of. TWO: COMPONENL TONEKS 22a. <cctmens ee 183
spatial frequency .....34, 37, 94, 193, 206, 208, LVPOOKADMYCOCING exc: meee reece tes
242, 245
SHECUALSONSIMVIT Vries eneetik
vee ere 60, 87 U
SECU MIMO NG lb eeent ceeeee 73-74, 80 LGR eer 2 So ceees 2 ice en veto hae 104
SPECULAIEGIOSS eae en een sant eee 230 UIMEVGMESS re ee teeta eee 208, 236-237
UMIVGIANT cece nara nh ann Ome 32-33
Speed <2 ease.cc24526.02,00) 09) (0leoo, 111 COW ACUIGIINGame deerterete Oca cea tee eet cert 158
STOIC] Saree eee hae ec 139
S Ue N UlMiceeceeaeeet crcetartikaee een sateen x164 V
SUMIKOgtHOUG Geneara nena sk ere a, 234 Vall Giese ccs eens aca re ore 72-74, 101-102
subtractive color formation.............. 49, 80-81 variable information printing .............. 124, 144
SUDETSUMUCTURO es ce metsues matce nnn a eo 167 VOTMIISTING escent nates encom tee erties 159, 223
SUfaGe MON UMITORMMILY :secece cer eect a eePSSM| VECLOMCOCING saat ce erate han eee 54
surface reflection......... 32, 194-196, 198, 202, Vegetable Ollie isath. te tene teense
eeemeee 286
207, 210, 235, 243, 248, 274 WIG Oteatean 18, 16,25, 43, 92-53, 56,66, 237
surface reflection indicatrix..............c00008 195 VISCOClASTICIIV are aeA cco 278-279, 281
Surface tension.............0.... 137, 281-282, 284 VISCOSIVaeeacameca soe 147, 178, 187, 258, 269,
274-275, 277-279, 281, 286
i ViSWal CUalltWawte aes 32, 36, 68, 96, 98, 116
TACK Geese te At cssetiai a:154, 277-278, 280-281 VISUaIMCStS pi ee keer 32, 36, 68, 98, 102
PARFIMO GNIS Siege cemeteries eeeeicee acceso 205 ViSUaliZatlON csccsresesecseceeet ey IS Pees
LEX Ceeceees 13, 15-22, 29-30, 37-41, 53-54, 58,
62, 81, 86, 111-116, 137, 140, 174 W
ROXIO Uae crc ne reuerieten Somat, 39, 53, 114 WANS eto rnse cou Meee ane eae eee eee 42-43
ANS tMaN CVE TANSClccsseccreresercesstecseess 188-189 Wate lDaS@dilkSwauce. cance: seeeme pera 169
(UTS IRL. o]Ceereees see eee mere epee 186, 191 Wevelenguy SPECULUM mews emer 70, 74
thermal ink transfer...............cccce 189, 237 web break frequency .................. 2M 220) cer
thermal transfer......... 128, 136, 141, 144, 178, Web: DKGaKSie...crcene 217, 219, 221, 225-227
188-189 web fed printing ........ 130, 140, 160, 164, 166,
thermodynamic wetting.............ccccscsercsere 220 216, 2200220
TITCHITIOSCHSIIVES cet ey sy eets.rasesernee eee 178 web tension....... 130, 216-219, 221, 225-226,
TETEXOUNO Deere sere teste, carats eases: 275-277 Zoe 200200
ITOLATS)pCols. cep ae anceae eae See one er 2/0
Web LensiOniprOnleeay tre eee ee 218
WETZON SWC ler trees Mame ti caahiteaganmee Oey, ales)
LONE EDO UCTIONACUL VG ecrrs.aroscsccnvecesecsre: 267
WIniTEMESS* OR DaADel rr.ds.t cancs eee ste 205
LOC lrsctrcenaen 141, 178-184, 190, 193, 209,
WOOCMNEE DAD leisncd wncuseulcrcneterse: 138, 2116
200; 202
LOMEMOCVElODINGN bits a sree eceetre 141, 178
Y
TOL: Pe CUES Te] hess ee Renae me norariaattin
acecee 199
yellow..69-70, 72, 74, 80-81, 86, 88-90, 104,
TOUCHE DLOOMECSS ate nated eee eee 230
106, 155, 266, 273
EVANSTCECADACIY eer eacmer mince 39, 42
VIC CVAlUC eerste aenceoectre eee tout, PAhey OMT
{TANSOUSSION. Hoxecastiens 13-14, 22-23, 33, 40,
42-43, 49, 53-54, 74, 83, 87-89,
1055120) 272

295
 Ces
a de ee
= a

Economics of the Pulp and Paper Industry Papermaking, Part 3 Finishing edited by
edited by Magnus Diesen Mikko Jokio €
Forest Resources and Sustainable Management Pigment Coating and Surface Sizing of Paper pl
edited by Seppo Kellomaki edited by Esa Lehtinen icTy
Forest Products Chemistry edited by Per Stenius Paper and Paperboard Converting edited by eee ome
Papermaking Chemistry edited by Leo Neimo Antti Savolainen
Mechanical Pulping edited by Jan Sundholm Printing edited by Pirkko Oittinen and Hannu Saarelma
Chemical Puloing edited by Johan Gullichsen Process Control edited by Kauko Leiviska
st wo
Ns»)
ic
1
Lae] and Carl-Johan Fogelholm Materials, Corrosion Prevention and Maintenance
—
»e)
ed
wo
‘ Secondary Fiber and Deinking edited by edited by Jari Aromaa and Anja Klarin
N laa
wn
ON Heikki Pakarinen and Lothar Gottsching Paper Physics edited by Kaarlo Niskanen
iP)
N
wn NN
1 Papermaking, Part 1 Stock Preparation and Pulp and Paper Testing edited by Jan-Erik Leviin
YP)
Oo
0 aD Wet End edited by Hannu Paulapuro Paper and Board Grades edited by Hannu Paulapuro
Nn
=
oO
Papermaking, Part 2 Drying edited by Environmental Control edited by Pertti Hynninen
Oo wow
Markku Karlsson