How to Shape

Light with Spatial

Light Modulators
SECOND EDITION

Carmelo Rosales-Guzmán
and Andrew Forbes
 Structured Light with Spatial Light Modulators
by Carmelo Rosales-Guzmán and Andrew Forbes
doi: http://dx.doi.org/10.1117/3.100024
PDF ISBN: 9781510674875

Published by

SPIE Press
P.O. Box 10
Bellingham, Washington 98227-0010 USA
Phone: +1 360.676.3290
Fax: +1 360.647.1445
Email: [email protected]
Web: http://spie.org

Copyright © 2024 Society of Photo-Optical Instrumentation Engineers (SPIE)

All rights reserved. No part of this publication may be reproduced or distributed in any
form or by any means without written permission of the publisher.

This SPIE eBook is DRM-free for your convenience. You may install this eBook on any
device you own, but not post it publicly or transmit it to others. SPIE eBooks are for
personal use only; for more details, see http://spiedigitallibrary.org/ss/TermsOfUse.aspx.

The content of this book reflects the work and thoughts of the author(s). Every effort has
been made to publish reliable and accurate information herein, but the publisher is not
responsible for the validity of the information or for any outcomes resulting from reliance
thereon.

Spotlight vol. SL74

Last updated: 6 May 2024
Table of Contents
Preface vii

1 Introduction 1
2 Spatial Light Modulators 1
2.1 Brief introduction to liquid crystals 1
2.2 SLM technology 2
3 Getting Started 5
3.1 Brief introduction to SLMs 5
3.2 Calibration 10
4 Beam Shaping Techniques 11
4.1 Introduction to Beam Shaping 11
4.2 Phase modulation 13
4.3 Amplitude modulation 13
4.4 Complex amplitude modulation 14
4.5 Lossless shaping 17
4.6 Some examples of experimentally generated beams 17
4.6.1 Vortex beams 17
4.6.2 Bessel–Gauss beams 19
4.6.3 Finite-energy Airy beams 21
4.6.4 Hermite–Gauss beams 23
4.6.5 Laguerre–Gauss beams 25
4.6.6 Ince–Gauss beams 27
4.6.7 Mathieu–Gauss beams 30
4.7 Vectorial beams 33
4.7.1 Generalities about vector beams 34
4.7.2 Vector beams in cylindrical coordinates 35
4.7.3 Vector beams in elliptical coordinates 37
4.7.4 Other types of vector beams 37
4.7.5 Experimental generation of vector beams 38
5 Characterization of Scalar and Vector Beams 40
5.1 Characterization of scalar beams 40

iii
Table of Contents iv

5.1.1 Modal decomposition 40

5.2 Characterization of vector beams 43
5.2.1 Beam quality factor 43
6 Applications of SLMs 45
6.1 Digital propagation and focusing 45
6.1.1 Digital propagation 45
6.1.2 Digital focusing 46
6.2 Mode multiplexing 48
6.2.1 Choosing the mode set 48
6.2.2 Detecting the mode set 52
6.3 Wavefront reconstruction 54
6.4 Simulation of atmospheric turbulence 57

7 Generation of Holograms 61
7.1 Phase-only modulation 61
7.1.1 Vortex beams 61
7.1.2 Bessel beams 63
7.1.3 Airy beams 64
7.2 Complex amplitude modulation 65
7.2.1 HG beams and LG beams 65
7.3 Digital propagation and focusing 67
7.3.1 Digital propagation 67
7.3.2 Digital focusing 68
7.4 Simulation of turbulence 69
8 Appendix A: Hermite Polynomials 69
9 Appendix B: Laguerre Polynomials 70
Acknowledgments 70
References 71
 SPIE Spotlight Series

Welcome to the SPIE Spotlight series! This growing collection of concise eBooks
serves as an entry point for particular topics in optics and photonics suitable for
researchers, engineers, managers, executives, and educators. Spotlights fill the
community need for timely and relevant references at a level of detail bridging
the gap between in-depth journal articles and broad fundamental tutorials.
Whatever your interest or need, we hope this series meets your expectations and
encourage you to submit your own ideas for future Spotlights online.

Craig Olson, Series Editor

L3 Technologies

Associate Editors

Brian Sorg Richard Blaikie

National Cancer Institute University of Otago
Biomedical Optics/Medical Imaging Semiconductor, Nano-, and Quantum Technology

Keith Kasunic Wei Li

Optical Systems Group Stanford University
Aerospace and Defense Energy and the Environment

Ronian Siew
inopticalsolutions
Aerospace and Defense
Preface
Structuring light is a ubiquitous laboratory tool, with applications in fields as
diverse as optical communications, optical metrology, optical manipulation, and
high-resolution microscopy, among many others. Recent technologies, particu-
larly spatial light modulators (SLMs), have developed computer-controlled devi-
ces capable of reshaping an input beam into almost any desired output. To
facilitate an easy introduction to the use of such devices, as well as accelerating
the learning curve for more-experienced users, this self-contained Spotlight guides
readers from the basics to a higher level of knowledge regarding beam shaping
with SLMs. Many worked examples have been included to make this guide more
comprehensive and to facilitate first-time use. The minimum requirements are an
undergraduate level of mathematics and a basic knowledge of programming.
The examples are MATLAB based, with the code provided, and can be easily
adapted to other programming languages.

vii
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 1

1 Introduction
The shaping of light fields, a topic of interest to the optics community for some
time, has taken many forms, e.g., from coherent field mapping to diffusing ele-
ments for incoherent light shaping.1–5 Since the advent of the laser, structuring
laser light in amplitude and phase has been achieved outside the laser cavity (with
refractive, adaptive, and diffractive elements) and inside the laser cavity (through
a variety of amplitude or phase objects that force the laser internally to oscillate
on particular transverse modes). We will not discuss in detail the contribution of
these tools for controlling light, but the reader is referred to the references pro-
vided for further details. Instead, we will consider a modern derivative of the
above, namely shaping light with computer-generated holograms (digital holo-
grams) using spatial light modulators (SLMs).6 Digital holography for structured
light has enabled many new advances, ranging from classical to quantum physics,
including communication, microscopy, imaging, metrology, and education.6–12
The advent of liquid crystal on silicon (LCoS) SLMs has made the aforemen-
tioned techniques accessible to the inexperienced researcher. LCoS devices, collo-
quially referred to as SLMs, have allowed researchers to display computer-
generated holograms (CGHs) as images; thus, controlling light digitally can be
realized with just a little know-how.13–15 Here, we will show how to “get started”
with SLMs for the creation and detection of structured light fields.
This guide focuses on the shaping of coherent light with these tools. We out-
line the means by which one can get started with digital holography as well as
introduce phase-only, amplitude-only, and complex amplitude modulation as tools
to create structured light fields in the laboratory.

2 Spatial Light Modulators

2.1 Brief introduction to liquid crystals
The different states, or phases, in which matter can be found are related to the
mobility of individual atoms or molecules. The most evident are the solid, liquid,
and gaseous states. For solids, molecules are kept close together at a fixed posi-
tion and orientation, with a well-defined shape, due to intermolecular forces.
Though the molecules are packed together in liquids, they have more mobility
and will not form a rigid body but rather will take the form of the container. In
the gaseous state, the molecules have even more freedom to move around and will
always expand to fill the container holding them. As a result, their intermolecular
spaces will be larger. Though these three categories seem obvious, the boundaries
between the different phases are not well defined; also, there exist a large number
of other intermediate phases. This is the case of liquid crystals (LCs), which are
described by a state of matter at the interface between the liquid- and solid-state
phases. LCs are composed of elongated, rod-like, organic molecules that exhibit
unique properties from both liquids and solids. At high temperatures, LC
2 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 1 Schematic representation of different molecular alignments in a LC: (a) liquid,

(b) nematic, (c) smetic, and (d) cholesteric.

molecules will be randomly orientated as in the liquid state, as shown in Fig. 1(a),
but under appropriate conditions, all of the molecules will tend to align in a pre-
ferred direction maintaining a certain degree of motion. Moreover, due to their
elliptical shape, LC molecules form dipoles, which easily respond to external
stimuli, e.g., mechanical or electrical.
Based on the molecular order or organizational constraints, LCs can be classi-
fied as nematic, smetic, or cholesteric. In nematic LCs, the molecules are ran-
domly located but orientated in the same direction, as shown in Fig. 1(b). In the
smetic case, molecules exhibit certain degree of positional order so that their cen-
ters of mass are arranged in layers and their movement is mainly confined to the
plane defined by these layers [see Fig. 1(c)]. In cholesteric LCs, the molecules
are arranged in a helix-like layered structure. Consequently, the orientation of
the molecules in each layer is rotated (by a given rotation angle), producing a spi-
ral-like twisting of the molecular axis; therefore, these are also known as twisted
nematic (TN) LCs, as shown in Fig. 1(d). The cholesteric type was first observed
in cholesterol derivatives from which the name was borrowed. Because these
structures are chiral, this type of LC is also known as chiral nematic. The binding
forces that form the LC are stronger in certain crystalline directions, giving rise to
birefringent materials. This material has variations in the refractive index, which
is maximal along the long axis of the molecules, known as the slow axis. Due
to this variation, the phase velocity of a propagating wave will vary according
to its polarization and propagation direction.

2.2 SLM technology

An SLM is a device that can modulate or manipulate properties of light, such as
amplitude, polarization, and phase. SLM technology is based on LC properties,
which can be implemented either upon reflection or transmission. In essence, an
SLM is a pixelated display formed by millions of LC-filled cells that can be con-
trolled independently. The beauty of SLMs relies on the fact that, in principle,
they do not require the use of any specific software: one simply needs to connect
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 3

them to a computer and use them as a second monitor. Gray-tone images are dis-
played on this second monitor, where the grayscale (the color) of the image is
used to “control” (by dictating a desired rotation) the LCs in each pixel and there-
fore to modulate the phase of any light beam interacting with it. Beam shaping
relies on finding the appropriate image, a CGH (sometimes referred to as a digital
hologram), to produce the desired result. The following explains how to generate
digital holograms to create some of the most widely known beam shapes, but first,
we will give a few more details about LC-based SLMs.
Among the most common LC display types commercially available are those
that are electro-optical, in which the molecular alignment (and therefore
the refractive index of the material) is controlled by means of an electric field.
The transmissive types use transparent liquid-crystal displays (LCDs), whereas
the reflective type uses LCoS displays. Another important characteristic of
LCDs is the molecular alignment within the pixels of the microdisplay, which
are vertically aligned nematic (VAN), parallel aligned nematic (PAN), or TN.
The type determines how the incident light beam is manipulated: phase-only,
amplitude-only, or coupled amplitude and phase modulation.
In the PAN and VAN displays, the LC molecules are sandwiched between a
silicon pixel and a transparent electrode. The orientation of the LCs within each
pixel will then depend on the magnitude of the applied voltage. Under normal
conditions with no voltage, the molecules tend to line up in a preferred orienta-
tion, but in the presence of an electric field, the molecules tend to rotate by an
angle ðV Þ according to the strength of the applied voltage V . As a result, the
effective extraordinary index of refraction ne ðθÞ will decrease in proportion to V
according to

2
2
1 cos θ sinθ
¼ þ , (1)
ne ðθÞ
2 ne no

whereas the ordinary index of refraction no ðθÞ will remain constant. Modulation
of the voltage applied to each pixel controls the rotation angle of each molecule
and thus the refractive index of the pixel. The device is then calibrated so that
for V ¼ V min the rotation angle θ is minimum (ideally zero), and for V ¼ V max
the rotation is maximum. The device acts as a digital phase retarder that induces
a linear phase shift ΔΦðV Þ to an incoming wave. ΔΦ can be computed using
the Jones matrix formulation, according to which the matrix representation of a
phase retarder is as follows:



exp½iϕe  0 exp½iðϕe − ϕo Þ 0
M¼ ¼ exp½iϕo  , (2)
0 exp½iϕo  0 1
where ϕe ¼ kdne and ϕo ¼ kdno are the phase offsets of the electric field in
the direction of ne and no , respectively, and d is the thickness of the LC.
4 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Further, k ¼ 2π∕λ is the wave vector, and λ is the wavelength. Therefore, the
phase shift as a function of the voltage is
2π
ΔΦðV Þ ¼ ðn − no Þd: (3)
λ e
From Eq. (2), only one polarization component of the incident field will acquire a
phase shift.
̭
For
̭
example,
̭
a linearly
̭
polarized (LP) input field of the form
Ein ¼ Ee e þ Eo o, where e and o represent the unitary vectors along ne and no ,
respectively, will emerge from the LC, ignoring the constant term ϕo , as


E e exp½iðϕe − ϕo Þ
Eout ¼ M Ein ¼ , (4)
Eo
from which we can see that only the component of polarization parallel to the
extraordinary axis acquires ̭ a phase shift. For the extreme case, an input field with
linear polarization along o, E out will acquire a zero-phase shift. Note that the
amplitude of the input field remains the same.
The voltage, and hence the phase shift, can be controlled using grayscale
images (usually 8-bit encoding, allowing 256 colors). In this way, each gray level
is directly associated with a discrete increment of the phase from 0 (black) to 2π
(white), and a linear relation for the intermediate values. Therefore, 2π/256 phase
increments can be encoded at each pixel on the LC display by displaying an 8-bit
(256) pixelated image, which can easily be generated using computer software.
The number of pixels on a typical SLM screen would be on the order of
1080 × 1920.
Figure 2 shows a schematic structure of an LCoS-SLM display. A LC layer is
sandwiched between two transparent alignment films, glued to a transparent elec-
trode layer, and covered with a flat glass substrate. At the bottom lies a silicon

Figure 2 Schematic representation of an LCoS SLM display. Here, E represents the applied
electric field.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 5

substrate, and above is an active matrix circuit directly connected to pixelated

metal electrodes, which controls the orientation of the LC molecules at each pixel
(see also Ref. 16 for more details).

3 Getting Started
This section describes some well-known and lesser-known issues with SLMs.
Specifically, we outline how to overcome them so that one can “get started” with
the SLM as quickly as possible.

3.1 Brief introduction to SLMs

As previously stated, SLMs are pixelated devices consisting of LC-filled cells
(pixels) that are electrically adjusted with an applied voltage. The device is cali-
brated such that for a given applied voltage, the LC molecules rotate by a prede-
fined angle, resulting in a form of birefringent material for a particular orientation
of incoming polarized light. The result is a phase shift that is proportional to the
applied voltage for one component of the light’s electric field (which we shall
refer to as the horizontal polarization component). Today, with typical commer-
cial devices, a phase shift of at least 2π can be expected for a particular wave-
length (note that it can vary depending on the wavelength and the device). The
optical calibration of the device ensures that for a written color (usually with
8-bit encoding), the phase is changed by an appropriate amount. For example, if
a grayscale image is loaded with colors ranging from black (0) to white (255)
through 256 levels of gray, then the rotation of the molecules results in a phase
change of 0 through 2π rad for the incoming light, although modern devices can
reach much higher phase changes. It is common that the device would have a cal-
ibration file with it (gamma file); later we will explain how to perform the calibra-
tion from scratch.
Despite the ease of use, where one simply creates an appropriate grayscale
image to represent the desired phase change, getting started with most commercial
SLMs requires some knowledge. In this section, we outline a practical “how to”
approach to getting started with digital holograms written to SLMs.
To begin with, SLMs are not 100% efficient, even for the correct incoming
polarization state. This is a result of the pixelation: light is diffracted in many
directions (see Fig. 3) due to the two-dimensional (2D) amplitude grating formed
by the pixel array, while the spacing between the pixels reduces the fill factor of
the device. Once all the undesired orders are removed and only the zeroth order
is left, the shaped light is produced on-axis with the zeroth order (undiffracted
light). The latter is due to imperfect calibration, reflection off the electrodes, and
so on. Consequently, there is always a component of the light that is “undif-
fracted,” i.e., it does not interact with the CGH representing the desired phase
change. This component can be anywhere from 5% to 20%, depending on the
device. If this is not separated from the diffracted light, interference of the two
6 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 3 The light diffracted off the SLM has been shaped to carry OAM. The insets show
the diffraction orders due to the screen pixelation: (left) a long exposure of the screen and
(right) a CCD image of the reflected beam.

components can occur. This can be useful in some applications, e.g., in finding
the intermodal phases of some unknown light field and even for calibration,17
but in general it is not desired. To overcome this limitation, one always adds a
blazed grating to the calculated hologram. The blazed grating moves the dif-
fracted (first order) light away from the undiffracted (zeroth order) light at an
angle that is proportional to the spatial frequency of the grating (inverse of the
grating period). The transfer function of such a grating takes the form:

tðx, yÞ ¼ exp½i2πðGx x þ Gy yÞ: (5)

Here, the grating is in the diagonal direction, with a frequency of Gx ¼ 1∕Lx in

the x direction and Gy ¼ 1∕Ly in the y direction, where Lx and Ly are the periods
of the gratings in the two directions. The form of Eq. (5) is that of a blazed gra-
ting, meaning a linear phase ramp, but it need not be. Other examples might be
binary gratings or sinusoidal gratings. Blazed gratings have the advantage that
100% of the light is diffracted into the desired order, whereas binary gratings
would create several orders, with the first order containing just more than 40%
of the power. An example of such a diagonal blazed grating, as well as the impact
it has on the resulting beam, is shown in Fig. 4. As the grating period becomes
smaller (higher spatial frequency), so the position of the desired beam shifts far-
ther off-axis in the detection plane.
In this example, the desired hologram, shown in Fig. 4(a), imparts an azimu-
thal phase with a topological charge of three to the beam (see Section 4.6.1),
which alone would have a transfer function of
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 7

Figure 4 A vortex beam of order 3 is created without adding a grating (a), whereby the vor-
tex splits into three vortices of unit charge due to the overlapping Gaussian mode (b). When
the grating is added, as shown on the insets of panels (b), (c), and (d). The shaped light
(vortex beam) moves off-axis, away from the undiffracted Gaussian beam, as shown in pan-
els (f), (g), and (h). The result is a “clean” image of the shaped light because the unwanted
Gaussian beam can be removed by an aperture. The higher the grating modulation is, the
farther from axis the beam is moved.

t ¼ exp½i3ϕ: (6)

Because the initial Gaussian beam is still present, interference results in the split-
ting of the l ¼ 3 vortex into three l ¼ 1 vortices.17 By adding the grating, the
desired beam is moved away from the undiffracted Gaussian beam (zeroth order);
the final transfer function is then
t ¼ exp½ið2πGx x þ 2πGy y þ 3ϕÞ: (7)
The transfer function acts on the incoming beam’s electric field. The SLM
displays the modulated phase of this transfer function, namely,
ΦSLM ¼ arg½t ¼ mod½2πGx x þ 2πGy y þ 3ϕ, 2π: (8)
Thus, the final hologram on the SLM is a grayscale image corresponding to
Eq. (8), shown in Figs. 4(b)–4(d) for three examples of grating periods. The
superposition principle in optics allows the same concept to be applied to several
holograms programmed simultaneously: one can create multiple beams, each with
their own grating that directs the said beam to a particular location. The result is a
“multiplexed” signal as shown in Fig. 5. Here, a spatial carrier frequency in the
form of a blazed grating, shown in Fig. 5(a), is added to the desired hologram,
8 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 5 Multiplexing of many beam shapes, where each is projected to a different position
in the detector plane with the aid of several gratings, one for each beam shape. Each of the
three different holograms (b), (d), and (f) is given unique spatial carrier frequencies (a), (c),
and (e) to produce a single multiplexed hologram (g). The result, in the far field, is three pat-
terns corresponding to holograms (b), (d), and (f), each located at a unique point in the
detector plane (h).

shown in Fig. 5(b). This is repeated for the three illustrative holograms
[Figs. 5(a)/5(b), 5(c)/5(d), and 5(e)/5(f)] to form the final multiplexed hologram
[Fig. 5(g)]. Illuminating these holograms produces a superposition whereby each
mode is projected to a different location in the detection plane, usually taken to
be the Fourier or far-field plane, as shown in Fig. 5(h).
The grating together with the hologram must be resolvable by the number of
pixels and pixel spacing in the SLM. If not, not only will the efficiency of the
transformation decrease but also in extreme cases the quality of the transformation
will be degraded. To illustrate the problem, consider the images in Fig. 6: in
Fig. 6(a), the grating period is within the resolution (pixel spacing and size) in
the device, and thus the desired beam (in this case a vortex beam) is produced,
as shown in Fig. 6(c). In Fig. 6(b), the same hologram is programmed but with
a grating that has a spatial frequency that is too high for the pixel spacing and
size, resulting in a severely distorted “vortex” beam, as shown in Fig. 6(d).
Essentially, the Moiré effect from the pixel period and grating period manifests
as a real amplitude modulation of the output beam, resulting in severe distortions.
Over 100 such multiplexed modes can be written to a single hologram by care-
fully subdividing the screen and judicious choice of gratings.18
Secondary issues would include correcting the SLM’s optical distortions and
selecting a suitable graphic format to encode the hologram. In the case of the for-
mer, SLM devices are not optically flat and so a phase correction is required for
each device in order to remove optical aberrations due to the surface. Examples
of measured wavefront aberrations for two different devices are shown in Fig. 7.
Regarding the graphic format, some image file types introduce digital arti-
facts, as shown in Fig. 8, which can affect the quality of the generated beam.
Therefore, some formats are more suitable for digital holograms than others.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 9

Figure 6 (a) An azimuthal hologram together with a low grating frequency produces a “neat”
forked grating; (b) the same azimuthal hologram but with a grating frequency that is too high
produces a Moire pattern. The consequences of these holograms are shown in panels
(c) and (d). (c) The beam produced is a high-quality Laguerre mode as desired, and (d) a
distorted “petal” structure is seen.

Figure 7 (a) and (b) Measured wavefront aberrations due to imperfections for two SLM sur-
faces. The wavefront error differs from device to device. These errors can be corrected by
applying a conjugate phase to the SLM.

Figure 8 One can observe additional structure in the jpeg image in comparison to the bmp
image. This can affect the quality of the generated beams in some cases and, conse-
quently, graphic quality is important.
10 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

3.2 Calibration
SLMs need to be calibrated. This simply means that one must adjust the applied
voltages to realize the desired phase change when the digital hologram is dis-
played. This is easily done with a simple lens, double pinhole or slit, and a cam-
era. The idea is shown in Fig. 9.
A uniform-amplitude “plane wave” is formed by expanding the initial test
beam (of the desired wavelength) through a beam-expanding telescope, taking
care not to introduce appreciable curvature on the wavefront. Next, the near-
collimated and near-uniform field is passed through a double-pinhole (or double-
slit) screen—the mask. The central region of this beam is usually selected to
ensure a uniform intensity, i.e., the beam may be much larger than the mask.

Figure 9 (a) Experimental setup for the calibration of an SLM. A mask is used to send two
equally weighted “plane waves” to the SLM screen. After reflection, the two beams are com-
bined to form an interference pattern (inset) at the detector. (b) As the phase on one side of
the SLM is changed (grayscale from 0 to 255), so the interference pattern shifts. From the
measured shift, the SLM can be calibrated.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 11

Note that this has to be a physical screen; trying to implement it digitally will not
work. The first part of this setup produces two identical beams and may be
replaced with any other setup that does this. Next, the two beams that emerge
are bounced off two parts of the SLM, each programmed with its own constant
phase screen, and then passed through a lens where they spatially overlap in the
far field and are recorded by a charge-coupled device (CCD) camera (or simple
webcam). Fringes appear in this plane due to interference, the position of which
is dependent on the relative phase difference between the two waves, which, in
turn, is determined solely by the phases programmed onto the two parts of the
SLM. As the phase on one screen is changed, so the fringe pattern at the detector
shifts laterally. The movement can be used to determine the change in phase as a
function of the hologram’s grayscale color. With this value measured, one can
either calculate the correction response manually or, as in the case of some man-
ufacturers, an automated process does this for you. Rather surprisingly, if you
elect to work in the first diffraction order with a blazed grating, then it is not nec-
essary to calibrate the SLM at all. In fact, the SLM will have the same response
across all wavelengths with only a minor amplitude effect.19
Once these effects are taken into account, and assuming that the device has
been correctly calibrated, the SLM can now be used to display CGHs digitally.
These are nothing more than grayscale pictures that execute the desired transfor-
mation of the incoming beam to produce the desired output beam. Section 4
describes how to shape light in amplitude and phase with digital holograms, i.e.,
how to calculate what the picture needs to look like.

4 Beam Shaping Techniques

4.1 Introduction to beam shaping
This section examines how to create custom modes of light, so-called structured
light, with digital holograms. In general, this requires a transmission function with
both phase and amplitude modulation. This is because most desired light fields
have both amplitude and phase terms, thus requiring complex amplitude modula-
tion to convert one beam’s amplitude and phase into that of another. If your
device already allows this, then your task is easier. Most modern SLMs are
phase-only devices, but fortunately, there are several techniques to convert from
a phase-only device response (which traces out a unit circle in the complex plane)
to a full amplitude–phase response (which represents a filled circle in the complex
plane). The problem is shown in Fig. 10 and has been addressed by many
approaches,20–32 with a review given in Ref. 6. We will not delve into these
approaches with great detail except to say that they exist and may be readily
applied for the creation of arbitrary structured light fields. Instead, we highlight
the basic concepts and how they may be applied quickly. All of the code to do
so is included in Appendices A and B.
12 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 10 Moving about the complex plane, from (a) a phase-only response that traces out
a unit circle, to (b) a full amplitude and phase response that fills the complex plane.

Figure 11 (a) When the field is modulated in amplitude and phase, the new field is created in
the near field, at the plane of the SLM. To measure it, one often uses a telescope to relay the
image of the SLM plane to the detector plane (screen) because it is usually not possible to
get close enough to the SLM with detectors due to space considerations. This has the advan-
tage that the desired diffraction order can be selected at the Fourier plane with an aperture,
removing the unwanted orders (higher or lower, depending on the coding scheme). (b) When
a phase-only shaping approach is used, the desired field is created in the far field. This can
be detected by placing the detector (screen) very far from the SLM, or as is more practical, at
the focal plane of a lens after the SLM (the Fourier plane). Here, f is the focal length of the lens.

Consider an optical field written in terms of amplitude and phase

u1 ¼ u01 expðiϕ1 Þ. If this beam were to be transformed into a new desired field,
e.g., u2 ¼ u02 expðiϕ2 Þ, then it is clear that the optical element executing this
change must have a transfer function given as
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 13

u2 u02
t¼ ¼ exp½iðϕ2 − ϕ1 Þ: (9)
u1 u01
There are some implicit consequences for this, which we discuss in detail in sec-
tions 4.3 and 4.4. Here, we will restrict ourselves to handling holograms for
phase-only devices, since this is now ubiquitous. However, it is possible to per-
form all such modulations on amplitude and phase devices (see, e.g., Ref. 29).
One important consideration is where to detect the desired field. As we will
show with numerous examples, complex amplitude modulation, as given in Eq.
(9), means that the desired beam is created at the plane of the SLM. For practical
reasons, one usually employs a telescope to relay this plane to a detector plane
(screen), as shown in Fig. 11(a). When only the phase is modulated, the plane
at the SLM does not yet have the desired beam; instead, the desired beam is found
in the far field of the SLM, which can be measured by placing the detector
(screen) at the focal plane of a lens after the SLM, see Fig. 11(b).

4.2 Phase modulation

When u01 = u02, the amplitude term of the transmission function is 1, indicating a phase-
only response is needed. For example, if the beam is merely redirected (tip/tilt) then a
phase-only blazed grating is needed; if the beam is to be focused, then a lens term is
needed; and if the beam is to be given orbital angular momentum (OAM), then an azimu-
thal phase is needed. These are all phase-only holograms. The phase programmed on the
hologram is simply mod½ϕ2 − ϕ1 , 2π. Often, the incoming phase is flat or an arbitrary
constant, in which case the hologram is simply the modulus of the desired phase
mod½ϕ2 , 2π. We will show some examples of this in Sections 4.6.1, 4.6.2 and 4.6.3.

4.3 Amplitude modulation

Modulation of both phase and amplitude can be achieved by what is known as
complex amplitude modulation of the input beam. It should be stressed that even
though the outcome is a desired amplitude change, the holograms themselves, by
definition, only change the phase of the light. Moreover, the SLM cannot create
light, so only a decrease in amplitude can be realized. This is possible by distrib-
uting the light away from the desired first diffraction order in a manner that is
spatially dependent. There are only two places for the undesired light to go: either
into the zeroth order or the higher diffraction orders. In the former, the depth of
the hologram is decreased (in a spatially varying manner),32 whereas in the latter,
the spatial frequencies are increased across the beam to increase the local diffrac-
tion angles (also in a spatially varying manner), e.g., as in Ref. 31.
Consider the simplest example of decreasing the phase depth locally across
the hologram. When the phase depth of the hologram at some position is changed
from 2π to 2πM, the efficiency at that position changes to
14 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 12 A linear phase ramp of depth 2π directs 100% of the light in the desired first dif-
fraction order (purple). If the ramp depth is zero (no ramp), then all of the light passes
through undiffracted, i.e., 0% in the first diffraction order (orange). A value between the
two extremes results in some light being diffracted into the first order and some not, i.e.,
amplitude modulation of the input light.

jC n j2 ¼ sinc2 ½πðn − M Þ, (10)

where n is the diffraction order and jC n j2 is the fraction of power in the n’th
order. For example, if M ¼ 0 (no grating depth) then 100% of the light is in the
zeroth order and none is in the first order. If M ¼ 1, then 100% of the light is
in the first order (our desired order). At a depth of πðM ¼ 0.5Þ, the desired first
order only has 41% of the light. It is clear that by inverting Eq. (10), any desired
amplitude modulation can be achieved. By first calculating the desired phase and
then modulating the depth according to Eq. (10) for each position (x, y), one can
realize spatially dependent amplitude and phase control on a phase-only device.
This principle is shown as an example of a grating in Fig. 12.

4.4 Complex amplitude modulation

As mentioned before, it is possible to change the amplitude and phase using com-
plex amplitude modulation. In most approaches, a high spatial frequency is added
to the field to produce the desired amplitude and phase. Because the spatial fre-
quency is high, the undesired light is found in the higher diffraction orders, as
shown in the bottom-right inset of Fig. 13(b), which can be filtered out by placing
an aperture between two telescope lenses, as shown in Fig. 11(a).
To see how this works, consider the complex plane diagrams in Fig. 13. It is
possible to realize any position on the unit circle in the complex plane since these
are phase-only values and therefore “allowed” by the SLM. The effect of beating
between two such values at positions A and B is to produce a new amplitude and
phase that is not on the unit circle, i.e., a non-unit amplitude with any desired
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 15

Figure 13 (a) A pure phase modulation is equivalent to tracing a unit circle in the complex
plane. To introduce amplitude modulation, the radius of the circle must be allowed to take
on values smaller than 1. This is equivalent to moving off the unit circle to any point inside
the unit circle. One example would be to modulate between two points on the circle
(A and B). The result of which would be a point inside the circle (star). (b) Measured ampli-
tude modulation using this approach is shown with the insets showing an example “checker-
board” as well as an experimental image of the resulting beam (Bessel beam in this case).
Note the unwanted light in the four diffraction orders at the edges.

phase. Say the desired complex value is t ¼ jtjeiϕ ; the two phase values that
would give the desired complex value inside the unit circle are given as
A ¼ exp½iϕ þ iα, (11)

B ¼ exp½iϕ − iα, (12)

with
jtj ¼ cosðαÞ: (13)
16 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

The idea is to create a “checkerboard” of such phases, has given by Eqs. (11)
and (12), to create the desired effect, 31 as illustrated in the top-left inset of
Fig. 13(b). Such a patterned grid should be at least 5 × 5 pixels, or more, to be effec-
tive. Thus, the ability to control amplitude comes at the expense of spatial resolution.
There are several approaches to complex amplitude modulation.17,18 For
example, consider a field of the form:
Uðx, yÞ ¼ Aðx, yÞ exp½iϕðx, yÞ, (14)
where Aðx, yÞ is the amplitude with values in the interval [0,1] and ϕðx, yÞ the
phase with values in ½−π, π. The goal is to rewrite Eq. (14) in a manner that
incorporates amplitude variations as phase variations, that is, a function hðx, yÞ
must be found such that
hðx, yÞ ¼ exp½iΨðA, ϕÞ, (15)
where ΨðA, ϕÞ takes into account both amplitude and phase variations. This
reduces to finding a phase function of the form ΨðA, ϕÞ. One way to do that is
by expressing hðx, yÞ as a Fourier series in the domain of Ψ, that is,
X
∞
hðx, yÞ ¼ cAq exp½iqϕ: (16)
q¼−∞

To recover the field U ðx, yÞ, only the first-order term of Eq. (16) is needed,
provided

cA1 ¼ Aa, (17)

where a is a positive constant. Moreover, for functions ΨðA, ϕÞ with odd sym-
metry, ΨðA, ϕÞ can be written as follows:
ΨðA, ϕÞ ¼ f ðAÞ sinðϕÞ: (18)

By rewriting hðx, yÞ ¼ exp½if ðAÞ sinðϕÞ as a Fourier series using the Jacobi–
Anger identity, this is,
X
∞
exp½if ðAÞsinðϕÞ ¼ J m ½f ðAÞ exp½imϕ, (19)
m¼−∞

where Jm is the m’th-order Bessel function; and by combining Eqs. (16), (17), and (19):

cA1 ¼ J 1 ½f ðAÞ: (20)

Therefore, from Eqs. (17) and (20)

Aa ¼ J 1 ½f ðAÞ: (21)
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 17

The function f ðAÞ can be obtained by numerical inversion of the above equation.
The maximum value of a for which this is true is 0.58, corresponding to the maxi-
mum value of the first-order Bessel function J 1 ðxÞ, that occurs at x0 ¼ 1.84. Thus,
f ðAÞ is restricted to the interval ½0, 1.84. The limiting values of f ðAÞ allow one to
implement this technique in the reduced domain ½−1.17π, 1.17π. This approach
has the added advantage of allowing beam shaping on SLMs when the possible
phase shifts are limited (not the full optical cycle of 2π). In modern-day devices,
it is possible to exceed 2π modulations, in which case free-form optics can be
approximated on SLMs.33

4.5 Lossless shaping

In many instances, the phase of the final field is not required to be defined
because only the intensity of the light is important, e.g., laser materials processing
where perhaps a flat-top intensity profile is desired at the target regardless of the
phase. If this is the case, then even though the transmission function requires an
amplitude change; it may be done by a phase-only transformation in a lossless
manner by leaving the phase in the target plane as a free variable. In this
approach, the initial phase is designed to create the desired intensity distribution
by interference of all the component waves that make up the initial field. In a
sense, nature converts the phase changes into intensity changes through diffrac-
tion. The required phase of the first element can be calculated analytically by
using the stationary phase approximation1 or numerically by an iterative approach.
However, the phase of the beam at the target plane is arbitrary and usually rather
complicated. If the phase is also desired to be well defined then, a second phase-
only element can be used to correct for the arbitrary phase at the plane of the
desired intensity: the conjugate of the phase at the target will result in a beam with
a flat wavefront. In such a scheme, two SLMs (or two reflections off one) are
used to execute amplitude and phase control in the target plane in a manner that
is, in principle, lossless (see, for example, Ref. 34 for this approach applied to
OAM detection). By exploiting this principle further, it is possible to structure
arbitrary amplitude, phase, and polarization states of light.

4.6 Some examples of experimentally generated beams

4.6.1 Vortex beams
One of the simplest modulations we can impose on a Gaussian beam is a transver-
sally varying azimuthal phase. This phase type generates a so-called vortex beam.
Incidentally, these beams carry a quantized amount of OAM of lℏ per photon,
where l is the topological charge of the vortex (how many twists it has about
the axis). The azimuthal phase is undefined at the origin and thus generates a
phase singularity along the propagation axis with an associated intensity null.
Such fields are also referred to as twisted beams due to their helical phase fronts.
Mathematically, such a twisted light beam can be described as5,34
18 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 14 (a) Intensity and (b) phase profiles of a vortex beam.

Eðρ, φÞ ¼ E0 exp½ilϕ, (22)

where ϕ is the azimuthal angle and, l ∈ Z is the topological charge. Figures 14(a)
and 14(b) show the intensity and phase profiles for fields with various topological
charges, respectively.
The experimental generation of such vortex beams can easily be accom-
plished with an SLM by encoding an azimuthal variation (of modulus 2π) and a
blazed grating to separate the first diffraction order from the rest. The final
expression takes the form

ΦSLM ¼ mod½lϕ þ 2πðGx x þ Gy yÞ, 2π, (23)

where mod is the modulus function, and Gx and Gy are the grating frequencies
along the x and y directions, respectively. Figure 15 shows the holograms dis-
played on the SLM [Fig. 15(a)] along with the corresponding measured intensity
profiles [Fig. 15(b)]. Because this is a phase modulation, the detection was done
in the far-field plane of the SLM, at the focal plane of a Fourier transforming lens,
as shown in Fig. 11(b). As can be seen, this phase-only modulation often produ-
ces undesired secondary rings because a pure azimuthal phase without a radial
intensity dependence is not an eigenmode of free-space (see Section 4.6.5). In
fact, these “vortex” beams are best described by hypergeometric beams,35,36 with
very little energy deposited into the desired vortex (without rings). This can be
corrected by full amplitude and phase control to remove the undesired rings (addi-
tional radial modes) to produce an ideal vortex beam.36 The MATLAB code
vortex.m, available online, can be used to generate these holograms.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 19

Figure 15 (a) Holograms encoded on the SLM and (b) experimentally measured intensity
profile of vortex beams.

4.6.2 Bessel–Gauss beams

Bessel beams have a transverse intensity profile based on the family of Bessel
functions. A zero-order Bessel beam has a transverse intensity profile with a
bright central core surrounded by bright concentric rings. Unlike Gaussian beams,
the transverse intensity profile of a Bessel beam does not spread as the beam
propagates over a finite distance.37,38 One of the most notable qualities of the
Bessel beam is that it can reconstruct itself around obstructions placed in its path.
This property makes the beam useful for stacking multiple objects along the
beam’s central core, for example, in an optical trapping experiment,39 and even
recovering quantum entanglement after obstacles.40
Bessel beams are exact solution to the free-space Helmholtz wave equation in
cylindrical symmetry and are mathematically described as37
Eðρ, φ, zÞ ¼ E0 J l ðk t ρÞexp½ik z z exp½ilϕ, (24)
where J l ðxÞ are the Bessel functions of order l, while k t and k z are the transverse
and longitudinal components of the wave vector k, respectively, obeying the rela-
tion jkj2 ¼ k 2t þ k 2z and jkj ¼ k. Figures 16(a) and 16(b) show the intensity and
phase patterns of these beams, respectively, consisting of an infinite number of
concentric rings that cover an infinite area, resulting in the beam carrying an
infinite amount of energy. Physically this is not possible; therefore, experimen-
tally, we generate an approximation of these beams known as Bessel–Gauss
(BG) beams given, at z ¼ 0, by the expression38

2 
ρ
Eðρ, ϕ, z ¼ 0Þ ¼ J l ðk t ρÞ exp − exp½ilϕ: (25)
ω0
20 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 16 (a) Intensity and (b) phase profile of Bessel beams.

There are two common approaches to generate BG beams: near field and far
field. The first uses a conical lens known as an axicon, whereas the second uses
a ring aperture. Even though both can be encoded as a hologram on an SLM,
the first is more efficient. The transfer function of an axicon is given as
tðρ, φÞ ¼ exp½ik t ρ: (26)
The transverse component of the wave vector k t can be expressed in terms of the
angle αax of the axicon (the apex angle of the conical prism) as
k t ¼ αax ðnR − 1Þk, (27)
where nR is the refractive index of the axicon. The BG beams generated in the
laboratory have a finite propagation distance given as37
k
zmax ¼ ω0 : (28)
kt
The transfer function of a high-order Bessel beam can be obtained by adding the
term exp½ilφ to Eq. (26), which after substitution of Eq. (27) takes the form
tðρ, ϕÞ ¼ exp½ikαax ðnR − 1Þρ exp½ilϕ: (29)

Thus, the mathematical expression to generate a hologram with the above transfer
functions takes the form
ΦSLM ¼ mod½kαðnR − 1Þρ þ lϕ þ 2πðGx x þ Gy yÞ, 2π: (30)
Figure 17(a) shows four digital holograms for an axicon with refraction index
nR ¼ 1.5 and topological charge values l ¼ 0, 1, 4, 5. The MATLAB code
Bessel.m, also available online, will reproduce similar images. The experimental
intensity profiles measured for these holograms are shown in Fig. 17(b).
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 21

Figure 17 (a) Holograms encoded on the SLM and (b) experimentally measured intensity
profiles of BG beams.

Using these holograms as the basis to build arbitrary superpositions allows

unprecedented control over the propagation dynamics of the resulting beam,
e.g., to create rotating,41 accelerating,42 and frozen beams.43

4.6.3 Finite-energy Airy beams

A first solution to the paraxial wave equation is given in terms of the Airy functions,
so-called Airy beams. The mathematical expression for these beams is given as44

2  
2  

ξ ξ iξ ξ3
Aðsx , sy , ξÞ ¼ Ai sx − Ai sy − exp s þ sy − , (31)
2 2 2 x 3
where Ai½x is the Airy function, sx ¼ x∕x0 and sy ¼ y∕y0 represent dimensionless
transverse coordinates, x0 and y0 are the transverse scale parameters, ξ ¼ z∕kx20 is
a normalized propagation distance, and k ¼ 2π∕λ is the wavevector. A very good
approximation to the ideal Airy beam can be realized in the optical regime by impos-
ing some restrictions to Eq. (31), namely
Aðsx , sy , ξ ¼ 0Þ ¼ Ai½sx Ai½sy  exp½bðsx þ sy Þ, (32)
where b is a positive parameter, typically smaller than 1, that limits the energy of
the Airy beam. The final expression for this “finite-energy Airy beam” takes the
following form:44

2  
2 
ξ ξ
A0 ðsx , sy , ξÞ ¼ Ai sx − þ ibξ Ai sy − þ ibξ
2 2
 
ξ 3 iξðsx þ s y Þ
× exp bðsx þ sy Þ − bξ2 þ ibξ2 − þ : (33)
6 2
22 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 18 (a) Phase and (b) intensity profiles of Airy beams for different values of b.

As a result of the finite energy, this beam maintains its nondiffraction properties
over a finite distance only. Figure 18 shows a theoretical plot of the phase along
with the intensity for the example case with parameters x0 ¼ y0 ¼ 3 mm for dif-
ferent values of b.
The experimental generation of these beams can be achieved by encoding a
hologram corresponding to the inverse Fourier transform of A0, which produces
a cubic phase of the form
 3 3 Þ
iðk x þ k y
FfA0 ðsx , sy Þg ∝ exp½−bðk 2x þ k 2y Þ exp , (34)
3
where k x and k y are the transverse components of the inverse Fourier transform,
denoted by FfA0 ðsx , sy Þg of the finite-energy Airy beam. Note that this expression
involves the product of a Gaussian beam [the first term in Eq. (34)] and a cubic
phase (second term). Therefore, to generate an Airy beam, a hologram with a
cubic phase should be illuminated by a Gaussian beam and Fourier transformed
with a lens placed in front of the SLM at a distance equal to the focal distance
f of the lens, as shown in Fig. 19.
In this case, the phase encoded on the SLM takes the following form:
 
X3 þ Y3
ΦSLM ¼ mod , 2π : (35)
3

Figure 19 Experimental setup to generate an Airy–Gauss beam. A hologram with a cubic

phase is encoded on the SLM, which is illuminated with a Gaussian beam. An Airy beam
will be observed in the far field, and the focal plane of a lens with focal length f .
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 23

Figure 20 Hologram displayed on the (a) SLM to generate (b) finite-energy Airy beams.

Note from Eq. (35) that it is not possible to change the parameter b, thus it is nec-
essary to change the size of the input Gaussian beam instead. Figure 20 shows the
encoded phase ΦSLM given by Eq. (35), generated with the MATLAB code
Airy.m, available online. The experimentally measured Airy beams for three dif-
ferent sizes of input Gaussian beam are also shown here.

4.6.4 Hermite–Gauss beams

The Hermite–Gauss (HG) modes are a set of solutions to the paraxial wave equa-
tion in Cartesian coordinates. Their mathematical representation is given in terms
of a Gaussian function and the Hermite polynomial H n ðxÞ as45
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi



1 2ð1−n−mÞ x y
HGnm ðx, y, zÞ ¼ H Hm
ωðzÞ πn!m! n ωðzÞ ωðzÞ

2   
ρ ikρ2
× exp½iðn þ m þ 1ÞξðzÞ exp − exp − exp½−ikz, ð36Þ
ωðzÞ 2R
where n and m are positive integers and λ is the wavelength. In addition,
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ρ ¼ x2 þ y2 , (37)

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

2ffi
z
ωðzÞ ¼ ω0 1 þ , (38)
zR


2 
z
RðzÞ ¼ z 1 þ R , (39)
z
rffiffiffiffiffiffiffi
λzR
ω0 ¼ z , (40)
π
24 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 21 (a) Intensity and (b) phase profile of HG modes.


z
ξðzÞ ¼ arctan : (41)
zR

Equation (36) represents a paraboloidal wave with radius of curvature R(z), beam
waist ω0 , and beam size ωðzÞ. The term zR is a constant known as the Rayleigh
range, which is used to measure the distance over which the beam remains colli-
mated. The parameter ξðzÞ, known as the Gouy phase, is an additional phase shift
that the wavefront acquires upon propagation through the beam waist.
Figures 21(a) and 21(b) show the intensity and phase profile of the HG beams:
HG01 , HG20 , HG31 , and HG23 , respectively.
These beams can be generated by employing complex amplitude modulation;
therefore, we will use them as a first example of this method. According to
Eq. (36), the amplitude term is given as
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi


 

1 2ð1−n−mÞ x y ρ 2
AHG ðx, y, zÞ ¼ H Hm exp − ,
ωðzÞ πn!m! n ωðzÞ ωðzÞ ωðzÞ
(42)
while the phase term is given as
 
ikρ2
ΦHG ðx, y, zÞ ¼ exp½iðn þ m þ 1ÞξðzÞ exp − exp½−ikz: (43)
2R

Thus, the encoded hologram would take the form

ΦSLM ¼ f HG sin½ΦHG þ Gx X þ Gy Y , (44)
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 25

Figure 22 (a) Complex amplitude modulation holograms encoded on the SLM and
(b) experimentally measured intensity profile of HG modes.

where f HG is the amplitude function and can be found by numerical inversion of

the following relation:
J 1 ½ f HG  ¼ AHG : (45)
Figures 22(a) and 22(b) show four examples of the holograms generated through
complex amplitude modulations, encoded on the SLM and their corresponding
measured intensity distribution, respectively. The MATLAB code Hermite_
Gaussian.m to generate these holograms is also available online.

4.6.5 Laguerre–Gauss beams

Laguerre–Gauss (LG) beams arise naturally as another solution to the paraxial
Helmholtz equation but cylindrical coordinates. Mathematically, LG are described
by the expression45
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffi  

ω 2p! 2ρ jlj l ρ 2
LGlp ðρ, φ, zÞ ¼ 0 Lp 2 exp½−ikz
ωðzÞ πðjlj þ pÞ! ωðzÞ ωðzÞ
× exp½ið2p þ jlj þ 1ÞξðzÞ (46)

  
ρ 2 ikρ2
× exp − exp − exp½−ilϕ,
ωðzÞ 2R
where Llp are the generalized Laguerre polynomials and the rest of the parameters
are the same as in the HG beams. Figures 23(a) and 23(b) show the intensity and
phase profile of the LG beams, respectively: LG02 , LG21 , LG12 , and LG22 .
26 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 23 (a) Intensity and (b) phase profiles of LG beams.

These beams can also be generated with SLMs, and they are approximated
using complex amplitude modulation. In this case, the amplitude is given as
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffi  "
# "
#
ω 2p! 2ρ jlj l ρ 2 ρ 2
ALG ¼ 0 Lp 2 exp − ,
ωðzÞ πðjlj þ pÞ! ωðzÞ ωðzÞ ωðzÞ
(47)

and the phase by

 
ikρ2
ΦLG ¼ exp½ið2p þ jlj þ 1ÞξðzÞ exp − exp½−ilϕ: (48)
2R
Thus, we encode on the SLM a hologram of the form

ΦSLM ¼ f LG sin½ΦLG þ Gx x þ Gy x, (49)

where f LG is obtained by numerical evaluation as follows:

J 1 ½f LG  ¼ ALG : (50)

Figures 24(a) and 24(b) show four examples of the holograms encoded on the
SLM, generated through complex amplitude modulation and their corresponding
measured intensity distribution, respectively. The MATLAB code to generate
these holograms (Laguerre_Gaussian.m.) is also available online.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 27

Figure 24 (a) Complex amplitude modulation holograms encoded on the SLM and
(b) experimentally measured intensity profiles of LG beams.

4.6.6 Ince–Gauss beams

The Ince–Gauss (IG) beams form an infinite set of elliptically shaped beams, a set
of solutions of the paraxial wave equation in elliptical cylindrical coordinates
r ¼ ðζ, η, zÞ.46 Here, ζ ∈ ½0, ∞Þ and η ∈ ½0, 2π are the radial and angular elliptical
coordinates, respectively, related to the Cartesian coordinates ðx, yÞ as
sffiffiffiffiffiffiffiffiffiffiffiffiffi
2
x¼ cosh ζ cosh η, (51)
εω2 ðzÞ

sffiffiffiffiffiffiffiffiffiffiffiffiffi
2
y¼ sinh ζ sinh η: (52)
εω2 ðzÞ

The parameter ωðzÞ is the beam size [Eq. (38)] and the parameter ε ¼ 2f 0 ∕ω20
∈½0, ∞Þ is the ellipticity of the beams. In this case, two types of solutions are
obtained, which are given in terms of even and odd Ince polynomials and are
q ð· , εÞ and S q ð· , εÞ, respectively. The set of beams obtained in both
denoted by C m m

cases are known as even and odd IG beams and given mathematically as46
 
Cω0 μ μ ρ2
IGeq,μ,ε ðrÞ ¼ C q ðiζ, εÞC q ðη, εÞ exp − 2
ωðzÞ ω ðxÞ
 2

ikρ
× exp½−iðkzÞ exp − exp½ðq þ 1ÞξðzÞ, (53)
2R
28 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

 
Sω0 μ μ ρ2
IGoq,μ,ε ðrÞ ¼ S q ðiζ, εÞS q ðη, εÞ exp − 2
ωðzÞ ω ðxÞ
 
ikρ2
× exp½−iðkzÞ exp − exp½ðp þ 1ÞξðzÞ, (54)
2R
where C and S are the normalization constants and the superscripts e and o stand
for even and odd, respectively. Further, the indices q, μ ∈ Zþ obey the relations
0 ≤ μ ≤ q for even beams and 1 ≤ μ ≤ q for odd beams. Figures 25(a) and
25(b) show the intensity and phase profile of odd IG beams of ellipticity ε = 2,
respectively. Note the π difference between consecutive intensity lobes.
In an analogous way, Figs. 26(a) and 26(b) show the intensity and phase pro-
file of even IG beams of ellipticity ε ¼ 2.
Of particular interest is the characteristic factor ε, which allows a smooth
transition between the LG (ε ¼ 0) and HG (ε ¼ ∞Þ beams. In the former, the
indices ðμ, qÞ are related to the indices ðp, lÞ through the relations l ¼ μ and
p ¼ ðq − μÞ∕2. In the latter, the indices of the even IG beams are related to
those of the HG by the relations n ¼ μ and m ¼ q − μ, whereas the odd through
the relations n ¼ μ − 1 and m ¼ q − μ þ 1. By way of example, Fig. 27 shows
the transition of an odd IG beam of parameters μ ¼ 5 and q ¼ 7, from an
LG to an HG beam, the corresponding ellipticities are, from left to right,
ε ¼ ½0, 1, 5, 100].
The superposition of even and odd IG beams with an intramodal phase
expðiπ∕2Þ gives rise to the helical IG (HIG) beams with a positive or negative hel-
icity, which are described mathematically as

Figure 25 Odd IG beams with ellipticity ε = 2. (a) Intensity and (b) phase profile.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 29

Figure 26 Even IG beams with ellipticity ε = 2. (a) Intensity and (b) phase profile.

Figure 27 Odd IG modes with varying ellipticity ε, from 0 to 100, evincing the transition from
LG to HG. Transverse (a) intensity and (b) phase profile.

q,μ,ε ðrÞ ¼ IGq,μ,ε ðrÞ  iIGq,μ,ε ðrÞ:

IGh e o (55)
A particular example of such superposition of even and odd IG modes with pos-
itive and negative helicities is shown in Fig. 28 for the specific case q ¼ 5,
μ ¼ 3, and ε ¼ 0.2:
Figure 29(a) shows four examples of the holograms encoded on the SLM that
generate the IG beams shown in Fig. 29(b). The MATLAB code to generate these
holograms can be downloaded directly from MathWorks in Ref. 47.
30 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 28 HIG beam obtained as the superposition of even and odd IG beams. (a) Intensity
and (b) phase distribution.

Figure 29 (a) Holograms encoded on the SLM, through complex amplitude modulation, and
(b) simulated intensity profiles of different types of IG beams.

4.6.7 Mathieu–Gauss beams

Another interesting set of beams are the Mathieu–Gauss (MG), which are
obtained as solutions to the Helmholtz equation in elliptical cylindrical coordi-
nates. 48 Here, the elliptical coordinates are related to the Cartesian by the
relations
x ¼ f cosh ζcosη, (56)

y ¼ f sinh ζsinη, (57)

Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 31

where again ζϵ½0, ∞Þ is the radial coordinate and ηϵ½0, 2π the angular. In addi-
tion, the semifocal distance f is given in terms of the major and minor axes, a
and b, respectively, as f 2 ¼ a2 − b2 and is related to the eccentricity e as Q1
e ¼ f ∕a. To solve the Helmholtz equation, it is separated into its longitudinal
and transverse components. The solution to the first is simply the function
expð−ik z zÞ, whereas the transverse part has the form49
 
∂2 ∂2 f 2 k 2t
þ þ ðcosh 2ζ − cos 2ηÞ uT ðζ, ηÞ ¼ 0, (58)
∂ζ2 ∂η2 2

where the parameter k t is the transverse component of the wave vector k. Using
separation of variables allows to split this equation into the radial and angular
Mathieu equations, the solution to which yields the nondiffracting even and odd
Mathieu beams:48,50

M eν ðζ, η, z; QÞ ¼ C ν J eν ðζ; QÞceν ðη; QÞ expðik z zÞ, (59)

M oν ðζ, η, z; QÞ ¼ S ν J oν ðζ; QÞseν ðη; QÞ expðik z zÞ: (60)

The parameters C ν and S ν are the normalization constants, whereas J eν ðζ; QÞ and
J oν ðζ; QÞ are the even and odd radial Mathieu functions, respectively, whereas
ceν ðη; QÞ, sev ðη; QÞ are the even and odd angular Mathieu functions. The sub-
index ν indicates the order of the function and is a positive integer, which takes
the value ν = 0, 1, 2, 3, . . . , for even and ν ¼ 1, 2, 3, : : : : for odd beams.
Further, Q is a dimensionless parameter related to the transverse component of
the wave vector k t as Q ¼ ðf k t ∕2Þ2 : In a similar way to Bessel beams, Mathieu
beams are also nondiffractive and carry an infinite amount and therefore cannot
be generated experimentally. Nonetheless, a finite-energy version can be gener-
ated by means of a Gaussian beam, from which it takes the name MG beam.
Crucially, the nondiffracting properties exist for a finite propagation distance
only. These modes are described mathematically as50



1 ik 2t z r2
MGeν ðζ̃, η̃, z; QÞ¼ exp − M ν ðζ̃, η̃, z; QÞ exp − 2 expðikzÞ,
e
T 2k T T ω0
(61)

2 z

2

1 ik t r
MGoν ðζ̃, η̃, z; QÞ¼ exp − M oν ðζ̃, η̃, z; QÞ exp − 2 expðikzÞ,
T 2k T T ω0
(62)
32 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

with the Cartesian coordinates redefined as


iz
x ¼ fo 1þ cosh ζ̃ cos η̃, (63)
zR


iz
y ¼fo 1þ sinh ξ̃ sin η̃, (64)
zR
and f 0 is the semifocal separation at z ¼ 0. The parameter μ is defined in terms of
the Rayleigh range zR as, T ¼ 1 þ iz∕zR , where ω0 is the waist radius of the
Gaussian beam. Figures 30(a) and 30(b) show the transverse intensity distribution
and phase profile of even MG beams with specific parameters z ¼ 0, k t ¼ 6,
e ¼ 0.9, and f 0 ¼ 0.9 and orders ν = 0, 1, 2, and 3.
Figure 31 shows four examples of odd MG beams, given by the same param-
eters, the intensity distribution is shown in Fig. 31(a) and the phase distribution in
Fig. 31(b).
In a similar way to IG beams, the superposition of even and odd MG beams
gives rise to the helical MG (HMG) beams, given mathematically as

ν ðζ̃, η̃, z; QÞ ¼ MGν ðζ̃, η̃, z; QÞ  iHGν ðζ̃, η̃, z; QÞ:

MGh e o (65)
Figures 32(a) and 32(b) show the transverse intensity profile and phase distribu-
tion of even, odd and HMG, MGhþ h−
ν and MGν , respectively, for the specific
parameter, k t ¼ 6, e = 0.5, f 0 ¼ 0.5, and ν ¼ 4.
Figure 33(a) shows four examples of the holograms encoded on the SLM to
generate the MG beams shown in Fig. 33(b). The first corresponds to an even
MG mode of order ¼ 6, the second to an odd beam of order 4, both with

Figure 30 Even MG beams. (a) Intensity distributions and (b) phase profile.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 33

Figure 31 Odd MG beams. (a) Intensity distribution and (b) phase profile.

Figure 32 HMG. (a) Transverse intensity distribution and (b) phase profile.

parameters k t ¼ 6, e ¼ 0.9, and f 0 ¼ 0.81. The third and fourth panels corre-
spond to an HMG beam of order ν ¼ 4 and eccentricity e ¼ 0.5. The third with
a positive helicity and parameters k t ¼ 4 and f 0 ¼ 0.75 while the fourth with neg-
ative helicity and parameters k t ¼ 6 and f 0 ¼ 1.35. The MATLAB code to gener-
ate these holograms can be downloaded directly from MathWorks in Ref. 51.

4.7 Vectorial beams

So far, we have only considered light beams with a transverse homogeneous
polarization distribution, in this section we will introduce a more general class
34 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 33 (a) Holograms generated through complex amplitude modulation, displayed on

the SLM. (b) Simulated intensity profiles of various MG beams.

of light beams characterized by a nonhomogeneous transverse polarization distri-

bution, commonly known as vector beams or “classically entangled” beams.52,53
Such beams have become a topic of late, in part due to their unique properties,
which are pioneering applications in fields as diverse as, optical tweezers, laser
remote sensing, optical communications, among others.54–61

4.7.1 Generalities about vector beams

Vector beams are solutions to the vectorial Helmholtz equation62,63 but are com-
monly regarded as a nonseparable weighted superposition of the spatial and
polarization degrees of freedom (DoFs).55,61 Mathematically, such superposition Q2
is represented as
̭ ̭
U ðrÞ ¼ cosΘuR ðrÞeR þ sinΘuL ðrÞeiδ eL , (66)
where the orthogonal functions uR ðrÞ and uL ðrÞ are associateḓ to the ̭ spatial DoF
with a weighting coefficient Θ. The orthogonal unitary vectors eR and eL represent
the right- and left-handed circular polarization states and the parameter δ is an inter-
modal phase between both polarization components.
Importantly, even though the polarization DoF is bounded to a bidimensional
space, the spatial DoF is not, and the various types of beams that we introduced in
the previous sections can be used to generate vector beams but perhaps the cylin-
drical BG and LG vector beams are the most popular. As such, in the following
section, we will briefly introduce such beams but we will also introduce vector
beams in elliptic cylindrical coordinates, in particular MG and IG vector beams,
as well as other types of vector beams.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 35

4.7.2 Vector beams in cylindrical coordinates

As a first example, we will focus on the set of LG vector (LGV) beams, which
can be written mathematically using Eq. (66) as
̭ ̭
U ðrÞ ¼ cosΘLGlp11 ðrÞeR þ sinΘLGlp22 ðrÞ expðiδÞeL : (67)
The set of LGV beams can be classified into two subsets defined by p1 ¼ p2 ¼ 0
and p1 ≠ 0, p2 ≠ 0. Each of these subsets can be split into two smaller subsets:
l1 ¼ l2 and l1 ≠ l2 , which can be classified by the type of disclinations (line
patterns) present in the orientation of the polarization ellipse which can be divided
into three morphologies: lemon, star, and monstar. Figure 34 shows four exam-
ples, with parameters f½l1 , p1 , ½l2 , p2 g given by: (a) f½0, 0, ½1, 0g with δ ¼ π,
known as lemon; (b) f½1, 0, ½0, 0g with δ ¼ 0, known as star; (c)
f½−2, 0, ½1, 0g, with δ ¼ 0, known as spider, and (d) f½−3, 0, ½3, 0g, with δ ¼ π
known as flower.
A particular subset of LGV beams are those given by the parameters
p1 ¼ p2 ¼ 0, l1 ¼ −l2 and δ ¼ 0, δ ¼ π, well known in the context of optical
fiber and often referred to as transverse electric (TE), characterized by a radial
polarization; transverse magnetic (TM), characterized by an azimuthal polariza-
tion; and hibrid electric odd (HEo) and hibrid electric even (HEe), characterized
by a hybrid polarization, described mathematically as55 Q3


1 ̭ ̭ ρ 2
TMðrÞ ¼ pffiffiffi ðexp½ilφeR þ exp½−ilφeL Þ exp − , (68)
2 ωðzÞ



1 ̭ ̭ ρ 2
TEðrÞ ¼ pffiffiffi ðexp½ilφeR − exp½−ilφeL Þ exp − , (69)
2 ωðzÞ



1 ̭ ̭ ρ 2
HEe ðrÞ ¼ pffiffiffi ðexp½−ilφeR þ exp½ilφeL Þ exp − , (70)
2 ωðzÞ

Figure 34 Examples of LGV beams: (a) lemon, (b) star, (c) spider, and (d) flower.
36 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 35 Cylindrical vector vortex beams, with the arrows depicting the local polarization
angle. From (a) to (d), they correspond to TM, TE, HEe , and HEo , respectively.

Figure 36 (a) Radial and (b) azimuthal vector vortex beams projected onto a linear polarizer
with its transmission axis at θp ¼ 0 deg , θp ¼ 45 deg , θp ¼ 90 deg , and θp ¼ 135 deg.


2 
1 ̭ ̭ ρ
HEo ðrÞ ¼ pffiffiffi ðexp½−ilφeR − exp½ilφeL Þ exp − : (71)
2 ωðzÞ
Figure 35 shows the transverse polarization distribution of the vector
beams described above, overlapped with their intensity profile, for this case
l ¼ 1. From Figs. 35(a)–35(d), they correspond to TM, TE, HEe , and HEo ,
respectively.
Figure 36 shows the intensity and polarization distribution of two of such
modes, namely, the TM and the TE, as well as their projection onto a linear polar-
izer with its transmission axis oriented at: θp ¼ 0 deg , θp ¼ 45 deg , θp ¼ 90 deg ,
and θp ¼ 135 deg.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 37

4.7.3 Vector beams in elliptical coordinates

Vector beams can also be generated in elliptical coordinates, using the IG or MG
beams as the spatial DoF.64,65 Here, we will show some examples of HIG vector
beams, which can be written mathematically using the general expression for vec-
tor beams [Eq. (66)] and the HIG beams as
̭ ̭
U ðrÞ ¼ cosΘIGhþ
q,μ,ε ðrÞeR þ sinΘIGq,μ,ε ðrÞ expðiδÞeL :
h− (72)
Figure 37 shows four specific examples of HIG vector beams with parameters: (a)
q ¼ 3, μ ¼ 3, ε ¼ 5, δ ¼ π; (b) q ¼ 3, μ ¼ 1, ε ¼ 1, δ ¼ π∕2; (c) q ¼ 4, μ ¼ 2,
ε ¼ 1, δ ¼ 0; and (d) q ¼ 5, μ ¼ 5, ε ¼ 5, δ ¼ π.

4.7.4 Other types of vector beams

Vector beams can be generated using any set of spatial modes. So far, we have
explicitly shown examples of LG and HIG vector beams. Nonetheless, in recent
times, other types have also been generated, each with unique properties.
Examples include parabolic-Gauss,66 accelerating,67 helicoconical,68 and abruptly
autofocusing69,70 vector beams. The generation process is not different from what
we have shown, which consists of the coaxial superposition of two orthogonal
spatial modes, each with orthogonal polarization.
As a first example, we would like to mention the parabolic vector beams,
whose spatial DoF is encoded in the so-called traveling parabolic-Gauss
beams.66 Such beams feature a free-space propagation-dependent separation of
both DoFs, evolving from a nonhomogeneous polarization distribution at the
plane z ¼ 0 to a homogeneous as z → ∞. As result, the local nonseparability
between the spatial and polarization DoFs decays to zero, maintaining a global
nonseparability. Figure 38 shows the transverse polarization distribution
overlapped with the intensity profile of a parabolic-Gauss beam at different propa-
gation distances, from (a) to (d): z ¼ 0, z ¼ 1.2zm , z ¼ 3zm , and z ¼ ∞,
with zm ¼ 1050 mm.
As a final example, we would also like to mention the recently introduced circular
Airy Gaussian vortex (CAGV) vector beams, characterized by a propagation-dependent

Figure 37 HIG vector beams with parameters (a) q ¼ 3, μ ¼ 3, ε ¼ 5, δ ¼ π; (b) q ¼ 3, μ ¼ 1,

ε ¼ 1, δ ¼ π∕2; (c) q ¼ 4, μ ¼ 2, ε ¼ 1, δ ¼ 0; and (d) q ¼ 5, μ ¼ 5, ε ¼ 5, δ ¼ π.
38 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 38 Parabolic–Gauss vector beams as a function of propagation. The transverse

polarization distribution overlapped with the intensity profile is shown at different propaga-
tion planes (a) z ¼ 0, (b) z ¼ 1.2zm (c) z ¼ 3zm , and (d) z ¼ ∞, where zm ¼ 1050 mm.

Figure 39 Abruptly autofocusing vector beams as a function of propagation distance. The

transverse polarization distribution overlapped with the intensity profile is shown at different
propagation planes: (a) z ¼ 0 mm, (b) z ¼ 600 mm, (c) z ¼ 720 mm, and (d) z ¼ 783 mm.

autofocusing behavior.69 These beams are generated from the CAGV beams,
which are the result of superimposing a large number of Airy beams in a circular
shape that upon propagation follow a parabolic trajectory toward the center of the
beam. As a result and upon propagation, the beam experiences a tunable abruptly
autofocusing behavior. Figure 39 shows a specific example of such beams, as a
function of propagation and for various propagation distances, namely, z ¼ 0,
600, 720, and 783 mm.

4.7.5 Experimental generation of vector beams

It is possible now to generate vector beams with LC q-plates, exploiting the geo-
metric phase,71 or with computer-controlled devices, such as SLMs72–74 or digital
micromirror devices (DMDs),75,76 exploiting the dynamic phase. Here, we will
briefly explain their generation through the Sagnac-based setup,74 schematically
shown in Fig. 40. To begin with, an expanded and collimated laser beam with
horizontal polarization is directed to two independent sections on the screen of
an SLM, each addressed with an independent digital hologram. Each hologram
is selected accordingly to generate the desired vector beam and superimposed
with a linear grating to separate and filter the first diffraction order of each
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 39

Figure 40 Schematic representation of a Sagnac-based experimental setup to generate

vector beams using an SLM. PBS, polarizing beam splitter; M, mirror; HWP, half wave-
plate; QWP, quarter wave-plate; L, lenses; SF, spatial filter; CCD, charge-coupled device
camera.

hologram. The two optical beams generated at the SLM are then passed through a
half wave-plate at 45 deg to rotate their polarization orientation to diagonal. Both
beams enter afterwards a common-path triangular Sagnac interferometer whose
main component is a polarizing beam splitter (PBS) that separates each beam into
its horizontal and vertical polarization components traveling along opposite direc-
tions inside the interferometer. The four beams exit the interferometer after a
round trip from the opposite side of the PBS, two with horizontal polarization
and two with vertical. To generate a vector beam, one beam of the first section
of the SLM with horizontal polarization is coaxially superimposed with the one
from the other section that carries vertical polarization. Fine alignment to ensure
the coaxial superposition can be done digitally through the linear gratings
encoded on the SLM. A quarter wave-plate can also be added to transform the
vector beam from the linear to the circular polarization basis.

Figure 41 Experimental Intensity profile of a TM cylindrical vector vortex mode projected

onto a linear polarizer oriented at different angles.
40 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 41 shows the experimentally generated TM vector vortex mode when

passed through a linear polarizer with its fast axis orientated horizontal, diagonal,
vertical, and antidiagonal.
5 Characterization of Scalar and Vector Beams
In this section, we will explain the process through which scalar and vector beams
can be characterized. First, we will explain how a scalar beam can be expanded in
terms of the basis elements of an orthonormal set, a process known as modal
decomposition. To continue, we will explain how the tools from quantum
mechanics allows to determine the quality of vector beams, through a measure
known as vector quality factor (VQF).

5.1 Characterization of scalar beams

5.1.1 Modal decomposition
We begin by noticing that it is possible to run the creation processes in reverse to
“unravel” an incoming light field into its “modes.” This powerful tool is known as
modal decomposition.77 The idea is that given some orthonormal basis set, it is
possible to expand any unknown function in terms of these basis elements.
Moreover, because the basis is orthonormal, it is also possible to find the expan-
sion coefficients without ambiguity. This is done by finding the inner product of
the unknown function and the basis element. It is well known that this can be
done optically, but only recently has this been done digitally too.78–83 The concept
is shown in Fig. 42. An incoming light field is modulated by the complex

Figure 42 The concept of modal decomposition. By performing an inner product on an

unknown field, it may be decomposed into an orthonormal basis. This is optically done by
passing the light through a lens and measuring the on-axis signal at the focal plane.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 41

conjugate of the basis function. The output is Fourier transformed and the signal
at the origin of the Fourier plane is evaluated. This signal equals the modal ampli-
tude. A similar procedure can be used for the modal phases. With the modal
amplitudes and phases known, the unknown field can be reconstructed and com-
pared to a measured field. The mathematical description of this technique is
explained next.
Imagine there is an unknown propagating field U ðx, yÞ whose power is nor-
malized to unity. This field can be written as a superposition of orthonormal func-
tions Ψn ðx, yÞ as follows:2,6
X
∞
U ðx, yÞ ¼ cn Ψn ðx, yÞ: (73)
n¼1

Due to their orthogonality, the inner product between any two functions can only
be zero or one, that is,
ZZ
Ψm ¼ Ψn ðx, yÞΨm ðx, yÞdxdy ¼ δnm : (74)

The modal coefficients cn are given as

ZZ
cn ¼ ρn exp½iϕn  ¼ hUi ¼ Ψn ðx, yÞU ðx, yÞdxdy, (75)

and satisfy the relation,

X
∞ X
∞
cn cn ¼ jcn j2 ¼ 1: (76)
n¼1 n¼1

The jcn j2 terms represent the power content of each mode in the expansion, e.g., if
the expansion was in terms of vortex beams then it would represent the power in
each vortex beam. The expression Ψn U inside the integrand of Eq. (75) is just the
incoming field after passing through a transfer function Ψn ðx, yÞ. If we note this as
ðx, yÞ, then its far-field diffraction at the focal plane of a lens will be given as
ZZ  
k0
W f ðX , Y Þ ¼ A0 W ðx, yÞ exp −i ðxX þ yY Þ dxdy, (77)
f
where A0 is a complex number, but as we shall see, it does not play any meaning-
ful role in the decomposition.
At the center of the beam (X,Y) = (0,0) and accordingly,

Ið0, 0Þ ¼ jW f ð0, 0Þj2 ¼ jA0 j2 jhU ij2 ∝ jcn j2 : (78)

Therefore, a hologram with a transfer function of the form T ðx, yÞ ¼ Ψðx, yÞ will
produce a diffraction pattern, whose axial center contains information about the
42 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

modal coefficients cn . Experimentally, we can place a detector at the focal plane

of the output beam to measure the on-axis intensity (at the origin) which should
be proportional to jcn j2 ¼ jρn j2 . The proportionality can be made exact by simply
adding all the signals of all holograms and normalizing to 1, as per Eq. (76). This
removes the constant A0. Changing the holograms displayed on an SLM very rap-
idly and measuring the on-axis intensity for each case can then achieve the char-
acterization of an input optical field. A more powerful method exploits the
linearity property of optics, and it simply involves adding (multiplexing) the holo-
gram of each mode into a final hologram, as discussed earlier and shown in
Fig. 5. Each hologram is encoded with a unique spatial carrier frequency (grating)
to spatially separate the signals at the Fourier plane. This method allows the meas-
uring of all the modal coefficients simultaneously. Moreover, the phase delay
between the modes can also be controlled digitally.
It is worth mentioning that both methods require a previous calibration to
define the position where the intensity must be measured. A transfer function of
the complex conjugate can send a known field (usually a Gaussian beam) through
the system and project on the SLM. The far-field pattern in this case would be a
Gaussian beam with an on-axis maximum intensity. The transverse coordinates
where the point of maximum intensity lies would then define the position where
the intensity should be measured.
Figure 43 shows results of the experimental implementation of modal decom-
position using the multiplexing principle.14 On the first SLM, a pure LG mode
was encoded using complex amplitude modulation (input mode). On a second
hologram, the complex conjugated of a set of 15 LG beams was encoded, each
with a different spatial carrier frequency (hologram). As result, a set of 15

Figure 43 Example of modal decomposition using the multiplexing technique. Here, 15

holograms, multiplexed in a single hologram, where illuminated with an LG02 mode. The
far-field intensity pattern shows 15 different shapes spatially separated, and the brightest
spot coincides with the LG02 mode.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 43

individual beams are displayed at the Fourier plane. All of them with different
intensity patterns, but only one with a bright spot in its center, the one matching
the incoming beam.

5.2 Characterization of vector beams

A proper characterization of vector beams is also desired in many scenarios, such
as for certain applications where information about their purity is crucial. This
task is not trivial but the mathematical similarity between quantum entanglement
and “classical entanglement” allows us to borrow some of the mathematical tools
employed in quantum mechanics to describe classical entanglement. More pre-
cisely, at the quantum level, the nonseparability of entangled photons can be
quantified through the concurrence C and at the classical level, it also allows to
quantify the nonseparability of classical systems providing information about the
vectorial nature of the beam, if it is purely scalar, purely vector, or an intermediate
state.84 Nonetheless, it is crucial to bear in mind that at the quantum level, entan-
glement can exist between systems that are spatially separated from each other
(nonlocal), whereas at the classical level, vector beams can only be entangled
locally within the internal DoFs of a system.85 In this context, scalar beams
receive the value 0 whereas vector beams receive the value 1. Such measurement
is carried out by projecting the input beam into the spatial and polarization DoFs
using either an SLM or a DMD.86,87 Finally, it is worth mentioning that, recently,
other techniques have been proposed for a basis-independent characterization,88 as
well as for the characterization of spatially disjoint vector beams.89 Below, we
will provide additional details about the theoretical and experimental measure-
ment of the concurrence through an SLM.

5.2.1 Beam quality factor

As previously stated, in a vector beam, the polarization and spatial DoFs
cannot be factorized as the product of two independent terms, a property we
refer to as classical entanglement, by analogy to entangled photons. Given the
similarity, we can use the entanglement entropy, described by the Von
Neumann entropy, of the reduced density matrix of one of the subsystems. For
vector beams, it can be any of the two constituting DoFs, polarization, or spatial
mode. For the sake of clarity, it has become common to write a vector beams
use Dirac’s notation as,
pffiffiffi pffiffiffiffiffiffiffiffiffiffiffi
jΨi ¼ ajΨ1 i ⊗ jRi þ 1 − ajΨ2 i ⊗ jLi, (79)

where, the kets jΨ1 i and jΨ2 i represent the spatial DoF, the kets jRi and jLi is the
polarization state, and ⊗ is the tensor product between the vectors. Further, the
parameter a ∈ ½0, 1 is a relative weighting factor that controls the degree of
entanglement. For a ¼ 0 and a ¼ 1, we obtain scalar beams, whereas for
44 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

a ¼ 1∕2 we get a vector beam. More precisely, the degree of non-separability can
be measured using the Von Neumann entropy as,



ð1 þ sÞ 1þs ð1 − sÞ ð1 − sÞ
EðjΨiÞ ¼ − log − log , (80)
2 2 2 2
where s represents the length of the Bloch vector and is defined as
!1
X
3 2

s¼ hσi i , (81)
i¼1

where σi is the expectation value of the Pauli operators which is defined as

σ1 ¼ jHihHj − jVihVj,
1
σ2 ¼ ðjH þ VihH þ Vj − jH − VihH − VjÞ,
2
σ3 ¼ jRihRj − jLihLj: ð82Þ
The degree of nonseparability can be determined by computing the real part of the
concurrence C, which in the context of classically entangled beams has been
termed VQF.

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VQF ¼ ReðCÞ ¼ Re ð1 − s2 Þ : (83)
Experimentally, the expectation values σi can be obtained from a series of 12
projective on-axis intensity measurements. Six spatial projections are for the
right-handed circular polarization and six are for the left-handed. The spatial pro-
jections are implemented through a series of six spatial filters in the form of digital
holograms displayed on an SLM. Two of them correspond to the digital holograms
that constitute the vector mode, i.e., jΨ1 i and jΨ2 i, whereas the other four corre-
spond to a superposition of these with a varying intramodal phase, namely, jΨ1 i þ
expðiαj ÞjΨ2 i with αj ¼ ½0, π∕2, π, 3π∕2, see Fig. 44(a). To realize this technique
experimentally, the two orthogonal polarization states are first separated using
polarization optical elements. Each polarization component is then projected onto
the six digital holograms. The on-axis intensity I ij of the 12 different combinations
is then measured in the far field with the help of a lens. To further illustrate this,
the Pauli operator can be written in terms of the experimental intensities as
σ1 ¼ ðI 13 þ I 23 Þ − ðI 15 þ I 25 Þ,
σ2 ¼ ðI 14 þ I 24 Þ − ðI 16 þ I 26 Þ, (84)
σ3 ¼ ðI 11 þ I 21 Þ − ðI 12 þ I 22 Þ:
By way of example, Fig. 44(b) (top) shows the 12 far-field intensities, corre-
sponding to a pure vector beam, captured with a CCD camera. The normalized
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 45

Figure 44 (a) Schematic representation of the projections required to compute the inten-
sities from which the VQF is obtained. (b) (top) Far-field intensity distribution of the twelve
required projections and (bottom) on-axis normalized intensity measurements.

on-axis intensity values are also shown in the bottom of Fig. 44(b). Incidentally,
the number of required measurement can be reduced to 8 by replacing the SLM
with a polarization-insensitive DMD.88

6 Applications of SLMs
In this section, we describe some of the applications that have been developed
over the years, based on the unique properties of SLMs. The first two are related
to very practical applications, which allow to propagate or focus a given beam by
simply refreshing the displayed hologram. We then describe a technique that
allows to generate several beams simultaneously and from a single hologram, a
technique known as mode multiplexing. As a final application, we explain in
detail the simulation of atmospheric turbulence of propagating light fields.

6.1 Digital propagation and focusing

6.1.1 Digital propagation
In some cases, it is necessary to analyze a beam as it propagates, e.g., when cal-
culating the beam quality factor of a laser beam. Even though one can always
mount a CCD camera on a translating stage, its alignment can be rather cumber-
some. Digital propagation offers the possibility to simulate the propagation of a
beam digitally without the need for any moving parts. This method is based on
the angular spectrum, according to which the propagation of an optical field U
along a distance z can be described as90
46 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

U ðr, zÞ ¼ F −1 fF fU ðr, 0Þg exp½ik z zg, (85)

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
with k z ¼ k − k 2x − k 2y , r ¼ ðx, yÞ, while F and F −1 denote the Fourier trans-
form and its inverse, respectively. The digital propagation of a beam can be
achieved by ensuring that the hologram encodes the Fourier transform of the
desired starting field, U ðr, 0Þ, together with the phase Ψ ¼ k z z. Next, the SLM
should be placed in the back focal plane of a lens and the camera at the front focal
plane. With this configuration, the camera will observe the propagation of the ini-
tial field as the z value of the hologram is changed. This approach has been used
to propagate scalar91,92 and vector beams.93 The MATLAB code is available on-
line as Digital_prop.m.
By way of example, Fig. 45(a) shows the generated holograms to propagate
an LG beam of parameters ωo ¼ 1 mm, p ¼ 2, and l ¼ 2, for the specific planes:
z ¼ 100, 200, 300, and 400 mm. Figure 45(b) shows the experimental intensity
profile of the propagated beam.

6.1.2 Digital focusing

In the same way, digital focusing of a beam can also be implemented using an
SLM, by encoding the transfer function of a lens91,94
 
k 2
tðx, yÞ ¼ exp i ðx þ y Þ ,
2 (86)
2f

where f is the focal length of the lens. The hologram encoded on the SLM then
takes the form

Figure 45 Holograms encoded on (a) an SLM to digitally propagate (b) an LG22 beam.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 47

Figure 46 (a) Holograms encoded on an SLM to digitally focus a Gaussian beam.

(b) Intensity and (c) phase profile.

 
k 2
ΦSLM ¼ mod ðx þ y Þ þ 2πðGx x þ Gy yÞ, 2π :
2 (87)
2f
Figure 46(a) shows four holograms for different focal distances f ¼ 100, 200,
300, and 400 mm. For the sake of clarity, in these holograms we used
Gx ¼ Gy ¼ 0, but a grating should be added to isolate the first diffraction order.
The MATLAB code is also available online as Digital_Lens.m. By fixing the
CCD camera, 400 mm away from the SLM and varying the focal length f from
100 to 400 mm, the size of the intensity profile of a Gaussian beam should
decrease as the value of f increases. This effect is shown in Fig. 46(b). The phase
profile of the beam is also shown for completeness in Fig. 46(c).
This technique can also be applied to other types of beams, e.g., with LG
beams. For this, we can use the same expression we used to generate LG beams
but with the addition of the transmission function of a lens. The expression to
generate the hologram will have the form
 
k 2
ΦSLM ¼ f LG sin ΦLG þ ðx þ y Þ þ Gx x þ Gy y :
2 (88)
2f
48 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

6.2 Mode multiplexing

A popular application of spatially structured light is to use it as an encoding
alphabet, where each pattern carries unique information, as shown in Fig. 47.
This is commonly referred to as mode division multiplexing (MDM), a variant
of the more general space division multiplexing (SDM). In SDM, modes might
be separated spatially to minimize overlap, e.g., in multicore optical fiber, and
in this instance do not need to be orthogonal. In MDM, the modes densely occupy
the physical space and so orthogonality is important. The use of spatial DoFs
offers an in-principle unbounded modal space with which to communicate, but
this requires a toolbox for the creation and detection of such modes, as well as a
deep understanding of their coupling in perturbative media, e.g., nonideal optical
fiber, turbulent free-space, and underwater, to name a few.

6.2.1 Choosing the mode set

For a common channel such as optical fiber or free space, the scalar mode set to
use is found from solutions to the paraxial Helmholtz equation

f∇2t þ n2R k 2 − β2 gEðx, yÞ ¼ 0, (89)

where nR is the refractive index profile, β is the propagation constant, E(x,y) is the
transverse electric field vector, and the subscript t indicates that the operator is on
the transverse coordinates only. In ideal free-space with no perturbations, we see
that it is the symmetry in which Eq. (89) is solved that determines the mode set,
e.g., HG modes in Cartesian symmetry, as shown in section 4.6.4. We will return
to the perturbative free-space solution shortly. In the case of optical fiber, we will
use two common types to discuss the mode set: step-index fiber and graded-index
fiber, as shown in Figs. 48(a) and 48(b).
In the case of step-index core, the solutions to Eq. (89) are the LP modes
whose electric field is given as

Figure 47 Concept of encoding information using spatial modes. Each mode could be a
pixel value, an alphabet, or a binary bit in a string.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 49

Figure 48 (a) Step index core and (b) parabolic graded index core with their mode sets
shown on the right. The modes supported are the LP mode set and LG mode set, respec-
tively. The LG modes are plotted only for p = 0.

LPl, p ðr, ϕÞ ¼ Rl, p ðrÞΦl ðϕÞ, (90)

where
8
> J jlj ðura Þ
>
>
< J jlj ðuÞ , r < a
Rl, p ðrÞ ¼ (91)
>
> K jlj ðwr
aÞ
>
: , r ≥ a,
K jlj ðwÞ


cosðlϕÞ, even
Φl ðϕÞ ¼ (92)
sinðlϕÞ, odd,
and a is the radius of the fiber core, J and K are Bessel and modified Bessel func-
tions, respectively, and u and w are normalized propagation constants given as
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u ¼ k 2 nco − β2 , (93)
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
w ¼ β2 − k 2 ncl , (94)

and nco and ncl are the refractive indices of the core and the cladding, respectively.
The intensity plots of the LP modes are shown in Fig. 48(a).
If the fiber has a parabolic graded index following:

n2R ðrÞ ¼ n2co ð1 − 2γ2 r2 Þ, (95)

50 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

with γ ¼ ðnco − ncl Þðanco Þ, then the mode set returns to that of the LG modes
shown earlier and again here in Fig. 48(b), expressed as
LGel, p ¼ Al, p ðrÞ cosðlϕÞ, (96)

LGol, p ¼ Al, p ðrÞ sinðlϕÞ, (97)

with the radial function given in Eq. (46) from Section 4.6.5. Although the func-
tional form of the LP and LG modes looks rather different, the LP modes can be
approximated by the LG modes by careful selection of mode size and mode
order.95
In the above analysis, the modes are selected based on ideal channels, but
what if the channel has some perturbation, for example, stressed optical fiber or
turbulence in the atmosphere? In such a case, a new set of modes will be stable;
these are the eigenmodes of the complex channel.96 Once the eigenmodes are
found, their creation follows immediately using the complex amplitude modula-
tion approaches in the earlier sections, as has been demonstrated experimentally
for static optical aberrated systems97 and atmospheric turbulence channels.98
So far, we have considered the scenario where one mode at a time is created,
but it is also possible to create an array of spatially structured beams. There are at
least two easy ways to do this: (1) spatial multiplexing or (2) angular multiplex-
ing. In the trivial case of subdividing space (option 1), the SLM screen is over-
filled with a plane wave (usually a large Gaussian beam) and the screen is
subdivided into multiple sections, each with its own hologram that creates the
desired mode. All the modes can then be given the same grating to ensure they
propagate unidirectionally. An example of such a hologram and the resulting field
is shown in Fig. 49 for a 4 × 4 array of LG modes with l = 3 and p = 0. One can
appreciate that since the SLM screen is partitioned into segments, that the

Figure 49 (a) 4 × 4 array hologram to create (b) an array of vortex beams. If the pixel num-
ber and size are not sufficient to resolve the hologram then unwanted diffraction orders will
appear in the output, shown as the dim light in (c).
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 51

resolution of the SLM and optical system will ultimately limit what is possible in
each segment.99 It is instructive to use OAM as an example to highlight this: as
the OAM increases, so the phase gradient at some radius also increases. One
can show that there is an optical vortex density limit within a circular region of
radius R (centered on the axis of the hologram), which is given as
2πR
jljmax ¼ NA, (98)
λ
where NA is the numerical aperture of the optical system. For modes with
|l| > |lmax|, evanescent waves are excited that quickly decay in amplitude. In the
case that the hologram spans a pixel array of Nx × Ny, each with a pitch of δx
and δy, then the largest circle that can be inscribed as
1
R¼ minðN x δx , N y δy Þ, (99)
2
Once the SLM and hologram parameters are set, it is the NA that is the limiting
factor, often by pin-holes in the optical system. There is also a technological
aspect: the finite pixel size and number means that the executed hologram is
always a stepwise approximation to the ideal kinoform version. Less pixels in
each segment can result in many extra diffraction orders and less efficiency in
the desired order. An example of this is shown in Fig. 49, where the desired beam
is overlayed with extra undesired orders [Fig. 49(c)].
The second option is to spatially overlap the holograms and then separate out
the desired modes by giving each a unique grating that directs the mode at a
desired angle: angular multiplexing, shown in Fig. 50. The angles can be reverted

Figure 50 In the angular multiplexed version, a single spatially overlapped hologram cre-
ates multiple beams that are directed at angles relative to one another. The angles can be
converted to positions by using a lens.
52 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

back to spatial position by a simple lens, so the final positions and sizes can be
determined by the optical system itself. For instance, say the hologram for mode
Mj is Hj(x,y), and in addition a linear grating is added to each mode for a final
transmittance function of the form
X
tðx, yÞ ¼ exp½i2πðuj x þ vj yÞ exp½iH j ðx, yÞ, (100)
j

then after passing through a lens of focal length, f, the desired mode will appear at
position Xj = ujλf and Yj = vjλf. In this approach, one has to take care that the beam
sizes and grating periods are selected to avoid spatial overlap at the focal plane of
the lens. Such an approach has been used to create over 200 modes on a single
hologram.45
The aforementioned techniques are all external to the laser, but it is possible
with intracavity SLMs to create single modes100,101 and arrays of modes,102–104
and the reader is referred to the references for more details.

6.2.2 Detecting the mode set

The detection can be viewed as the creation step in reverse but exploiting modal
decomposition. In the angular multiplexed approach, as shown graphically in
Fig. 51, multiple input modes arrive overlapped with one another or in a time
series and are modulated by the hologram displayed on the SLM. The hologram
acts as a match filter (see Section 5.1.2) to send the detected modes to specific

Figure 51 The detection of the mode set can be viewed as the creation in reverse, shown
here for the example of an incoming modal set distributed to spatial locations on a detector.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 53

positions on the detector by using Eq. (100) and Xj = ujλf and Yj = vjλf, but in
detector rather than creator mode. This is done by replacing the hologram for Mj
with that for Mj*; in other words, the complex conjugate of the desired mode
but the same grating rules apply. Exactly such a scheme has been used to detect
100 modes simultaneously across three wavelengths.18 It is also possible to use
a matched set of spatially arrayed beams, following approach (1) from the crea-
tion step. For example, a detection hologram would be carefully aligned with
the creation hologram to unravel each mode in the array, forming an array of
spots. One can think of this as a “lock and key” system, where information is
placed into spatial and modal positions as a “lock” and then “unlocked” only with
the correct and matching key. This has been useful for encryption using spatial
modes.105,106
It is also possible to do this in a parallelized approach, where the usual place-
ment of the SLM and detector is inverted: the SLM is placed in the Fourier plane
between two lenses (far-field) and the detector in the image plane (near field) of
the incoming array. Using an array of OAM beams as an example, each vortex
in the array is imaged to the same far-field profile regardless of its position in
the array, as schematically shown in Fig. 52. The pattern in the far field can be
easily understood using the convolution theorem: since a vortex array corresponds
to the convolution of a Dirac comb with a vortex beam, its Fourier transform is
simply given by the pointwise product of the Fourier transforms of the Dirac
comb and the vortex, resulting in a single, large pixelated vortex. If the SLM is
programmed to be a match filter for the desired vortex beam and this is present,

Figure 52 Parallel detection of arrays of modes by placing the SLM in the focal plane of the
lens and the detector in the image plane. This is modal decomposition but exploiting the
convolution theorem for arrays of modes. The example shows a 10 × 10 array of l ¼ 1
detection signals and no signal in adjacent OAM values. Adapted under the CC-BY
license.107 Q4
54 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

then an array of Gaussian peaks will appear in the detector plane—analyzing all
array positions at once.107

6.3 Wavefront reconstruction

Wavefront reconstruction is a highly topical field, and SLMs have long lent them-
selves to wavefront sensing and reconstruction, e.g., as adaptive optics solutions.
The wavefront of an optical beam is defined as the continuous surface that is nor-
mal to the time average of the direction of energy propagation, i.e., normal to the
time average of the Poynting vector, P. Since there might be no continuous sur-
face that fulfills this condition, the wavefront can be more generally defined as
the continuous surface that minimizes the power density weighted deviations of
the direction of its normal vectors to the direction of energy flow in the measure-
ment plane,108 given as
ZZ  2
 Pt 

jPj − ∇t w dA → min : (101)
jPj
Here, w is the wavefront and Pt = [Px, Py, 0] and the subscript t makes clear that
all components are in the transverse plane only. For simple scalar beams, the
wavefront is directly related to the phase (θ) of the beam through
λ
w¼ θ: (102)
2π
It is not possible to directly measure the wavefront or phase of light, and instead,
all wavefront and phase measurements rely on a conversion to a detectable DoF,
usually the intensity of the light. The simplest approach is to perform a direct
mimic of a Shack–Hartmann (wavefront) sensor (SHS) made of a 2D array of
lenslets, all of the same focal length. This device works on the principle that
lenses map angles to position, so local tilts to the wavefront will see a displaced
focal spot for each lenslet proportional to this tilt. The tilt is proportional to the
local gradient of the wavefront, so the displaced focal spots essentially return the
derivative of the wavefront function. Integrating the interpolated version of this func-
tion returns the wavefront itself. Holographically, only an array of lenses needs to be
programmed, following all the procedures and algorithms from SHS can then be
directly applied without any modification.109 An example of such a digital wavefront
sensor is shown in Fig. 53.
An alternative approach is to use the transport of intensity equation (TIE).
Because light propagation is a phase-dominated process, monitoring the intensity
change during propagation is sufficient to deduce what the initial phase must have
been. It can be shown that the phase is related to the gradients in the transverse
and longitudinal intensity following:
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 55

Figure 53 A 4 × 4 digital lenslet array for wavefront sensing by the Shack–Hartmann

approach.


−k∂z I
θ ≈ ∇−2
t , (103)
I þC
where ∇−2 t is the inverse of the transverse Laplacian operator, which can be
numerically solved using fast Fourier transforms, and C is a constant chosen by
the user to deal with regions where I = 0 and should be chosen to be small enough
not to distort the phase outcome (typically 1% of the peak intensity works well).
To find the term ∂z ðx, y, zÞ, images of the intensity at multiple planes along the
propagation axis are required, e.g., at δz, 0, − δz, which can easily be acquired
by the digital propagation approach of Section 6.1.1. The gradient in the intensity
is then approximately
I out ðx, y, δzÞ − I out ðx, y, − δzÞ
∂z Iðx, y, zÞ ≈ : (104)
2δz
56 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 54 For the implementation of the TIE, two SLMs are needed. The first creates the
digital aberration and the second performs the digital propagation. If the beam is already
aberrated and only the wavefront is needed, then only step two is required. When used
together the first SLM can act as an adaptive mirror, programmed to correct for the incoming
aberration through a real-time iterative routine shown in the bottom panel.

The phase θ(x, y, 0) can then be deduced from Eqs. (103) and (104). This has
been executed on digital devices for fast retrieval of wavefronts110 and is shown
in Fig. 54. The set-up usually requires two SLMs although for just wavefront
reconstruction, one is sufficient: SLM 2 performs digital propagation on the
incoming aberrated beam to recover a wavefront by TIE, measuring the beam at
three planes and attempting to reconstruct the phase. In Fig. 53, the first SLM is
added to either create an aberrated beam or to be used as a phase corrector:
SLM 1 then becomes a digital adaptive mirror to program the conjugate of the
phase found from the TIE in order to correct for aberrations.
A final approach is to return to the modal decomposition of the optical field,
given previously as Eq. (73) and reproduced here for convenience
X
∞
U ðx, yÞ ¼ cn Ψn ðx, yÞ: (105)
n¼1
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 57

If the unknown complex coefficients cn are known (amplitude and phase) then
since the left-hand side of the equation is the optical field and not just the inten-
sity, the right-hand side contains all the information required to extract a wave-
front. It is in fact simply extracting the phase from the arg function applied to
the reconstructed field: θ = arg(U). In the earlier modal decomposition section,
we were only interested in the modal weights of the coefficients, i.e., |cn|2 which
gives the modal power of each expansion term. To find the complex term cn ¼
ρn expðiΔϕn Þ requires two additional measurements with hologram transmittance
functions given as
1   1  
n ¼ pffiffiffi ðΨ0 þ Ψn Þ and T n ¼ pffiffiffi ðΨ0 þ i Ψn Þ:
T cos sin (106)
2 2
Note that this is essentially an interference of two modes, a reference mode usu-
ally taken to be n = 0 and the desired (n’th) mode whose phase we wish to find.
From the correlation of the incoming field with the above transmittance functions,
n ¼ ρ0 þ ρn þ 2ρ0 ρn sinðΔϕn Þ
one finds outcome intensities at the detector of I cos 2 2

and I n ¼ ρ0 þ ρn þ 2ρ0 ρn cosðΔϕn Þ, from which the intermodal phase is easily

sin 2 2

extracted as


n − ρn − ρ0
2I sin 2 2
Δϕn ¼ −arctan : (107)
n − ρn − ρ0
2I cos 2 2

This is the phase of each mode relative to the n= 0 mode. The modal approach
comes with the benefit that the resolution of the reconstruction is not
dependent on the number of measurements (always three per mode: one for the
modal power and two to extract the intermodal phase) nor the resolution of the
detector, since only single complex numbers are measured. The resolution resides
in the basis functions of the expansion, functions that are computationally imple-
mented in the addition, and therefore the wavefront can be plotted to any resolu-
tion. An example of this is shown in Fig. 55. An unknown laser beam is first
measured with an SHS to extract the intensity and wavefront. One can see the
pixelation due to the finite size and array number of the lenslets (one data point
per lens). In contrast, the modal approach with CGHs on SLMs returns excellent
resolution with just three modes in the expansion, each measured with a modal
weight and modal phase. The two wavefront reconstructions using the phase
directly (Phs) and the minimum surface approach (Min) return the same outcome.

6.4 Simulation of atmospheric turbulence

Understanding how light behaves in random or fluctuating media is a very active
area of research, e.g., in free-space optical communication, where turbulence
degrades the information signal.111,112 Here, we will describe a technique to simu-
late the effect of a turbulent atmosphere on propagating optical fields by using
58 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 55 A comparison of intensity (a), (b) and wavefront measured with an SHS (d), and
by a modal approach (c). Wavefront determined from a phase reconstruction (e) and by a
minimization approach (f). Adapted under the CC-BY license.82 Q5

SLMs.113–116 In this approach, we consider the effect of small temperature fluctu-

ations in the atmosphere, which give rise to variations in its refractive index n(t).
These variations originate as fluctuations in phase but after propagation this trans-
lates to both phase and amplitude perturbations on the optical field.
We can simplify the problem by considering only phase fluctuations. This
approximation is known as the “thin phase screen approximation” and allows
the turbulence to be approximated by a single phase screen. In this approximation,
the turbulence strength can be fully described by the dimensionless parameter,
D/r0. Here, D is the diameter of the aperture or beam and r0, known as Fried’s
parameter, is the atmospheric coherence length also known as the “seeing param-
eter,” and is given as112

2 3
λ 5
r0 ¼ 0.185 2 : (108)
cn z
The strength of a turbulent medium can be characterized by the Strehl ratio (SR),
defined at the receiver plane as111
I
SR ¼ , (109)
I0
where I and I0 are the on-axis mean irradiance of a point source in the presence
and absence of turbulence, respectively. This definition is very general and can
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 59

be applied to both the weak and strong turbulence regimes. SR will reach its
maximum value of 1 when I = I0, i.e., no turbulence, and its minimum SR = 0
when the medium is highly turbulent. The SR can be related to Fried’s parameter
r0 and the aperture diameter D as
1
SR ≅ 5 , (110)
1þ D
r0
3

and can be approximated for a single-phase screen in terms of the beam’s waist
size ω0 as follows:
1
SR ≅ 5 : (111)
ω0
1 þ 6.88 r0
3

Given that fluctuations in the atmospheric refractive index are stochastic, a statis-
tical description is needed. There are two approaches: (1) use appropriate random
drawings from known aberration weightings to create, in the spatial domain,113
the desired screen or (2) by working in the spatial frequency domain, make ran-
dom drawing give a known frequency power law. Here, we describe the latter
case. This is best represented in the Fourier domain as the Fourier transform of
the covariance function of the refraction index. In the Fourier domain, the power
spectral density of the refractive-index fluctuations is given as112

Φn ðkÞ ¼ 0.0033C 2n k − 3 :
11
(112)
Commonly known as the Kolmogorov power spectral density. Here, k is the sca-
lar spatial frequency. The power spectral density function tells you how to ran-
domly draw spatial frequencies so as to best mimic the atmosphere. Hence,
turbulence phase screens can be generated by encoding the Fourier transform of
the product of a random function with the power spectrum above. Figure 56
shows several phase screens generated using the above approach, for the case
SR = 0.9. Here, we also provide a MATLAB code (Turb.m, also available online)
that generates this phase screens.

Figure 56 Different realizations of phase screens for an SR = 0.9.

60 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Figure 57 Phase screen as function of SR. The turbulence increases as SR decreases.

Figure 58 Numerical simulation of the effect of turbulence on a vortex beam. (a) Intensity,
(b) phase profile, and (c) hologram encoded on the SLM to simulate the effect of turbulence.

Figure 57 shows the effect of the SR on the phase screen. As can be seen, as
the SR decreases, the phase changes become more abrupt. Thus, the phase of a
beam impinging on an SLM, where these phase screens are displayed, will also
change accordingly. As a result, the intensity profile of the beam will degrade.
For these examples, SR = 0.9, 0.7, 0.5, and 0.1.
As a final example, we illustrate the effect of turbulence on a vortex beam.
This effect can be simulated by adding the turbulence phase screen and the phase
of a vortex beam, as shown in Fig. 58. Figure 58(a) shows the intensity profile for
decreasing values of the SR (from 1.0 to 0.1), as expected, the intensity profile of
the vortex beam degrades. Figure 58(b) shows the corresponding phase for each
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 61

case and Fig. 58(c) shows the hologram displayed on the SLM to simulate the
effect of turbulence.
The MATLAB file Vortex_Turbulence.m (also available online) generates the
phase screens shown in Fig. 58(c). Here, the aperture size is taken to be the beam
waist of the input beam.

7 Generation of Holograms
This section explains in detail some of the MATLAB codes that are used to gen-
erate the holograms shown in sections 4.6.1 to 4.6.5. These codes, along with the
rest of the codes, are available online. Even though these examples are orientated
to MATLAB, the process is general to any programming language. A full list of
all the codes included is as follows:
(1) vortexbeam.m. This code generates a vortex beam using phase-only
modulation.
(2) Bessel.m. This code generates a BG beam by encoding on the SLM the
transmittance function of an axicon.
(3) Airy.m. This function generates the finite-energy Airy beams.
(4) Hermite_Gaussian.m. This function generates the set of HG beams, using
complex amplitude modulation. This code uses the function Hermite.m to
evaluate the different Hermite polynomials and the file fx2.mat, where the
values of the numerical inversion of Eq. (45) are stored. Both are also avail-
able online.
(5) Hermite.m. This function evaluates the Hermite polynomial.
(6) Laguerre_Gaussian.m. This function generates the set of LG beams, using
complex amplitude modulation. This code uses the function Laguerre.m to
evaluate the associated Laguerre polynomials as well as the file fx2.mat.
neeed to invert the function given by Eq. (50).
(7) Laguerre.m. This function evaluates the Laguerre polynomial.
(8) Digital_Lens.m. This code implements a digital focusing of any beam.
(9) Digital_prop.m. This code implements a digital propagation of any beam.
(10) Vortex_Turbulence.m. This code generates atmospheric turbulence
phase-screens.
(11) Turb.m. This code illustrates how to simulate the effect of turbulence on a
vortex beam.

7.1 Phase-only modulation

7.1.1 Vortex beams
The first examples will explain in detail how to generate the holograms using phase-
only modulation. It is important to mention that these holograms should have the res-
olution of the SLM. For example, 1920 × 1080 in the case of the PLUTO version
from HoloEye or 600 × 800 in the case of the X10468-01 from Hamamatsu.
62 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

The following lines will generate a matrix with the resolution of the PLUTO
SLM from HoloEye.

1. clear all; close all;

2. H=1920; V=1080; %%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1); y=-V/2:1:(V/2-1);
4. [X,Y]=meshgrid (x, y); %%Meshgrid

X and Y are two 1080 × 1920 matrices with values from –H/2 to H/2-1
and –V/2 to V/2-1, respectively. An azimuthally varying phase, phi can be
generated as

5. phi=angle(X+1i*Y); %%Azimuthal angle

6. l=1;%%Topological charge

The topological charge (l) of the beam can take any positive or negative
values.

7. nx=100; ny=100;%%Number of horizontal and vertical grooves.

8. gx=nx/H; gy=ny/V;

The number of grooves nx and ny in the vertical and horizontal directions can
take any integer value but are limited by the resolution of the SLM. The following
line is one of the most important as it encodes the hologram:

9. Hol=mod(n*phi+2*pi*(gy*Y+gx*X),2*pi);

where again, mod(x, 2*pi) is the 2π modulo operation that creates the blazed
grating as explained in Section 2.1. The following lines will normalize the gray
levels of Hol to fully cover the gray levels available by the SLM.

10. SLM=Hol/max(Hol(:))*255; % Grayscale normalization

The resultant image can be displayed as an image on the SLM (once

connected to the PC) directly from MATLAB as described below. It is very
important to first set the arrangement of the screens (PC and SLM) as “extended
monitor:”
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 63

11. fig=figure(1);
12. set(fig,'Position',[1200 0 1920 1080], 'MenuBar',
'none','ToolBar','none','resize','off');
13. set(gca,'position',[0 0 1 1],'Visible','off')
14. imagesc(SLM)
15. colormap gray
16. axis off;

In line 13, the command (. . . 'Position',[xc yc H V]. . . ) specifies the

position (xc yc) and size (H, V) of the hologram, which should be displayed on
the SLM. The position of the images needs to be adjusted according to the reso-
lution of your monitor, which can vary from one computer to another. The com-
mand scrsz = get(0,'ScreenSize') can be useful to find out what is
the resolution of the available monitor. The commands . . . 'MenuBar',
'none', 'ToolBar', 'none', 'resize','off', remove unnecessary
options from the images as is the case of the tool and menu bars.

7.1.2 Bessel beams

The phase encoded on the SLM to generate a BG beam is given by Eq. (30) and
repeated here for the sake of clarity:
ΦSLM ¼ mod½kαðn − 1Þ þ lϕ þ 2πðGx X þ Gy Y Þ, 2π:

The first three lines of the code are the same as in the previous case. The following
lines will rescale everything to mm for which we need to know the size of the pixel,
a parameter that might differ from device to device. In the case of the PLUTO
version of HoloEye, the pixel size is 8 μm. We also have to define the refraction
index ni, the wave number k, the wavelength λ, and the axicon angle α.

1. clear all; close all;

%Initializing Hologram Matrices
2. H=1920; V=1080; %%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1); y=-V/2:1:(V/2-1);
4. x=x*8e-3;%(in mm) Scales the hologram in the V direction
5. y=y*8e-3;%(in mm) Scales the hologram in the H direction
6. [X,Y]=meshgrid(x, y);
7. lambda=633e-6; % Wavelength in mm
8. k=2*pi/lalambda; %Wavenumber
9. rho=sqrt(X.^2+Y.^2);
10. phi=angle(X+1i*Y);
11. ni=1.5; % Index of refraction of glass
12. l=2; % Topological charge
13. gama=.48*pi/180; % Axicon angle
14. kr=k*(ni-1)*gama;
64 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

The following lines are the same as in the previous case:

15. nx=0;
16. ny=0;
17. gy=ny/(V*8e-3);
18. gx=nx/(H*8e-3);

The following line (19) defines the transmission function of an axicon, in which
we also encode the phase of a vortex to generate high-order BG beams

19. Hol=mod(-l*phi+kr*rho+2*pi*(X*gx+Y*gy),2*pi);
20. SLM=Hol/max(Hol(:))*255;
21. fig=figure(1);
22. set(fig,'position',[1900 0 1920 1080],'MenuBar',
'none','ToolBar','none','resize','off');
23. set(gca,'position',[0 0 1 1],'Visible','off')
24. imagesc(SLM)
25. colormap gray
26. axis off;

7.1.3 Airy beams

Airy beams can also be generated in the same way as before, for this we have to
use Eq. (35), which is repeated here for the sake of clarity:
 
X3 þ Y3
ΦSLM ¼ mod , 2π :
3

The full MATLAB code to generate this type of beam is as follows:

1. clear all; close all;

%Initializing Hologram Matrices
2. H=1920; V=1080; %%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1); y=-V/2:1:(V/2-1);
4. [X,Y]=meshgrid (x, y); %%Meshgrid
5. x=x*8e-3;%(in mm) Scales the hologram in the V direction
6. y=y*8e-3;%(in mm) Scales the hologram in the H direction
7. [X,Y]=meshgrid(x, y);
8. A= (X.^3+Y.^3)/3; %Cubic phase
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 65

9. Hol=mod(A,2*pi);
10. SLM=Hol/max(Hol(:))*255;
11. fig=figure(1);
12. set(fig,'Position',[1920 0 1920 1080],'MenuBar',
'none','ToolBar','none','resize','off');
13. set(gca,'position',[0 0 1 1],'Visible','off')
14. imagesc(SLM)
15. colormap gray
16. axis off;

7.2 Complex amplitude modulation

7.2.1 HG beams and LG beams
Complex amplitude modulation requires more lines of coding. Here, we will
describe only how to generate HG modes; the rest can be generated in an analo-
gous way.

1. clear all; close all;

%Initializing Hologram Matrices
2. H=1920; V=1080; %%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1); y=-V/2:1:(V/2-1);
4. [X,Y]=meshgrid (x, y); %%Meshgrid
5. x=x*8e-3;%(in mm) Scales the hologram in the V direction
6. y=y*8e-3;%(in mm) Scales the hologram in the H direction
7. [X,Y]=meshgrid(x, y);
8. phi=angle(X+1i*Y);
9. rho=sqrt(X.^2+Y.^2);
10. lambda=633e-6;
11. w0=1;%% Input beam waist.
12. z=.00001;
13. k=2*pi/lambda;
14. zr=pi*w0^2/lambda;
15. w=w0*sqrt(1+(z/zr)^2);
16. R=z*(1+(zr/z)^2);
17. m=2;
18. n=0;

In the above, n and m define the order of the Hermite polynomials, which can be
computed using the function Hermite.m, mentioned above. The functions are
also available online and can be found in Appendix A.
66 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

19. Hx=polyval(HermitePoly(m),sqrt(2)*X/w);
20. Hy=polyval(HermitePoly(n),sqrt(2)*Y/w);

The following lines are required to compute the amplitude and phase of the
HG beam:

21. rc=sqrt(2^(1-n m)/(pi*factorial(n)*factorial(m)))/w;

22. HG=rc*Hx.*Hy.*exp(1i*(n+m+1)*atan(z/zr)).*exp(-rho.^2/
w^2).*exp(-1i*k*rho.^2/(2*R))*exp(1i*k*z);
23. HG=HG/sqrt(sum(sum(abs(HG).^2)));
24. ph=angle(HG); %%Phase of HG modes
25. A=abs(HG)/max(max(abs(HG)));%%Amplitude normalized to
unity

In the above, A and ph are the amplitude (A) and phase (φ), respectively, to
generate the required hologram using the expression:
ΦSLM ¼ f ðAÞ sinðϕÞ,
where f ðAÞ can be found by numerical inversion of Eq. (45). The following lines
would compute f ðAÞ and store it in the function F. For this, the function
“fx2.mat,” also provided online, needs to be loaded.

26. load fx2.mat;

27. aux=round(A*800+1);
28. for mh=1:V;
29. for nh=1:H;
30. temp=aux(mh,nh);
31. F(mh,nh)=fx(temp);
32. End
33. end
34. nx=50;
35. ny=0;
36. gy=ny/(V*8e-3);
37. gx=nx/(H*8e-3);
38. Hol=F.*sin(ph+2*pi*(X*gx+Y*gy));
39. Hol=Hol-min(Hol(:));
40. SLM=Hol/max(Hol(:))*255;
41. fig=figure(1);
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 67

42. set(fig,'Position',[1900 0 1920 1080],'MenuBar',

'none','ToolBar','none','resize','off');
43. set(gca,'position',[0 0 1 1],'Visible','off')
44. imagesc(SLM)
45. colormap gray
46. axis off;

7.3 Digital propagation and focusing

7.3.1 Digital propagation
The full MATLAB code to digitally propagate a beam is described next. As in the
previous codes, everything has been normalized to millimeter.

1. clear all; close all;

%Initializing Hologram Matrices
2. H=1920; V=1080; %%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1); y=-V/2:1:(V/2-1);
4. [X,Y]=meshgrid (x, y); %%Meshgrid
5. x=x*8e-3;%(in mm) Scales the hologram in the V direction
6. y=y*8e-3;%(in mm) Scales the hologram in the H direction
7. [X,Y]=meshgrid(x, y);
8. rho=sqrt (X.^2+Y.^2);
9. nx=100;
10. ny=100;
11. gy=ny/(V*8e-3);
12. gx=nx/(H*8e-3);
13. lambda=0.63e-3;
14. k=2*pi/lambda;
15. kz=2*pi*sqrt(1/lambda^2-rho.^2);%
16. fig=figure(1);
17. set(fig,'Position',[1920 0 1920 1080],'MenuBar',
'none','ToolBar','none','resize','off');
18. set(gca,'position',[0 0 1 1],'Visible','off')
19. for zp=1:500;% Propagation distance in mm
20. Hol=mod(kz*zp+2*pi*(X*gx+Y*gy),2*pi);
21. Hol=Hol-min(Hol(:));
22. SLM=Hol/max(Hol(:))*255;
23. imagesc(SLM)
24. colormap gray
68 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

25. axis off;

26. drawnow;
27. end

7.3.2 Digital focusing

The full MATLAB code to digitally focus a beam is described next. As in the pre-
vious codes, everything is normalized to mm so that the focal distance should also
be given in millimeter.

1. clear all; close all;

%Initializing Hologram Matrices
2. H=1920; V=1080; %%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1); y=-V/2:1:(V/2-1);
4. [X,Y]=meshgrid (x, y); %%Meshgrid
5. x=x*8e-3;%(in mm) Scales the hologram in the V direction
6. y=y*8e-3;%(in mm) Scales the hologram in the H direction
7. [X,Y]=meshgrid(x, y);
8. rho=sqrt (X.^2+Y.^2);
9. nx=100;
10. ny=100;
11. gy=ny/(V*8e-3);
12. gx=nx/(H*8e-3);
13. lambda=0.63e-3;
14. k=2*pi/lambda;
15. ff=400;% Focal distance in mm
16. T=pi/lambda/ff*rho.^2;%Transmittance function of a Lens
17. Hol=mod(T+2*pi*(X*gx+Y*gy),2*pi);
18. Hol=Hol-min(Hol(:));
19. SLM=Hol/max(Hol(:))*255;
20. fig=figure(1);
21. set(fig,'Position',[1920 0 1920 1080],'MenuBar',
'none','ToolBar','none','resize','off');
22. set(gca,'position',[0 0 1 1],'Visible','off')
23. imagesc(SLM)
24. colormap gray
25. axis off;
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 69

7.4 Simulation of turbulence

The following code will generate a turbulence phase screen. This code uses the
function turb.m. In order for this code to accurately simulate the effect of atmos-
pheric turbulence on a beam, it is important to measure the beam’s waist w0 at the
SLM’s screen.

1. clear all; close all;clc

%Initializing Hologram Matrices
2. H=960; V=1080;%%Number of Horizontal and Vertical pixels
3. x=-H/2:1:(H/2-1);y=-V/2:1:(V/2-1);
4. [X,Y]=meshgrid(x, y);
5. phi=angle(X+1i*Y);%%Azimuthal angle
6. n=1;%%Topological charge
7. nx=50;ny=50;%%Number of horizontal and vertical grooves
8. gx=nx/H; gy=ny/V;
9. SR=.5;
10. w0=1;%%Beam size at the SLM in mm
11. Pixel=8;%Pixel size in microns
12. [turb] = Turb(H,V,SR, w0,Pixel);
13. Hol=mod(turb+n*phi+2*pi*(Y*gy+X*gx),2*pi);
14. SLM=Hol/max(Hol(:))*255;
15. fig=figure(1);
16. set(fig,'Position',[1200 0 H V],'MenuBar',
'none','ToolBar','none','resize','off');
17. set(gca,'position',[0 0 1 1],'Visible','off')
18. imagesc(SLM)
19. colormap gray
20. axis off;

8 Appendix A: Hermite Polynomials

The following function computes the Hermite polynomial H_n for positive integer
values n. The result is a vector whose m’th element is the coefficient of xðnþ1−mÞ :

1. function H = HermitePoly(n)
2. if n==0
3. H = 1;
4. elseif n==1
5. H = [2 0];
70 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

6. else
7. hpm2 = zeros(1,n+1);
8. hpm2(n+1) = 1;
9. hpm1 = zeros(1,n+1);
10. hpm1(n) = 2;
11. for k=2:n
12. H = zeros(1,n+1);
13. for r=n-k+1:2:n
14. hk(r) = 2*(hpm1(r+1) - (k-1)*hpm2(r));
15. end
16. H(n+1) = -2*(k-1)*hpm2(n+1);
17. if k<n
18. hpm2 = hpm1;
19. hpm1 = H;
20. end
21. end
22. end

9 Appendix B: Laguerre Polynomials

The following function generates the Laguerre functions, which is stored in the
variable “y.” The inputs are the indices “p”, “l,” and the vector “x.”

1. function y=laguerre(p,l,x)
2. y=zeros(p+1,1);
3. if p==0
4. y=1;
5. else
6. for m=0:p
7. y(p+1-m)=((-1)^m*(factorial(p+l)))/(factorial(p-
8. m)*factorial(l+m)*factorial(m));
9. end
10. end
11. y=polyval(y,x);

Acknowledgments
We would like to thank the South African National Research Foundation, the
Claude Leon Foundation, and CONACyT for financial support. We also thank
Dr. Angela Dudley of the CSIR and Dr. Hiao-Bo Hu of the Hangzhou Zhejiang
Sci-Tech University, for assistance with some of the figures.
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 71

References
1. F. M. Dickey, Laser Beam Shaping: Theory and Techniques, CRC Press, New York (2014).
2. A. Forbes, Laser Beam Propagation: Generation and Propagation of Customized Light, CRC
Press, New York (2014).
3. D. L. Andrews, Structured Light and Its Applications, Elsevier (2008).
4. J. P. Torres and L. Torner, Twisted Photons, Wiley-VCH (2011).
5. H. Rubinsztein-Dunlop et al., “Roadmap on structured light,” J. Opt. 19, 013001 (2017).
10.1088/2040-8978/19/1/013001
6. A. Forbes, A. Dudley, and M. McLaren, “Creation and detection of optical modes with spatial
light modulators,” Adv. Opt. Photonics 8, 200–227 (2016). 10.1364/AOP.8.000200
7. A. E. Willner et al., “Optical communications using orbital angular momentum beams,” Adv.
Opt. Photonics 7, 66–106 (2015). 10.1364/AOP.7.000066
8. G. Li et al., “Space-division multiplexing: the next frontier in optical communication,” Adv.
Opt. Photonics 6, 413–487 (2014). 10.1364/AOP.6.000413
9. C. Maurer et al., “What spatial light modulators can do for optical microscopy,” Laser
Photonics Rev. 5, 81–101 (2011). 10.1002/lpor.200900047
10. G. Nehmetallah and P. P. Banerjee, “Applications of digital and analog holography in three-
dimensional imaging,” Adv. Opt. Photonics 4, 472–553 (2012). 10.1364/AOP.4.000472
11. P. Memmolo et al., “Recent advances in holographic 3D particle tracking,” Adv. Opt. Photonics
7, 713–755 (2015). 10.1364/AOP.7.000713
12. W. Osten et al., “Recent advances in digital holography invited,” Appl. Opt. 53, G44–G63
(2014). 10.1364/AO.53.000G44
13. B. R. Boruah, “Dynamic manipulation of a laser beam using a liquid crystal spatial light modu-
lator,” Am. J. Phys. 77, 331–336 (2009). 10.1119/1.3054349
14. D. Huang et al., “A low-cost spatial light modulator for use in undergraduate and graduate
optics labs,” Am. J. Phys. 80, 211–215 (2012). 10.1119/1.3666834
15. N. Savage, “Digital spatial light modulators,” Nat. Photonics 3, 170–172 (2009). 10.1038/
nphoton.2009.18
16. Y. Q. Yang, A. Forbes, and L. C. Cao. “A review of liquid crystal spatial light modulators:
devices and applications,” Opto-Electron, Sci. 2, 230026 (2023). 10.29026/oes.2023.230026
17. A. Dudley et al., “Implementing digital holograms to create and measure complex-plane optical
fields,” Am. J. Phys. 84, 106–112 (2016). 10.1119/1.4935354
18. A. Trichili et al., “Optical communication beyond orbital angular momentum,” Sci. Rep. 6,
27674 (2016). 10.1038/srep27674
19. D. M. Spangenberg et al., “White light wavefront control with a spatial light modulator,”
Opt. Express 22, 13870–13879 (2014). 10.1364/OE.22.013870
20. R. W. Cohn and M. Liang, “Approximating fully complex spatial modulation with pseudoran-
dom phase-only modulation,” Appl. Opt. 33, 4406–4415 (1994). 10.1364/AO.33.004406
21. R. W. Cohn, “Pseudorandom encoding of complex-valued functions onto amplitude-coupled
phase modulators,” J. Opt. Soc. Am. A 15, 868–883 (1998). 10.1364/JOSAA.15.000868
22. V. Arrizón, “Complex modulation with a twisted-nematic liquid-crystal spatial light modulator:
double-pixel approach,” Opt. Lett. 28, 1359–1361 (2003). 10.1364/OL.28.001359
23. V. Arrizón et al., “Pixelated phase computer holograms for the accurate encoding of scalar
complex fields,” J. Opt. Soc. Am. A 24, 3500–3507 (2007). 10.1364/JOSAA.24.003500
24. E. G. Van Putten, I. M. Vellekoop, and A. P. Mosk, “Spatial amplitude and phase modulation
using commercial twisted nematic LCDs,” Appl. Opt. 47, 2076–2081 (2008). 10.1364/AO.47.
002076
25. V. Arrizón, “Optimum on-axis computer-generated hologram encoded into low-resolution
phase-modulation devices,” Opt. Lett. 28, 2521–2523 (2003). 10.1364/OL.28.002521
72 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

26. T. W. Clark et al., “Comparison of beam generation techniques using a phase only spatial light
modulator,” Opt. Express 24, 6249–6264 (2016). 10.1364/OE.24.006249
27. J. A. Davis et al., “Encoding amplitude information onto phase-only filters,” Appl. Opt. 38,
5004–5013 (1999). 10.1364/AO.38.005004
28. J. A. Davis, K. O. Valadéz, and D. M. Cottrell, “Encoding amplitude and phase information
onto a binary phase-only spatial light modulator,” Appl. Opt. 42, 2003–2008 (2003). 10.1364/
AO.42.002003
29. V. Arrizón, G. Méndez, and D. Sánchez-de La-Llave, “Accurate encoding of arbitrary complex
fields with amplitude-only liquid crystal spatial light modulators,” Opt. Express 13, 7913–7927
(2005). 10.1364/OPEX.13.007913
30. E. Bolduc et al., “Exact solution to simultaneous intensity and phase encryption with a single
phase-only hologram,” Opt. Lett. 38, 3546–3549 (2013). 10.1364/OL.38.003546
31. D. W. K. Wong and G. Chen, “Redistribution of the zero order by the use of a phase checker-
board pattern in computer generated holograms,” Appl. Opt. 47, 602–610 (2008). 10.1364/AO.
47.000602
32. R. Bowman et al., “Optimisation of a low cost SLM for diffraction efficiency and ghost order
suppression,” Eur. Phys. J. Spec. Top. 199, 149–158 (2011). 10.1140/epjst/e2011-01510-4
33. M. Bawart, E. Bernet, and M. Ritsch-Marte, “Programmable freeform optical elements,”
Opt. Express 25, 4898–4906 (2017). 10.1364/OE.25.004898
34. G. C. G. Berkhout et al., “Efficient sorting of orbital angular momentum states of light,”
Phys. Rev. Lett. 105, 153601 (2010). 10.1103/PhysRevLett.105.153601
35. E. Karimi et al., “Hypergeometric-Gaussian modes,” Opt. Lett. 32, 3053–3055 (2007). 10.1364/
OL.32.003053
36. B. Sephton, A. Dudley, and A. Forbes, “Revealing the radial modes in vortex beams,”
Appl. Opt. 55, 7830–7835 (2016). 10.1364/AO.55.007830
37. J. Durnin, “Diffraction-free beams,” Phys. Rev. Lett. 58, 1499–1501 (1987). 10.1103/
PhysRevLett.58.1499
38. M. Mazilu et al., “Light beats the spread: non-diffracting beams,” Laser Photonics Rev. 4, 529–
547 (2010). 10.1002/lpor.200910019
39. V Garces-Chavez et al., “Simultaneous micromanipulation in multiple planes using a self-
reconstructing light beam,” Nature 419, 145–147 (2002). 10.1038/nature01007
40. M. McLaren et al., “Self-healing of quantum entanglement after an obstruction,” Nat. Commun.
5, 4248 (2014). 10.1038/ncomms4248
41. J. Tervo and J. P. Turunen, “Rotating scale-invariant electromagnetic fields,” Opt. Express
9, 9–15 (2001). 10.1364/OE.9.000009
42. C. Schulze et al., “Accelerated rotation with orbital angular momentum modes,” Phys. Rev. A
91, 043821 (2015). 10.1103/PhysRevA.91.043821
43. M. Zamboni-Rached, “Stationary optical wave fields with arbitrary longitudinal shape by
superposing equal frequency Bessel beams: frozen waves,” Opt. Express 12, 4001–4006
(2004). 10.1364/OPEX.12.004001
44. G. A. Siviloglou et al., “Observation of accelerating airy beams,” Phys. Rev. Lett. 99, 213901
(2007). 10.1103/PhysRevLett.99.213901
45. C. Rosales-Guzmán et al., “Multiplexing 200 spatial modes with a single hologram,” J. Opt.
19, 113501 (2017). 10.1088/2040-8986/aa8b8e
46. M. A. Bandres and J. C. Gutiérrez-Vega, “Ince–Gaussian modes of the paraxial wave equation
and stable resonators,” J. Opt. Soc. Am. A 21(5), 873–880 (2004). 10.1364/JOSAA.21.000873
47. Miguel A. Bandres, “Ince Gaussian Beam” (2014). https://www.mathworks.com/matlabcentral/
fileexchange/46222-ince-gaussian-beam
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 73

48. J. C. Gutiérrez-Vega, M. D. Iturbe-Castillo, and S. Chávez-Cerda, “Alternative formulation for

invariant optical fields: Mathieu beams,” Opt. Lett. 25(20), 1493–1495 (2000). 10.1364/OL.25.
001493
49. J. C. Gutiérrez-Vega and R. M. Rodríguez-Dagnino, “Mathieu functions, a visual approach,”
Am. J. Phys. 71(3), 233–242 (2003). 10.1119/1.1522698
50. J. C. Gutiérrez-Vega and M. A. Bandres, “Helmholtz–Gauss waves,” J. Opt. Soc. Am. A 22,
289–298 (2005). 10.1364/JOSAA.22.000289
51. E. Cojocaru, “Mathieu Functions” Toolbox v.1.0 (2008). https://www.mathworks.com/
matlabcentral/fileexchange/22081-mathieu-functions-toolbox-v-1-0
52. A. Forbes, A. Aiello, and B. Ndagano, “Classically entangled light,” in Progress in Optics,
Taco D. Visser, Ed., pp. 99–153, Elsevier (2019).
53. T. Konrad and A. Forbes, “Quantum mechanics and classical light,” Contemp. Phys. 60, 1–22
(2019). 10.1080/00107514.2019.1580433
54. Q. Zhan, “Cylindrical vector beams: from mathematical concepts to applications,” Adv. Opt.
Photonics 1(1), 1–57 (2009). 10.1364/AOP.1.000001
55. C. Rosales-Guzmán, B. Ndagano, and A. Forbes, “A review of complex vector light fields and
their applications,” J. Opt. 20(12), 123001 (2018). 10.1088/2040-8986/aaeb7d
56. B. Ndagano et al., “Creation and detection of vector vortex modes for classical and quantum
communication,” J. Light. Technol. 36, 292–301 (2018). 10.1109/JLT.2017.2766760
57. L. Fang et al., “Vectorial Doppler metrology,” Nat. Commun. 12, 4186 (2021). 10.1038/
s41467-021-24406-z
58. S. Berg-Johansen et al., “Classically entangled optical beams for high-speed kinematic sens-
ing,” Optica 2, 864–868 (2015). 10.1364/OPTICA.2.000864
59. B. Ndagano et al., “Characterizing quantum channels with non-separable states of classical
light,” Nat. Phys. 13, 397–402 (2017). 10.1038/nphys4003
60. A. Forbes, M. de Oliveira, and M.R. Dennis, “Structured light,” Nat. Photonics 15, 253–262
(2021). 10.1038/s41566-021-00780-4
61. Y. Shen and C. Rosales-Guzmán, “Nonseparable states of light: from quantum to classical,”
Laser Photonics Rev. 16, 2100533 (2022). 10.1002/lpor.202100533
62. D. G. Hall, “Vector-beam solutions of Maxwell’s wave equation,” Opt. Lett. 21, 9–11 (1996).
10.1364/OL.21.000009
63. M. A. Bandres and J. C. Gutierrez-Vega, “Vector Helmholtz–Gauss and vector Laplace–Gauss
beams,” Opt. Lett. 30(16), 2155–2157 (2005). 10.1364/OL.30.002155
64. Y. L. Yao-Li et al., “Classically entangled Ince–Gaussian modes,” Appl. Phys. Lett. 116(22),
221105 (2020). 10.1063/5.0011142
65. C. Rosales-Guzmán et al., “Experimental generation of helical Mathieu–Gauss vector modes,”
J. Opt. 23, 034004 (2021). 10.1088/2040-8986/abd9e0
66. X.-B. Hu et al., “Free-space local nonseparability dynamics of vector modes,” Photonics Res.
9, 439–445 (2021). 10.1364/PRJ.416342
67. B. Zhao et al., “Parabolic-accelerating vector waves,” Nanophotonics 11, 681–688 (2022).
10.1515/nanoph-2021-0255
68. E. Medina-Segura et al., “Helico-conical vector beams,” Opt. Lett. 48, 4897–4900 (2023).
10.1364/OL.497773
69. X.-B. Hu et al., “Experimental generation of arbitrary abruptly autofusing circular Airy
Gaussian vortex vector beams,” Sci. Rep. 12, 18274 (2022). 10.1038/s41598-022-23157-1
70. X.-B. Hu et al., “Tunable longitudinal spin–orbit separation of complex vector modes,”
Opt. Lett. 48, 2728–2731 (2023). 10.1364/OL.486699
71. L. Marrucci, C. Manzo, and D. Paparo, “Optical spin-to-orbital angular momentum conversion
in inhomogeneous anisotropic media,” Phys. Rev. Lett. 96, 163905 (2006). 10.1103/
PhysRevLett.96.163905
74 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

72. C. Maurer et al., “Tailoring of arbitrary optical vector beams,” New J. Phys. 9, 78 (2007).
10.1088/1367-2630/9/3/078
73. C. Rosales-Guzmán, N. Bhebhe, and A. Forbes, “Simultaneous generation of multiple vector
beams on a single SLM,” Opt. Express 25, 25697–25706 (2017). 10.1364/OE.25.025697
74. B. Perez-Garcia et al., “On-demand tailored vector beams,” Appl. Opt. 56, 6967–6972 (2017).
10.1364/AO.56.006967
75. L. Perumal and A. Forbes, “Broadband structured light using digital micro-mirror devices
(DMDs): a tutorial,” J. Opt. 25, 074003 (2023). 10.1088/2040-8986/acd563
76. X.-B. Hu and C. Rosales-Guzmán, “Generation and characterization of complex vector modes
with digital micromirror devices: a tutorial,” J. Opt. 24, 034001 (2022). 10.1088/2040-8986/
ac4671
77. V. A. Soifer and M. A. Golub, Laser Beam Mode Selection by Computer Generated
Holograms, CRC Press (1994).
78. D. Flamm et al., “Mode analysis with a spatial light modulator as a correlation filter,” Opt. Lett.
37, 2478–2480 (2012). 10.1364/OL.37.002478
79. D. Flamm et al., “All-digital holographic tool for mode excitation and analysis in optical
fibers,” J. Lightwave Technol. 31, 1023–1032 (2013). 10.1109/JLT.2013.2240258
80. C. Schulze et al., “Measurement of the orbital angular momentum density of light by modal
decomposition,” New J. Phys. 15, 073025 (2013). 10.1088/1367-2630/15/7/073025
81. I. A. Litvin et al., “Azimuthal decomposition with digital holograms,” Opt. Express 20, 10996–
11004 (2012). 10.1364/OE.20.010996
82. C. Schulze et al., “Wavefront reconstruction by modal decomposition,” Opt. Express
20, 19714–19725 (2012). 10.1364/OE.20.019714
83. J. Pinnell et al., “Modal analysis of structured light with spatial light modulators: a practical
tutorial,” J. Opt. Soc. Am. A 37, C146–C160 (2020). 10.1364/JOSAA.398712
84. M. McLaren, T. Konrad, and A. Forbes, “Measuring the nonseparability of vector vortex
beams,” Phys. Rev. A 92, 023833 (2015). 10.1103/PhysRevA.92.023833
85. E. Toninelli et al., “Concepts in quantum state tomography and classical implementation with
intense light: a tutorial,” Adv. Opt. Photonics 11, 67–134 (2019). 10.1364/AOP.11.000067
86. B. Ndagano et al., “Beam quality measure for vector beams,” Opt. Lett. 41, 3407 (2016). 10.1364/
OL.41.003407
87. B. Zhao et al., “Determining the non-separability of vector modes with digital micromirror
devices,” Appl. Phys. Lett. 116, 091101 (2020). 10.1063/1.5142163
88. A. Selyem et al., “Basis-independent tomography and nonseparability witnesses of pure com-
plex vectorial light fields by Stokes projections,” Phys. Rev. A 100, 063842 (2019). 10.1103/
PhysRevA.100.063842
89. A. Aiello et al., “A non-separability measure for spatially disjoint vectorial fields,” New J.
Phys. 24, 063032 (2022). 10.1088/1367-2630/ac77ab
90. J. W. Goodman, Introduction to Fourier optics, Standford University Press (1996).
91. C. Schulze et al., “Beam-quality measurements using a spatial light modulator,” Opt. Lett.
37, 4687–4689 (2012). 10.1364/OL.37.004687
92. J. Webster, C. Rosales-Guzmán, and A. Forbes, “Radially dependent angular acceleration of
twisted light,” Opt. Lett. 42, 675–678 (2017). 10.1364/OL.42.000675
93. A. Dudley et al., “All-digital wavefront sensing of structured light beams,” Opt. Express
22, 14031–14040 (2014). 10.1364/OE.22.014031
94. J. Perez-Vizcaino et al., “Free-motion beam propagation factor measurement by means of a
liquid crystal spatial light modulator,” J. Disp. Technol. 8, 539–545 (2012). 10.1109/JDT.
2012.2200453
95. R. Brüning et al., “Overlap relation between free-space Laguerre Gaussian modes and step-
index fiber modes,” J. Opt. Soc. Am. A 32, 1678–1682 (2015). 10.1364/JOSAA.32.001678
Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators 75

96. A. Klug, C. Peters, and A. Forbes, “Distortion-free forms of structured light,” Opt. Photonics
News 34, 51 (2023).
97. W. T. Buono et al., “Eigenmodes of aberrated systems: the tilted lens,” J. Opt. 24, 125602
(2022). 10.1088/2040-8986/ac9f22
98. A. Klug, C. Peters, and A. Forbes, “Robust structured light in atmospheric turbulence,”
Adv. Photonics 5, 016006 (2023). 10.1117/1.AP.5.1.016006
99. J. Pinnell, V. Rodríguez-Fajardo, and A. Forbes, “Probing the limits of orbital angular
momentum generation and detection with spatial light modulators,” J. Opt. 23, 015602
(2020). 10.1088/2040-8986/abcd02
100. S. Ngcobo et al., “A digital laser for on-demand laser modes,” Nat. Commun. 4, 2289 (2013).
10.1038/ncomms3289
101. C. Tradonsky et al., “High-resolution digital spatial control of a highly multimode laser,”
Optica 8, 880–884 (2021). 10.1364/OPTICA.423140
102. A. Pando et al., “Improved laser phase locking with intra-cavity adaptive optics,” Opt. Express
31, 6947–6955 (2023). 10.1364/OE.482196
103. N. Davidson et al., “Complex-light lasers,” Opt. Photonics News 33 (5), 26–33 (2022).
10.1364/OPN.33.5.000026
104. M. Piccardo et al., “Roadmap on multimode light shaping,” J. Opt. 24, 013001 (2021).
10.1088/2040-8986/ac3a9d
105. H. Wang et al., “Coloured vortex beams with incoherent white light illumination,” Nat.
Nanotechnol. 18, 264–272 (2023). 10.1038/s41565-023-01319-0
106. A. Forbes and L. Perumal, “Twisted light gets a splash of colour,” Nat. Nanotechnol. 18, 221–
222 (2023). 10.1038/s41565-023-01322-5
107. M. Piccardo et al., “Vortex laser arrays with topological charge control and self-healing of
defects,” Nat. Photonics 16, 359–365 (2022). 10.1038/s41566-022-00986-0
108. ISO, “ISO 15367-1:2003 lasers and laser-related equipment – test methods for determination
of the shape of a laser beam wavefront – Part 1: Terminology and fundamental aspects,”
(2003).
109. L. Zhao et al., “Efficient implementation of a spatial light modulator as a diffractive optical
microlens array in a digital Shack–Hartmann wavefront sensor,” Appl. Opt. 45, 90–94
(2006). 10.1364/AO.45.000090
110. K. Singh, A. Dudley, and A. Forbes, “Versatile all-digital transport-of-intensity based wave-
front sensor and adaptive optics using a DMD,” Opt. Express 31, 8987–8997 (2023).
10.1364/OE.481767
111. L. C. Andrews and R. L. Phillips, Laser Beam Propagation through Random Media, SPIE
Press, Bellingham, WA (1998).
112. L. C. Andrews, Field Guide to Atmospheric Optics, SPIE Press, Bellingham, WA (2004).
113. L. Burger, I. Litvin, and A. Forbes, “Simulating atmospheric turbulence using a phase-only
spatial light modulator,” S. Afr. J. Sci. 104, 129–134 (2008).
114. G. T. Bold et al., “Practical issues for the use of spatial light modulators in adaptive optics,”
Opt. Commun. 148, 323–330 (1998). 10.1016/S0030-4018(97)00586-5
115. D. C. Dayton et al., “Theory and laboratory demonstrations on the use of a nematic liquid
crystal phase modulator for controlled turbulence generation and adaptive optics,” Appl. Opt.
37, 5579–5589 (1998). 10.1364/AO.37.005579
116. G. D. Love, “Wavefront correction and production of Zernike modes with a liquid crystal spa-
tial light modulator,” Appl. Opt. 36, 1517–1524 (1997). 10.1364/AO.36.001517
76 Rosales-Guzmán and Forbes: Structured Light with Spatial Light Modulators

Carmelo Rosales-Guzmán received his PhD with Cum

Laude distinction from ICFO—The Institute of Photonic
Sciences, Spain, in 2015. He then spent 3 years as a
research fellow with Professor Andrew Forbes, at the
University of the Witwatersrand, South Africa. In 2018, he
was hired as a professor and principal investigator by the
University of Science and Technology of Harbin, China. In
2020, he became a principal investigator at Centro de
Investigaciones en Optica, Mexico, where he is establishing one of the first struc-
tured light laboratories of the country. He holds a level-2 rating by the National
System of Researchers of Mexico. He is currently a member of the editorial panel
of Journal of Optics and Frontiers in Physics, and in 2023 he was elected by APL
Photonics as a member of the Early Career Editorial Advisory Board. In his short
academic career, he has coauthored more than 80 research articles, which include
several reviews and tutorials. His outstanding contributions to the field of struc-
tured light granted him the 2023-SPIE Early Career Achievement Award-
academic focus.

Andrew Forbes has at various times in his career found

himself as a teacher, janitor, secretary, receptionist, web-
master, systems engineer, sales rep, manager, director, and
sometimes a scientist. He is presently a distinguished pro-
fessor in the School of Physics at the University of the
Witwatersrand, South Africa, where in 2015 he established
a new laboratory for Structured Light. He is active in pro-
moting photonics in Africa, a founding member of the
Photonics Initiative of South Africa and initiator of South Africa’s Quantum
Roadmap. He is a fellow of SPIE, Optica, the South African Institute of Physics
(SAIP), and an elected member of the Academy of Science of South Africa. He
holds an A-rating by the South African NRF, 4 honorary professorships, is edi-
tor-in-chief of the IoP’s Journal of Optics, and sits on the editorial board of five
other international journals. He has won several awards, including the NSTF
national award for his contributions to photonics in South Africa, the Georg
Forster prize from the Alexander von Humboldt Foundation for outstanding con-
tributions to photonics, the SAIP Gold Medal, the highest award for physics in
South Africa, making him the youngest winner to date, the Sang Soo Lee award
from Optica and the Korean Optical Society and the TWAS prize for physics.
 Queries
Q1. Please note that the variable “e” has been changed as lightface italics. Please check
and confirm.

Q2. Please note that “degrees of freedom/degree of freedom” has been changed as
“DoFs/DoF” in thoughout the article. Please check and confirm.

Q3. Please note that the variables “eR” and “eL” has been changed as bold roman as in
equation 66. Please check and confirm.

Q4. Please check whether the insertion of “reference 107” citation in accurate here. If
not please remove the citation of reference 107.

Q5. Please check whether the insertion of “reference 82” citation in accurate here. If not
please remove the citation of reference 82.