Optical Design with Zemax

for PhD - Basics

Lecture 5: Aberrations I
2013-06-06
Herbert Gross

Summer term 2013 www.iap.uni-jena.de

 2

Preliminary Schedule

No Date Subject Detailed content

Zemax interface, menus, file handling, system description, editors,

preferences, updates, system reports, coordinate systems, aperture, field,
1 02.05. Introduction
wavelength, glass catalogs, layouts, raytrace, system insertion, scaling,
component reversal
Diameters, stop and pupil,pick ups, solves, variables, ray fans, quick focus,
2 16.05. Fundamentals
3D geometry, ideal lenses, vignetting, footprints, afocal systems,
Aspheres, gratings and diffractive surfaces, special types of surfaces,
3 23.05. Properties of optical systems I
telecentricity
4 30.05. Properties of optical systems II Ray aiming, Delano diagram, lens catalogs
Representations, geometrical aberrations, spot, Seidel, transverse aberration
5 06.06. Aberrations I
curves, Zernike wave aberrations
6 13.06. Aberrations II PSF, MTF, ESF

7 20.06. Imaging Fourier imaging, geometrical images

8 27.06. Advanced handling I Slider, universal plot, I/O of data, multi configurations

9 04.07. Optimization Algorithms, merit function, methodology, correction process, examples

10 11.07. Correction I Principles, simple systems

 3

Contents

1. Representations
2. Geometrical aberrations
3. Aberration expansions
4. Primary aberrations
5. Wave aberrations
Optical Image Formation
 Perfect optical image:
All rays coming from one object point intersect in one image point
 Real system with aberrations:
1. tranverse aberrations in the image plane
2. longitudinal aberrations from the image plane
3. wave aberrations in the exit pupil

object image
plane wave plane
aberrations

transverse
aberrations
optical
system

longitudinal
aberrations
Representation of Geometrical Aberrations

 Longitudinal aberrations Ds

Gaussian image
plane
reference ray logitudinal aberration Gaussian
along the reference ray image
plane
Dl' ray
ray

reference
U' point
optical axis optical axis
Dl'o

longitudinal aberration
system
projected on the axis
system D s'

longitudinal
aberration

 Transverse aberrations Dy
reference ray D y' transverse
(real or ideal chief ray) aberration

ray

U'

optical axis
reference
plane
system
Representation of Geometrical Aberrations

 Angle aberrations Du
ideal reference ray angular aberration
DU'

real ray

optical axis

system

x
reference sphere Gaussian
wavefront reference
 Wave aberrations DW W>0
plane

paraxial ray

real ray

U'
C z

R D y'

D s' < 0
Angle Aberrations

 Angle aberrations for a ray bundle:

deviation of every ray from common direction of the collimated ray bundle
 Representation as a conventional spot diagram
 Quantitative spreading of the collimated
bundle in mrad / °

real beam
Du
real angle
perfect spectrum
collimated
z
Longitudinal Aberration Chart

spherical aberration coma in zone coma in full field

4 colors 4 colors 4 colors

secondary
chromatic

chromatical difference distortion image shells/astigmatism

in magnification main color 4 colors
Transverse Aberrations

 Typical low order polynomial contributions for:

defocus, coma, spherical aberration, lateral color
 This allows a quick classification of real curves

Dy '  K 'r ' p cos  p Dy '  S 'r '3p cos  p

Dy '  C ' y 'r '2p (2  cos 2 P )

linear: quadratic: cubic: offset:
defocus coma spherical lateral color
Transverse Aberrations

 Characteristic chart for the representation of transverse aberrations

axis field zone full field
Dy Dy Dy meridional
deviation

yp yp yp

Dx Dx sagittal
wavelengths: deviation
365 nm
480 nm
546 nm
644 nm
xp xp
Spot Diagram

 All rays start in one point in the object plane

 The entrance pupil is sampled equidistant
 In the exit pupil, the transferred grid
may be distorted
 In the image plane a spreaded spot
diagram is generated

object plane entrance pupil optical exit pupil image plane

point equidistant grid system transferred grid spot diagram

yo yp y'p y'

xo xp x'p x'

z
Spot Diagram

 Variation of field and color

 Scaling of size:
1. Airy diameter (small circle)
2. 2nd moment circle (larger circle, scales with wavelength)
3. surrounding rectangle 486 nm 546 nm 656 nm

axis

field
zone

full
field
Gaussian Moment of Spot

 Spot pattern with transverse aberrations Dxj and Dyj

1. centroid 1 1
D xS   D x j D yS   D y j
N j N j
2. 2nd order moment
1

j  D x j  D xS    
2 2
MG  D r 2   D y j  D yS
N
3. diameter D  2  M G

 Generalized:
Rays with weighting factor gj:
corresponds to apodization

MG  D r 2 
1
NG j

 g j  D x j  D xS  
2
 D y j  D yS 
2

 Worst case estimation:

size of surrounding rectangle Dx=2Dxmax, Dy = 2Dymax
Aberrations of a Single Lens
y

 Single plane-convex lens,

BK7, f = 100 mm, l = 500 nm
 Spot as a function of field position
 Coma shape rotates according to circular symmetry x

 Decrease of performance with the distance to the axis

 Example HMD without symmetry

y
image 8

free formed
surface 6

total 2
internal
reflection 0 x

eye -2
pupil
-4

free formed
surface -6

field angle 14° -8

-8 -6 -4 -2 0 2 4 6 8
Fermat Principle
x
 Fermat principle: P1
the light takes the ray path, which corresponds
to the shortest time of arrival ds
C
 The realized path is a minimum and therefore
the first derivatives vanish
P2 P1
 L    n( x, y, z ) ds  0
z

P1

 Several realized ray pathes have the same

optical path length
P2
 
L   n  s  dr  const.
P1
n d
1 1
P
 The principle is valid for n d
3 3
smooth and discrete n d
2 2 n d
index distributions 4 4

n d
5 5

P'
Paraxial Point Eikonal for a Refracting Surface

 Refracted ray by a spherical

dielectric interface
L(r )  n  ( z  s) 2  (r  x) 2  n' ( z  s' ) 2  (r  x' ) 2
path difference

 Paraxial approximation:
Taylor expansion for small x, x‘, r, z
 nx 2 n' x'2 2s'   nx n' x'  2 1  n n' n'n 
L(r )   ns  n' s'    r   s  s '   r  2  s  s '  R 
 2 s     

 Stationary phase dL(r )

 B  2r  C  0
condition dr

rx r  x'
 Angles u, u‘ u , u' 
s s'

surface
Q
a
a'
P u
n n' R u' P'
r
x x'
z
(-)s s'
 17

Notations for an Optical System

x, y object coordinates, especially object height
x', y' image coordinates, especially image height
xp,yp coordinates of entrance pupil
y'
x'p, y'p coordinates of exit pupil
s object distance form 1st surface Dx'
x' P'0 P' z
s' image distance form last surface
y'
p entrance pupil distance from 1st surface Dy'
p' exit pupil distance from last surface y'p s'
Dx' sagittal transverse aberration
Dy' meridional transverse aberration y x'P p'
image
p
plane

xP y'p

system x'P
surfaces yp exit
xP pupil
y
p
entrance
x s pupil

object
plane y

P
Rotational Invariants
 General case : two coordinates in object plane and pupil
 Rotational symmetry: 3 invariants
1. Scalar product of field vector and pupil vector
 
w  P  F  P  F  cos( F   P )  x p  x  y p  y
yp

2. Square of field height

   z

u  F  F  F 2  x2  y2 xp P
xp P yp

3. Square of pupil height

 
v  P  P  P2  xp  y p
2 2 y


 Therefore: F x Pupi
l
F
Only special power y
x
combinations are
physically meaningful

Obje
ct
Power Series Expansion of Aberrations

 General case of Taylor expansion W a

k ,l , m , n
klmn x kp y lp x m y n

 Expansion with selection rules:

only powers of the rotational invariants can occur W  W (u , v, w)

 Simple expansion according

to this scheme

W  a0  b1v  b2 w  b3u
 c1v 2  c2 wv  c3 w2  c4uv  c5uw  c6u 2
 d1v 3  d 2 wv 2  d 3 w2 v  d 4uv 2  d 5uwv  d 6 w3  d 7uw2  d 8u 2 v  d 9u 2 w  d10u 3  ...

 Explicite equation in real coordinates

W  a0  b1 ( x p  y p )  b2 yy p  b3 y 2
2 2

 c1 ( x p  y p ) 2  c2 yy p ( x p  y p )  c3 y 2 y p  c4 y 2 ( x p  y p )  c5 y 3 y p  c6 y 4
2 2 2 2 2 2 2

 d1 ( x p  y p ) 3  d 2 yy p ( x p  y p ) 2  d 3 y 2 y p ( x p  y p )  d 4 y 2 ( x p  y p ) 2  d 5 y 3 y p ( x p  y p )
2 2 2 2 2 2 2 2 2 2 2

 d 6 y 3 y p  d 7 y 4 y p  d 8 y 4 ( x p  y p )  d 9 y 5 y p  d10 y 6  ...
3 2 2 2
 20

Primary Aberrations

 Expansion of the transverse aberration Dy on image height y and pupil height r

 Lowest order 3 of real aberrations: primary or Seidel aberrations
 Spherical aberration: S
- no dependence on field, valid on axis D y  r 3  S  r 2  y  r 2  cos   C
- depends in 3rd order on apertur
 y 2  r  cos 2   A  y 2  r  P
 Coma: C
- linear function of field y  y3  D
- depends in 2rd order on apertur with azimuthal variation
 Astigmatism: A
- linear function of apertur with azimuthal variation
- quadratic function of field size
 Image curvature (Petzval): P
- linear dependence on apertur
- quadratic function of field size

 Distortion: D
- No dependence on apertur
- depends in 3rd order on the field size
Polynomial Expansion of Aberrations

 Representation of 2-dimensional Taylor series vs field y and aperture r

 Selection rules: checkerboard filling of the matrix
 Constant sum of exponents according to the order
Image Primary
location aberrations /
Seidel
Field y Secondary
Spherical Coma Astigmatism aberrations
y0 y1 y2 y3 y4 y 5
5
y cos y 3 cos y cos
Distortion r0 Tilt Distortion Distortion
primary secondary
r1 y 2 r 1 cos2 y 4 r 1 cos2
r 1
Defocus y2r1
Astig./Curvat. y4r 1
Aper-
ture y r 2 cos y 3 r 2 cos3
2
r r Coma
primary y 3 r 2 cos
r3 y 2 r 3 cos2
3
r Spherical
primary y2 r3
y r 4 cos
4
r Coma
secondary
r5
5
r Spherical
secondary
 22

Transverse Aberrations of Seidel

 Transverse deviations
D x' 

x' p x'2p  y '2p s'4  S '
2 x' x' x'
p p  
 y ' y ' p   x' x'2p  y '2p s '3 s' p
C'
3 3
2n' R ' 2n' R '
 Sum of surface p p

contributions

x' x' x' p  y ' y ' p s'2 s'2p
A'
 
x' p x'2p  y '2p s'2 s'2p
P'
3 3
k
n' R ' 2n' R '
S'  S j p p

j 1

 
x' x'2  y '2 s' s'3p
D'
k 3
C'   C j 2n' R ' p
j 1
k
A'   A j
j 1 D y' 

y ' p x'2p  y '2p s '4 S '
2 y ' x' x'
p p  y ' y ' p   y ' 
x ' 2
p  y ' 2
p
s '3
s' p
C'
k 2n' R'3p 2n' R'3p
P'   Pj
j 1

y ' x' x' p  y ' y ' p s'2 s'2p
A'
 
y ' p x'2p  y '2p s'2 s'2p
P'
3 3
k n' R ' 2n' R '
D'   D j
p p

j 1

 
y ' x'2  y '2 s' s'3p
D'
2n' R'3p
Surface Contributions

 Spherical aberration  1 1 
Sj  Q
4  2

j  n' s ' n s 
j
 j j j j 

 1 1  n1 pjQ pj  1 1 
Cj   Q4 2
   
 Coma j n' s ' n s   Q
j s s 
 j j j j  j j  1 p1 

2 2
 Astigmatisms 4 2
 1 1   n1 pjQ pj  1 1 
Aj   j Q j     
 n' s ' n s    Q  s s 
 j j j j   j j   1 p1 

 Field curvature  1
2 2
1   n1 pjQ pj  1 1  1  1 1
Pj   4j Q 2j          
 n' s ' n s    Q   s s  r  n' n 
 j j j j   j j   1 p1  r  j j 
 Distortion

  1 1   n  Q 
2
 1 1 
2
1  1 1  n1 pjQ pj  1 1 
4 2       
D j   j Q j       
1 pj pj

  n' j s' j n j s j    j Q j   s1 s p1  rr  n' j n j   j Q j  s1 s p1 

       
  
Spherical Aberration

 Spherical aberration:
On axis, circular symmetry
 Perfect focussing near axis: paraxial focus
 Real marginal rays: shorter intersection length (for single positive lens)
 Optimal image plane: circle of least rms value

plane of the medium

smallest image
waist plane

As

2 As

marginal plane of the paraxial

ray focus smallest focus
rms-value
 25

Caustic with Spherical aberration

 Growing spherical aberration shows an asymmetric behavior around the nominal image
plane for defocussing

c9 = 0 c9 = 0.3 c9 = 0.7 c9 = 1
 26

Spherical Corrected Surface

 1 1 
 Seidel contribution of spherical aberration Sj  Q  4 2

 n' s ' n s 
j j
with
hj  1 1  j j j j 

j  Qj  nj    
R s 
h1  j j 

2
 1 1  1 1 
4
h 
 Result S j   j   n 2j      
R s   n' s ' n s 
 h1   j j   j j j j 

 Vanishing contribution:
1. first bracket: vertex ray hj  0

2. second bracket: concentric Rj  s j

3. bracket: aplanatic surface n' j s ' j  n j s j

 Discussion with the Delano formula
i 'u
n U sin u1 n i'i 2i  sin
Ds'SPH  DsSPH  1 1  j  h  sin  2
n'k U 'k sin u 'k j n' j 2 U ' j sin u ' j

2. concentric corresponds to i' = i

3. aplanatic condition corresponds to i' = u
Aplanatic Surfaces with Vanishing Spherical Aberration

 Aplanatic surfaces: zero spherical aberration:

1. Ray through vertex

s'  s  0
2. concentric
s'  s und u  u'
3. Aplanatic
ns  n' s'

sphere

sphere
hyperboloid oblate ellipsoid oblate ellipsoid prolate ellipsoid
 Condition for aplanatic + power series + power series + power series + power series
surface: Ds'
0.1
ns n' s' ss'
r   aplanatic S
n  n' n  n' s  s' 0
vertex concentric

-0.1
 Virtual image location
-0.2
 Applications:
-0.3
1. Microscopic objective lens
2. Interferometer objective lens -0.4

-0.5
0 50 100 150 200 250 300
Aplanatic Lenses

A-A :
 Aplanatic lenses parallel offset

 Combination of one concentric and

one aplanatic surface:
zero contribution of the whole lens to
A-C :
spherical aberration convergence enhanced

 Not useful:
1. aplanatic-aplanatic
2. concentric-concentric
bended plane parallel plate,
C-A :
nearly vanishing effect on rays convergence reduced

C-C :
no effect
Lens Contributions of Seidel

 Spherical aberration 1  n3 n2  2(n 2  1)

2
 n 2 (n  1) 
Slens      X   M   M 
2

32n(n  1) f 3  n  1 n  1  n  2  n  2 

 Special impact on correction: W sph

n = 1.5
1. Special quadratic dependence on
bending X
Minimum at
X sph min  

2 n2  1
M

n2

2. No correction for small n and M

3. Correction for large M=0
M=-3 M=3
n: infrared materials M=6
M=-6
M: virtual imaging
n n  2
Limiting value M s2 0  X
 n  12
Spherical Aberration: Lens Bending

object principal image

 Spherical aberration and focal spot diameter plane plane plane

as a function of the lens bending (for n=1.5)

 Optimal bending for incidence averaged
incidence angles
 Minimum larger than zero:
usually no complete correction possible

diameter

bending
X
 31

Astigmatism

 Geometrical caustic of astigmatism

 Sagittal and meritional focal lines in different z-location
 Midepoint between focal lines: circle of least confusion
 32

Astigmatism

 Imaging of a polar grid in different planes

 Tangential focus:
- blur in azumuthal direction
- rings remain sharp image space
circle sagittal line
 Sagittal focus: sagittal
tangential line
focus
- blur in radial direction exit
pupil
- spokes remain sharp
entrance
best focus
pupil

tangential
focus

object
Wavefront for Astigmatiusm with Defocus

 Astigmatic wavefront with defocus

 Purely cylindrical for focal lines in x/y
 Purely toroidal without defocus: circle of least confusion

cylindrical x

c4 = -1.0
toroidal
c4 = -0.5 c7 only

c4 = -0.25

c4 = 0 cylindrical y

c4 = 0.25

c4 = 0.5

c4 = 1.0
Astigmatism: Lens Bending

Astigmatism
 Bending effects astigmatism Seidel coefficients
 For a single lens 2 bending with 20 in [l]

zero astigmatism, but remaining 15

Sum
field curvature 10
Surface 2

0
Surface 1
-5

-0. 04 -0. 03 -0. 02 -0. 01 0 0. 01

Curvature of surface 1

T S TS S T ST T S

-2.5 0 2.5 -2.5 0 2.5 -2.5 0 2.5 -2.5 0 2.5 -2.5 0 2.5

Ref : H. Zügge
 35

Field Curvature and Image Shells

 Imaging with astigmatism:

Tangential and sagittal image shell sharp depending on the azimuth
 Difference between the image shells: astigmatism
 Astigmatism corrected:
It remains one curved image shell,
Bended field: also called Petzval curvature
image surfaces
 System with astigmatism: sagittal
y'

Petzval sphere is not an optimal shell

tangential ideal
surface with good imaging resolution shell image
 No effect of lens bending on curvature, plane
important: distribution of lens
powers and indices
Correction of Astigmatism and Field Curvature

 Different possibilities for the correction of astigmatism and field curvature

 Two independend aberrations allow 4 scenarious

a) bended b) bended c) flattened d) flattened

image plane image plane image plane image plane
residual corrected residual corrected
astigmatism astigmatism astigmatism astigmatism

T S y T S y S y T ST

Dz Dz Dz Dz
-2.5 0 2.5 -2.5 0 2.5 -2.5 0 2.5 -2.5 0 2.5
 37

Field Curvature

 Focussing into different planes of a system with field curvature

 Sharp imaged zone changes from centre to margin of the image field

focused at field boundary focused in field zone focused in center

(mean image plane) (paraxial image plane)

y'

receiving
planes

image
sphere
Petzval Theorem for Field Curvature

 Petzval theorem for field curvature: 1 n 'n

  nm '  k k
k nk  nk 'rk
1. formulation for surfaces R ptz
1 1

2. formulation for thin lenses (in air)
R ptz j nj  f j
 Important: no dependence on
bending

 Natural behavior: image curved object

towards system plane

 Problem: collecting systems

with f > 0:
If only positive lenses:
Rptz always negative R

ideal
optical system real image
image plane
shell
 39

Ray Caustic of Coma

 A sagittal ray fan forms a groove-like only sagittal
surface in the image space coma rays

 Tangential ray fan for coma: bended ray

caustic fan
Building of Coma Spot

 Coma aberration: for oblique bundels and finite aperture due to asymmetry
 Special problem: coma grows linear with field size y
 Systems with large field of view: coma hard to correct
 Relation of spot circles
and pupil zones as shown
chief ray
zone 1
zone 2
coma zone 3
blur

lens / pupil

axis
Coma

 Coma deviation, elimination of the azimuthal dependence:

circle equation
y'
 Diameter of the circle and position variiation with rp2
Every zone of the circlegenerates a circle in the tan 0°/180°

image plane
tangential
coma
 All cricels together form a comet-like shape

 The chief ray intersection point is at the tip of

the cone 45° 135°

 The transverse extension of the cone shape has

a ratio of 2:3
rp =1.0
the meridional extension is enlarged and gives
a poorer resolution
sag 90°

rp = 0.75

sagittal
coma rp = 0.5

30°
x'
Stop Position Influence for Corrected Spherical

Achromat 4/100, w = 10°, y‘ = 17.6, (Asphäre für Sph. Aberr. = 0)

Transverse aberr. +- 2.0 Spot

Asphäre w = 0° w= 10° w = 10°
Dy‘ Dy‘ Dy‘ Dy‘

Ref: H. Zügge
Coma Correction: Symmetry Principle

 Perfect coma correction in the case of symmetry

 But magnification m = -1 not useful in most practical cases

Image height: y’ = 19 mm

Symmetry principle Pupil section: meridional sagittal

Transverse Dy' Dy'
Aberration: 0.5 mm 0.5 mm

(a)

(b)

Ref: H. Zügge
Psf with Coma

 Separation of the peak and the centroid position in a point spread function with coma
 From the energetic point of view coma induces distortion in the image

c7 = 0.3 c7 = 0.5 c7 = 1

centroid
Distortion Example: 10%
 What is the type of degradation of this image ?
 Sharpness good everywhere !

Ref : H. Zügge
Distortion

 Distortion. change of magnification over the field

 Corresponds to spherical aberration of the chief ray
 Measurement: relative change of image height

y  yideal y'
V  real
yideal
Dh 3
ideal h3
 No image point blurr image
aberration
only geometrical shape deviation h2
Dh 2
 Sign of distortion:
h1
1. V < 0 : barrel, x'
image Dh 1
lens with stop in front height
2. V > 0: pincushion,
lens with rear stop

real
image
Distortion

 Purely geometrical deviations without any blurr

 Distortion corresponds to spherical aberration of the chief ray
 Important is the location of the stop:
defines the chief ray path y'
 Two primary types with different sign:
1. barrel, D < 0 rear
stop pincussion
front stop y lens x' distortion
2. pincushion, D > 0
D>0
rear stop
 Definition of local
x
magnification image
changes y'

front
y'  y'
D real ideal object y stop
x'
barrel
distortion
y 'ideal
D<0
x
Reasons of Distortion

 Distortion occurs, if the magnification depends on the field height y

 In the special case of an invariant location p‘ of the exit pupil:
the tangent of the angle of the chief ray should be scaled linear
 Airy tangent condition: tan w'
necessary but not sufficient condition for distortion corection:  const
tan w
 This corresponds to a corrected angle of the pupil imaging
form entrance to exit pupil

Dy'
O2 O'2
EnP ExP
real
y2 O1 w2 w'2 O'1 y2

wo'2
ideal
p p'
Reasons of Distortion

 Second possibility of distortion:

the pupil imaging suffers from longitudial spherical aberration
 The location of the exit pupil than depends on the field height
 With the simple relations y  p  tan w , y'  p' tan w'

we have the general expression y' p' ( y)  tan w'  po ' Dp' ( y)  tan w'
m( y)       
for the magnification y p  tan w  p p  tan w
 For vanishing distortion:
1. the tan-condition is fulfilled (chief ray angle)
2. the spherical aberration of the pupil imaging is corrected (chief ray intersection point)

Dy'
O2 O'2
EnP ExP
real
y2 O1 w2 w'2 O'1 y2

ideal
p po'
Dp'
 50

Axial Chromatical Aberration

 Axial chromatical aberration:

Higher refractive index in the blue results in a shorter intersection length for a single lens
 The colored images are defocussed along the axis
 Definition of the error: change in image location /
D s'CHL  s'F '  s'C'
intersection length
 Correction needs several glasses with different dispersion
 Single lens: normal dispersion
blue intersection length is shorter
than red P'

 Notations: white
1. CHL = chromatical longitudinal
2. AXCL = axial chromatic

s'
F'

s'e blue

green
s'C'

red
 51

Axial Chromatical Aberration

 Longitudinal chromatical aberration for a single lens

 Best image plane changes with wavelength

l best image plane

l= 648 nm

l= 546 nm

l= 480 nm

defocus
z
-2 -1 0 1 2

Ref : H. Zügge
 52

Lateral Color Aberration

 Dispersion of the chief ray deviation in the lens

 Effect ressembles the disperion of a prism in the upper part of the lens
 In the image plane, the differences in the colored ray angles cause changes in
the ray height
 The lateral color aberrations corresponds to a change of magnification with the
wavelength

dispersion y
prism effect
DyCHV
chief
ray

image
stop plane
 53

Chromatic Variation of Magnification

 Lateral chromatical aberration:

Higher refractive index in the blue results in a stronger D y'CHV  y'F '  y'C '
ray bending of the chief ray for a single lens
y 'F '  y 'C '
 The colored images have different size, D y 'CHV 
the magnification is wavelength dependent y 'e
 Definition of the error: change in image height/magnification
 Correction needs several glasses with different dispersion
 The aberration strongly depends on the stop position

stop
D y'
CHV
red

blue

reference
image
plane
 54

Chromatic Variation of Magnification

Representation of CHV: spot

1. Spot diagram diagram
chromatical field
2. Magnification m(l) magnification height
3. Transverse aberration: difference Y‘

offset of chief ray reference

CHV
0.047 0.108

transverse aberration curves

field field
axis
tangential sagital
Dy Dy Dx

Dy

yp xp
yp
 55

Chromatic Variation of Magnification

 Impression of CHV in real images

 Typical colored fringes blue/red at edges visible
 Color sequence depends on sign of CHV

original

without
lateral
chromatic
aberration

0.5 % lateral
chromatic
aberration

1 % lateral
chromatic
aberration
Law of Malus-Dupin
wave fronts
 Law of Malus-Dupin:
- equivalence of rays and wavefronts
- both are orthonormal
- identical information
rays
 Condition:
No caustic of rays

 Mathematical:
Rotation of Eikonal
vanish

rot  n  s   0 object image
plane plane
 Optical system:
Rays and spherical phase
L = const
waves orthonormal
y0 L = const

rays s
y1

z0 z1
 57

Relationships

 Concrete calculation of wave aberration: AP


addition of discrete optical path lengths lOPL   n  dr
(OPL) OE

 Reference on chief ray and reference

sphere (optical path difference)
DOPD ( x, y)  lOPL ( x, y)  lOPL (0,0)

 Relation to transverse aberrations W D y' D y'

 
 yp R W R

 Conversion between longitudinal R D y' R2 W (xp , y p )

transverse and wave aberrations Ds'   D y '   
yp sin u ' yp  yp
 Scaling of the phase / wave aberration:
E ( x)  A( x)  ei ( x )
1. Phase angle in radiant
2. Light path (OPL) in mm E ( x)  A( x)  eik DOPD ( x )
3. Light path scaled in l E ( x)  A( x)  e 2 iW ( x )
 58

Relationship to Transverse Aberration

 Relation between wave and transverse aberration
 Approximation for small aberrations and small aperture angles u
 Ideal wavefront, reference sphere: Wideal
 Real wavefront: Wreal
 Finite difference DW  W  Wreal  Wideal
W
 Angle difference   tan  
y p
 Transverse aberration D y'   R  
W D y' D y'
 Limiting representation  
 yp R W R
yp
reference sphere
reference
wave front W(yp)
plane


R, ideal ray

u C z
D y'
real ray
Wave Aberration in Optical Systems

 Definition of optical path length in an optical system:

Reference sphere around the ideal object point through the center of the pupil
 Chief ray serves as reference
Difference of OPL : optical path difference OPD
 Practical calculation: discrete sampling of the pupil area,
real wave surface represented as matrix
y yp y'p y'

wave
aberration image
point
W
optical
upper chief
system
coma ray ray
w'
z
chief
ray wave
front
object lower coma reference
point ray sphere

Object plane Entrance pupil Exit plane Image plane

Op EnP ExP Ip
 60

Wave Aberration

 Definition of the peak valley value WPV

 Reference sphere corresponds to perfect imaging
 Rms-value is more relevant for performance evaluation

pv-value
of wave
aberration

image
plane
wave
aberration

phase front

reference
sphere
exit
aperture
 61

Pupil Sampling

 All rays start in one point in the object plane

 The entrance pupil is sampled equidistant
 In the exit pupil, the transferred grid
may be distorted
 In the image plane a spreaded spot
diagram is generated

object plane entrance pupil optical exit pupil image plane

point equidistant grid system transferred grid spot diagram

yo yp y'p y'

xo xp x'p x'

z
 62

Wave Aberration Criteria

 Mean quadratic wave deviation ( WRms , root mean square )

Wrms  W2  W 
2 1
AExP  Wx p , y p   Wmean x p , y p  2
dx p dy p
with pupil area
AExP   dxdy
 Peak valley value Wpv : largest difference


Wpv  max Wmax x p , y p Wmin x p , y p  
 General case with apodization:
weighting of local phase errors with intensity, relevance for psf formation

Wrms 
1
( w)  ExP
AExP
I x p , y p  W x p , y p   W ( w)
mean x p , y p 
 2
dx p dy p
 63

Tilt of Wavefront
W

 Change of reference sphere:

tilt by angle 
linear in yp DWtilt  n  y p 
 y
p

 Wave aberration
yp
due to transverse y'
aberration Dy‘ wave aberration
DW < 0

yp
DWtilt   D y'
RRe f reference
sphere
transverse
aberration
 Is the usual description D y'
of distortion z
 tilt angle

wave front

pupil image
plane plane
 64

Defocussing of Wavefront

Paraxial defocussing by Dz:

Change of wavefront

n  rp2 1
DWDef   2
 D z '   n  D z ' sin 2
u
2Rref 2
yp
y'
wave aberration
DW > 0

wave front

reference sphere D z'

defocus
pupil image
plane plane
 65

Expansion of the Wave Aberration

 Table as function of field and aperture

 Selection rules:
checkerboard filling of the matrix Image Primary
location aberrations /
Seidel
Field y Secondary
Spherical Coma Astigmatism aberrations
y0 y1 y2 y3 y4 y 5
5
y r cos y 3 r cos y r cos
Distortion r1 Tilt Distortion Distortion
primary secondary
r2 y 2 r 2 cos2 y 4 r 2 cos2
r 2
Defocus y2r2
Astig./Curvat. y4r 2
Aper-
ture y r 3 cos y 3 r 3 cos3
3
r r Coma
primary y 3 r 3 cos
r4 y 2 r 4 cos2
4
r Spherical
primary y2 r4
y r 5 cos
5
r Coma
secondary
r6
6
r Spherical
secondary
 66

Zernike Polynomials
 Expansion of wave aberration surface
into elementary functions / shapes m=+8
cos
n +7

W (r ,  )  nm n (r,)
c Z m
+6
n m n
+5

 Zernike functions are defined in circular +4

coordinates r,  +3

 sin (m ) for m  0

 +2


Z nm (r , )  Rnm (r )  cos (m ) for m  0 +1

1 for m  0

0

-1

 Ordering of the Zernike polynomials by -2

indices:
-3
n : radial
m : azimuthal, sin/cos -4

-5

 Mathematically orthonormal function

-6
on unit circle for a constant weighting
-7
function sin
-8

 Direct relation to primary aberration types

n= 0 1 2 3 4 5 6 7 8
 67

Zernike Polynomials

 Advantages of the Zernike polynomials:

1. usually good match of circular symmetry to most optical systems
2. de-coupling of coefficients due to orthogonality
3. stable numerical computation
4. direct measurement by interferometry possible
5. direct relation of lower orders to classical aberrations
6. optimale balancing of lower orders (e.g. best defocus for spherical aberration)
7. fast calculation of Wrms and Strehl ratio in approximation of Marechal

 Problems and disadvantages of the Zernike polynomials:

1. computation on discrete grids
2. non circular pupils often occur in practice
3. different conventions can be found, conversion is quite confusing
4. calculation not stable for very high orders
5. Zernike functions are no eigenfunctions of wave propagation,
if the measurement is not made exactly in the pupil, the coefficients are erroneous
 68

Zernike Polynomials: Fringe Convention

Nr Cartesian representation Circular representation

1 1 1
2 x r sin 
3 y r cos
2 2
4 2x +2y -1 2 r² - 1
5 2xy r² sin 2
2 2
6 y -x r² cos 2
2 2
( 3r - 2r ) sin 
3
7 ( 3x + 3 y - 2 ) x
2 2 3
8 ( 3x + 3 y - 2 ) y ( 3r - 2r ) cos
2 2 2 2 2 4
9 6 (x +y ) -6 (x +y ) +1 6r - 6r² + 1
2 2
10 ( 3y -x ) x r³ sin 3
2 2
11 ( y -3x ) y r³ cos 3
2 2 4
12 (4x +4y -3) 2xy ( 4r - 3r² ) sin 2
2 2 2 2 4
13 (4x +4y -3) (y - x ) ( 4r - 3r² ) cos 2
2 2 2 2 2
( 10r - 12r³ + 3r ) sin 
5
14 [10(x +y ) -12(x +y )+3] x
2 2 2 2 2 5
15 [10(x +y ) -12(x +y )+3] y ( 10r - 12r³ + 3r ) cos
2 2 3 2 2 2 2 2 6 4
16 20 (x +y ) - 30 (x +y ) + 12 (x +y ) - 1 20r - 30r + 12r² - 1
2 2 4
17 (y -x ) 4xy R sin 4
4 4 2 2 4
18 y +x -6x y R cos 4
 69

Orthogonality
 Expansion of the wave aberration on a circular area
n
W (r , )  c nm Z nm (r , )
n m  n

1 2
 1 m0 
  Z n (r,)Z n' (r,) drdr    nn' mm '
m m '*
 Orthonormality for Fringe 2(n  1)
0 0
convention
 nn'
1

 Rn ( r )  Rn' ( r ) r dr 
m m
 Orthogonality of radial functions
0
2( n  1)
1 2
 Determination of coefficients 2(n  1)
cnm    W (r ,) Z nm* (r ,)d rdr
 1   m 0  0 0

 Necessary requirements for orthogonality:

1. pupil shape circular
2. uniform illumination of pupil (corresponds to constant weighting)
3. no discretization effects (finite number of points, boundary)

 Orthogonality perturbed in reality by:

1. real non-circular boundary (vignetting)
2. apodization (laser illumination)
3. discretization (calculation by a discrete finite ray set)
,
 70

Zernikepolynomials: Different Conventions

 Different standardizations used concerning:
1. indexing
2. scaling / normalization
3. sign of coordinates (orientation for off-axis field points)

 Fringe - representation
1. CodeV, Zemax, interferometric test of surfaces
2. Standardization of the boundary to ±1
3. no additional prefactors in the polynomial
4. Indexing according to m (Azimuth), quadratic number terms have circular symmetry
5. coordinate system invariant in azimuth

 Standard - representation
- CodeV, Zemax, Born / Wolf
- Standardization of rms-value on ±1 (with prefactors), easy to calculate Strehl ratio
- coordinate system invariant in azimuth

 Original - Nijboer - representation

- Expansion: 1 k k n k n
W (r ,  )  a00 
2
 a
n 0
0n R 
0
n a
n 0 m 1
nm R cos(m )  
m
n b
n 0 m 1
nm Rnm sin(m )
nm nm
gerade gerade
- Standardization of rms-value on ±1
- coordinate system rotates in azimuth according to field point
 71

Balance of Lower Orders by Zernike Polynomials

 Mixing of lower orders to get the minimal Wrms

 Example spherical aberration:

1. Spherical 4th order according to W (rp )  6rp4  6rp2  1
Seidel
2. Additional quadratic expression:
Optimal defocussing for edge
correction W
3. Additional absolute term 4
Minimale value of Wrms 4th order
(Seidel)
3
 Special case of coma:
2 4th, 2nd and 0th order
Balancing by tilt contribution,
(Zernike) rms is minimal
corresponds to shift between peak 1
and centroid + +
rp
0 +1
_
-1
4th and 2nd order
_
-2
 72

Calculation of Zernike Polynomials

 Assumptions:
1. Pupil circular
2. Illumination homogeneous
3. Neglectible discretization effects /sampling, boundary)

 Direct computation by double integral:

1. Time consuming 1 2
1
2. Errors due to discrete boundary shape
3. Wrong for non circular areas
cj 
   W (r , ) Z
0 0
j *(r , ) d rdr

4. Independent coefficients
2
 LSQ-fit computation:  N 
1. Fast, all coefficients cj simultaneously obtained   i  j j i   min
i 1 
W  c Z ( r )
j 1 
2. Better total approximation
3. Non stable for different numbers of coefficients,
if number too low
 T

c Z Z 1 T
Z W


4. Stable for non circular shape of pupil

 
 Calculation by Fourier transform A(k , )   W ( r, )  e 2ir k
d 2r
r 1

N 2 1
1  
cnm 

 Al (k , ) U (k , )
l 0
*
nm U nm (k , )   Z nm ( r, )  e 2ir k
d 2r
 73

Performance Description by Zernike Expansion

 Vector of cj
linear sequence with runnin g index

cj

 Sorting by symmetry 0.5

0.4

0.3

0.2

0.1

-0.1

-0.2

m
0 circular 1 2 3 4 -1 -2 -3 -4
symmetric cos terms sin terms
m=0 m>0 m<0