Zernike Polynomial Definitions:

+3.0
+2.5 j
Zernike Polynomials Fringe Zernike Polynomials:
( s +2.0
(n, m) Zernike polynomials are often used to describe wavefront error on an optical system. Mathematically they are an infinite set of orthogonal The index numbers A and the pair g, f have the following properties:
Zj (ρ, θ) = Znm (ρ, θ) = Nnm Rn|m| (ρ)
sin (|m|θ) m < 0
where Nnm =
2(n + 1) +1.5 A functions over the unit disk. Any arbitrary function defined on a unit disk can be described as a linear combination of Zernike polynomials. $r %
cos (mθ) m ≥ 0 1 + δm,0 (g, f ) 1
+1.0 On the left are the definitions for the Zernike polynomials that follow the Optical Society of America (OSA) standard. On the right are the A = (g − 1)2 + f ; g = A− + 1; f = A − (g − 1)2
2
The index numbers j and the pair n, m have the following properties: +0.5
Fringe Zernike polynomials. Below is a tree of the index values for the Zernike polynomials. In the top of each box are the index values for
0.0 Index conversion between the Fringe Zernikes and the OSA Zernikes:
j ∈ (0, 1, 2, · · · , ∞); n ∈ (0, 1, 2, · · · , ∞); m ∈ (−n, −n+2, · · · , n−2, n); −0.5
the OSA standard Zernike polynomials. In the bottom of each box are the index values of the Fringe Zernike polynomials. Rows of constant
n are shown in blue and light blue. Groups of Fringe Zernike polynomials with constant g are red and light red. Below each index box is
 
√ −1.0 f −1
n = (g − 1) +
 
n(n + 2) + m −3 + 9 + 8j
j= ; n= ; m = 2j − n(n + 2); −1.5 a contour plot of the Zernike polynomial and a logarithmically scaled point spread function (PSF) that would result from an optical system 2
2 2   
−2.0 f −1
Relationships for the radial polynomial
|m|
Rn (ρ): −2.5
with wavefront error of the given Zernike polynomial. All the PSF plots have the same spatial extent. m = (−1)f +1 (g − 1) −
2
−3.0 m=0 n + |m|
n−|m| g= +1
X2
(−1)s (n − s)! 0 2
Rn|m|  ρn−2s
 
(ρ) = (0, 0) 1 2m + 1
The Fourier transform of the Zernike polynomials defined as
1  1
s! 2 (n + m) − s ! 2 (n − m) − s ! f = n − |m| + 1 + 1−
s=0
The Recursion relationship between three radial polynomials, 1 Z 2 |2m + 1|
1 1
Γ (n + 1) 2 F1 − 2 (|m| + n) , 2 (|m| − n) ; −n; ρ−2


1 h i (1, 1) Sj (κ, φ) = Snm (~κ) = F [Znm (~ρ)] = Znm (~ρ) e−2πi~κ·~ρ d~ρ
|m| |m|+1 |m|+1
Rn|m| (ρ) = ρn Rn (ρ) = (n + |m| + 2) Rn+1 (ρ) + (n − |m|) Rn−1 (ρ)
1 2ρ(n + 1)

1

Γ 2 (2 + n + m) Γ 2 (2 + n − m) is as follows where Jn (2πκ) is a Bessel function of the first kind A g f j Fringe Zernikes Aberration Name∗
m = −1 m=1
{|m|,0} connects the following indexes shown in their relative location for
Rn|m| (ρ) = (−1)(n−|m|)/2 ρ|m| P(n−|m|)/2 (1 − 2ρ2 ) 1 2
p  
|m| (n+|m|)/2 J (2πκ)
m > 0 on the Zernike tree. For m < 0 the mirror image of the 2(n + 1)i (−1) n+1
sin(|m|φ) m < 0 1 1 1 0 1 Piston

n=1

(1, −1) (1, 1)
 κ
Pn{α,β} (x) figure below is used. Any one of the polynomials can be written in

Where is a Jacobi polynomial √
  
J (2πκ)
(−1)n −α −β d
n h
α+n β+n
i terms of the other two. 3 2 Snm (κ, φ) = n + 1(−1)n/2 n+1κ m=0 2 2 1 2 ρ cos(θ) x-Tilt
{α,β} (2, 2) (2, 1)
Pn (x) = n (1 − x) (1 + x) (1 − x) (1 + x)

3 2 2 1 ρ sin(θ) y-Tilt
  
2 n! dxn (n − 1, |m| + 1)
 p J (2πκ)
 2(n + 1)im (−1)(n+m)/2
 n+1
cos(mφ) m>0
κ 4 2 3 4 2ρ2 − 1 Defocus
Orthogonality and integral relationships for Zernike polynomials: m = −2 m=2
Therefore:
Z 1 3 4 5 5 3 1 5 ρ2 cos(2θ) Astigmatism 0◦ or 90◦
1 (n, |m|) ∞

n=2
|m|
Rn|m| (ρ) Rn′ (ρ) ρ dρ = δn n′ (2, −2) (2, 0) (2, 2) X 6 3 2 3 ρ2 sin(2θ) Astigmatism ±45◦
2 (n + 1) F [f (ρ, θ)] = aj Sj (κ, φ)
0
6 4 5 7 3 3 8 (3ρ3 − 2ρ) cos(θ) x-Coma
Z 1 j=0
Jn+1 (ν) (3, 2) (2, 3) (3, 1) 8 3 4 7 (3ρ3 − 2ρ) sin(θ) y-Coma
Rn|m| (ρ) J|m| (νρ) ρ dρ = (−1)(n−|m|)/2 were
0 ν (n + 1, |m| + 1) 9 3 5 12 6ρ4 − 6ρ2 + 1 Primary Spherical
Z 2π Z 1 ∞
m = −3 m=3
′
X
Znm (ρ, θ) Znm′ (ρ, θ) ρ dρ dθ = πδn,n′ δm,m′ f (ρ, θ) = aj Zj (ρ, θ)
10 4 1 9 ρ3 cos(3θ)
0 0
i−n π iz cos θ
6 7 8 9 j=0
11 4 2 6 ρ3 sin(3θ)

n=3
Z
(3, −3) (3, −1) (3, 1) (3, 3)
Jn (z) = e cos(nθ)dθ The Zernike polynomials are complete over the unit disk, 12 4 3 13 (4ρ4 − 3ρ2 ) cos(2θ) Secondary Astigmatism 0◦ or 90◦
π 0
!2 11 8 7 10 13 4 4 11 (4ρ4 − 3ρ2 ) sin(2θ) Secondary Astigmatism ±45◦
Jn (−z) = (−z)n z −n Jn (z) Z 1 Z 2π N
X (4, 2) (3, 4) (3, 3) (4, 1)
14 4 5 18 (10ρ5 − 12ρ3 + 3ρ) cos(θ) Secondary x-Coma
lim f (ρ, θ) − aj Zj (ρ, θ) ρdρdθ = 0
N →∞ 0 0 15 4 6 17 (10ρ5 − 12ρ3 + 3ρ) sin(θ) Secondary y-Coma
j=0
Table of OSA Zernike Polynomials:
m = −4 m=4 16 4 7 24 20ρ6 − 30ρ4 + 12ρ2 − 1 Secondary Spherical
where f (ρ, θ) is an arbitrary function defined on the unit
The indexes, j, n and m are shown below along with disk and the expansion coefficients aj are 10 11 12 13 14

n=4
the main index for the Fringe Zernikes, A and the
(4, −4) (4, −2) (4, 0) (4, 2) (4, 4) 17 5 1 14 ρ4 cos(4θ)
1 1 2π
Z Z
18 5 2 10 ρ4 sin(4θ)
standard Code V Zernikes, jCV . aj = f (ρ, θ)Zj (ρ, θ)ρdρdθ 18 13 9 12 17
π 0 0 (5, 2) (4, 4) (3, 5) (4, 3) (5, 1) 19 5 3 19 (5ρ5 − 4ρ3 ) cos(3θ)
j n m A jCV OSA Zernikes 20 5 4 16 (5ρ5 − 4ρ3 ) sin(3θ)
Therefore: ∞
21 5 5 25 (15ρ6 − 20ρ4 + 6ρ2 ) cos(2θ) Tertiary Astigmatism 0◦ or 90◦
m = −5 m=5
X
0 0 0 1 1 1 f (ρ, θ) = aj Zj (ρ, θ) 22 5 6 23 (15ρ6 − 20ρ4 + 6ρ2 ) sin(2θ) Tertiary Astigmatism ±45◦
j=0 15 16 17 18 19 20 23 5 7 32 (35ρ7 − 60ρ5 + 30ρ3 − 4ρ) cos(θ) Tertiary x-Coma
n=5

(5, −5) (5, −3) (5, −1) (5, 1) (5, 3) (5, 5)

1 1 -1 3 3 2ρ sin(θ) 24 5 8 31 (35ρ7 − 60ρ5 + 30ρ3 − 4ρ) sin(θ) Tertiary y-Coma
2 1 1 2 2 2ρ cos(θ) 27 20 15 14 19 26 25 5 9 40 70ρ8 − 140ρ6 + 90ρ4 − 20ρ2 + 1 Tertiary Spherical
(6, 2) (5, 4) (4, 6) (4, 5) (5, 3) (6, 1)
√ 2
3 2 -2 6 6 26 6 1 20 ρ5 cos(5θ)
√6ρ sin(2θ)
4 2 0 4 5 2
√ 2 − 1)
3 (2ρ m = −6 m=6 27 6 2 15 ρ5 sin(5θ)
5 2 2 5 4 6ρ cos(2θ) 28 6 3 26 (6ρ6 − 5ρ4 ) cos(4θ)
21 22 23 24 25 26 27
n=6

(6, −6) (6, −4) (6, −2) (6, 0) (6, 2) (6, 4) (6, 6) 29 6 4 22 (6ρ6 − 5ρ4 ) sin(4θ)
√ 3 30 6 5 33 (21ρ7 − 30ρ5 + 10ρ3 ) cos(3θ)
6 3 -3 11 10 2√2ρ sin(3θ) 38 29 22 16 21 28 37
7 3 -1 8 9 2√2 (3ρ3 − 2ρ) sin(θ) (7, 2) (6, 4) (5, 6) (4, 7) (5, 5) (6, 3) (7, 1) 31 6 6 30 (21ρ7 − 30ρ5 + 10ρ3 ) sin(3θ)
32 6 7 41 (56ρ8 − 105ρ6 + 60ρ4 − 10ρ2 ) cos(2θ)
8 3 1 7 8 2√2 (3ρ3 − 2ρ) cos(θ)
33 6 8 39 (56ρ8 − 105ρ6 + 60ρ4 − 10ρ2 ) sin(2θ)
9 3 3 10 7 2 2ρ3 cos(3θ) m = −7 m=7 34 6 9 50 (126ρ9 − 280ρ7 + 210ρ5 − 60ρ3 + 5ρ) cos(θ)
√ 4 28 29 30 31 32 33 34 35 35 6 10 49 (126ρ9 − 280ρ7 + 210ρ5 − 60ρ3 + 5ρ) sin(θ)
n=7

10 4 -4 18 15 √10ρ sin(4θ) (7, −7) (7, −5) (7, −3) (7, −1) (7, 1) (7, 3) (7, 5) (7, 7)
36 6 11 60 252ρ10 − 630ρ8 + 560ρ6 − 210ρ4 + 30ρ2 − 1
4 2
11 4 -2 13 14 10 (4ρ − 3ρ ) sin(2θ)
12 4 0 9 13
√ 4 2 51 40 31 24 23 30 39 50
√5 (6ρ 4− 6ρ 2+ 1) (8, 2) (7, 4) (6, 6) (5, 8) (5, 7) (6, 5) (7, 3) (8, 1) 37 7 13 84 924ρ12 − 2772ρ10 + 3150ρ8 − 1680ρ6 + 420ρ4 − 42ρ2 + 1
13 4 2 12 12 √10 (4ρ − 3ρ ) cos(2θ) ∗
The words “orthonormal Zernike” are to be associated with these names.
14 4 4 17 11 10ρ4 cos(4θ)
m = −8 m=8
√ 5
15 5 -5 27 21 2√3ρ sin(5θ) 36 37 38 39 40 41 42 43 44
n=8

(8, −8) (8, −6) (8, −4) (8, −2) (8, 0) (8, 2) (8, 4) (8, 6) (8, 8)
16 5 -3 20 20 2√3 (5ρ5 − 4ρ3 ) sin(3θ)
17 5 -1 15 19 2√3 (10ρ5 − 12ρ3 + 3ρ) sin(θ) 66 53 42 33 25 32 41 52 65 The Fringe Zernike polynomials follow an indexing pattern that is
(9, 2) (8, 4) (7, 6) (6, 8) (5, 9) (6, 7) (7, 5) (8, 3) (9, 1) distinctly different than the indexing of the OSA standard Zernike
18 5 1 14 18 2√3 (10ρ5 − 12ρ3 + 3ρ) cos(θ)
polynomials. The Fringe Zernike polynomials are only the 37
19 5 3 19 17 2√3 (5ρ5 − 4ρ3 ) cos(3θ)
m = −9 m=9 that are shown in the table above. The numbering of the Fringe
20 5 5 26 16 2 3ρ5 cos(5θ) Zernike polynomial number 37 does not follow the same rules
√ 6 45 46 47 48 49 50 51 52 53 54 as the rest of them. This 37th Fringe Zernike polynomial
n=9

(9, −9) (9, −7) (9, −5) (9, −3) (9, −1) (9, 1) (9, 3) (9, 5) (9, 7) (9, 9)
21 6 -6 38 28 √14ρ sin(6θ) is shown in the tree as the red box on the bottom cen-
22 6 -4 29 27 √ 14 (6ρ6
− 5ρ 4
) sin(4θ) 83 68 55 44 35 34 43 54 67 82 ter. If one follows the indexing rule set up by the first
6 4 2 (10, 2) (9, 4) (8, 6) (7, 8) (6, 10) (6, 9) (7, 7) (8, 5) (9, 3) (10, 1)
23 6 -2 22 26 √14 (15ρ6 − 20ρ4 + 6ρ2 ) sin(2θ) 36 Fringe Zernike polynomials the generalized Fringe
24 6 0 16 25 √7 (20ρ 6− 30ρ 4+ 12ρ2 − 1) Zernike polynomials can be defined and are shown
25 6 2 21 24 √14 (15ρ6 − 20ρ + 6ρ ) cos(2θ) m = −10 m = 10 in the index tree as the green and light green
26 6 4 28 23 √14 (6ρ − 5ρ4 ) cos(4θ) 55 56 57 58 59 60 61 62 63 64 65 boxes. The Fringe Zernike polynomials with a
n = 10

27 6 6 37 22 14ρ6 cos(6θ) (10, −10) (10, −8) (10, −6) (10, −4) (10, −2) (10, 0) (10, 2) (10, 4) (10, 6) (10, 8) (10, 10) constant g form “V” shapes in the index tree.
102 85 70 57 46 36 45 56 69 84 101
(11, 2) (10, 4) (9, 6) (8, 8) (7, 10) (6, 11) (7, 9) (8, 7) (9, 5) (10, 3) (11, 1)

m = −11 m = 11
66 67 68 69 70 71 72 73 74 75 76 77
n = 11

(11, −11) (11, −9) (11, −7) (11, −5) (11, −3) (11, −1) (11, 1) (11, 3) (11, 5) (11, 7) (11, 9) (11, 11)

123 104 87 72 59 48 47 58 71 86 103 122

(12, 2) (11, 4) (10, 6) (9, 8) (8, 10) (7, 12) (7, 11) (8, 9) (9, 7) (10, 5) (11, 3) (12, 1)

m = −12 m = 12
78 79 80 81 82 83 84 85 86 87 88 89 90
n = 12

(12, −12) (12, −10) (12, −8) (12, −6) (12, −4) (12, −2) (12, 0) (12, 2) (12, 4) (12, 6) (12, 8) (12, 10) (12, 12)

146 125 106 89 74 61 49 37 60 73 88 105 124 145

(13, 2) (12, 4) (11, 6) (10, 8) (9, 10) (8, 12) (7, 13) (8, 11) (9, 9) (10, 7) (11, 5) (12, 3) (13, 1)
Made by Brad Paul (bradpaul@rochester.rr.com)
with Asymptote and LATEX 2010

fy
w w
MTF
2λf
− nλf

Pupil Function PSF

The Root Mean Square (RMS) of an arbitrary two dimensional The Pupil function P (x, y; λ, t) is a complex function of the form PSF for a diffraction-limited incoherent imaging system with a clear circular aperture is y y y ′
y′
function f (~x) over the domain D is given by  2 w w nλf
w
− 2πi − − λf nλf

P (x, y; λ, t) = T (x, y; λ, t)e−iφ(x,y;λ,t) = T (x, y; λ, t)e ψ(x,y;λ,t)
= T (x, y; λ, t)e−2πiW (x,y;λ,t)

πDr
πD 2  2J1 λf 
λ 2 2 n
sR 
f 2w w
f (~x)2 d~x PSF(r) = nλf
RMS [f (~x)] = DR
where T (x, y; λ, t) is a real function representing the optical transmission of the aperture, 4λf πDr w w
d~x λf w − 2λf
D 0 ≤ T (x, y; λ, t) ≤ 1 and φ(x, y; λ, t) is the wavefront relative to the Gaussian reference D
x z x′ w
In the special case where the domain of the function is a disk of sphere in radians, ψ(x, y; λ, t) in length or W (x, y; λ, t) in waves. The PSF is the scaled where D is the diameter, λ is the wavelength of light, and f is the focal length of the λf
radius R then the RMS of the function can be written as modules squared of the Fourier transform of the pupil function. system. The MTF of the system is λf
w w
D n
′ 2
′ ′
 ′
 ( h i
2
v p
u∞
uX ∞ PSF(x , y ; λ, t) = F{x,y} [P (λf x, λf y; λ, t)] (x , y )

π
cos−1 (ρn ) − ρn 1 − ρ2n 0 ≤ ρn ≤ 1 − w2 − nλf
MTF(ρn ) =
X
2w
RMS [f (~r)] = t hj a2j where f (~r) = f (ρR, θ) = aj Zj (ρR, θ) w
The OTF can be calculated from the PSF via a Fourier transform 0 ρn > 1 nλf
j=0 j=0 w

where The pupil grid is sampled from − w2 to w2 − wn at steps of wn as shown on the y axis. The wavefront
Where the orthogonality relationship for Zj (ρ, θ) is OTF(fx , fy ; λ, t) = F{x′ ,y′ } [PSF(x′ , y ′; λ, t)] (fx , fy ) λf is relative to the Gaussian reference sphere. The horizontal axis of the PSF is sampled as shown
ρn = ρ
D on the y ′ axis. The Fourier transform of the PSF is the OTF which is Hermitian because the PSF
Z 2π Z 1 When using FFT’s to perform the Fourier transforms one must ensure that the width of
Zj (ρ, θ)Zk (ρ, θ)ρdρdθ = πhj δj,k and ρ is the spacial frequency that produces non-zero MTF values over the range of is real. The MTF and PTF are shown at the far right. The horizontal axis of the OTF is sampled PTF
the pupil sampling grid is at least twice the size as the enclosing diameter of the pupil,
0 0 D
0 ≤ ρ ≤ λf . as shown on the fy axis.
w ≥ D. The size and sampling of the axis are shown in the figure to the far right.