Stigmatic Optics

IOP Series in Emerging Technologies in Optics and Photonics

Series Editor
R Barry Johnson a Senior Research Professor at Alabama A&M
University, has been involved for over 50 years in lens design,
optical systems design, electro-optical systems engineering, and
photonics. He has been a faculty member at three academic
institutions engaged in optics education and research, employed
by a number of companies, and provided consulting services.

Dr Johnson is an IOP Fellow, SPIE Fellow and Life Member, OSA Fellow, and was
the 1987 President of SPIE. He serves on the editorial board of Infrared Physics &
Technology and Advances in Optical Technologies. Dr Johnson has been awarded
many patents, has published numerous papers and several books and book chapters,
and was awarded the 2012 OSA/SPIE Joseph W Goodman Book Writing Award for
Lens Design Fundamentals, Second Edition. He is a perennial co-chair of the annual
SPIE Current Developments in Lens Design and Optical Engineering Conference.

Foreword
Until the 1960s, the field of optics was primarily concentrated in the classical areas of
photography, cameras, binoculars, telescopes, spectrometers, colorimeters, radio-
meters, etc. In the late 1960s, optics began to blossom with the advent of new types of
infrared detectors, liquid crystal displays (LCD), light emitting diodes (LED), charge
coupled devices (CCD), lasers, holography, fiber optics, new optical materials,
advances in optical and mechanical fabrication, new optical design programs, and
many more technologies. With the development of the LED, LCD, CCD and other
electo-optical devices, the term ‘photonics’ came into vogue in the 1980s to describe
the science of using light in development of new technologies and the performance of
a myriad of applications. Today, optics and photonics are truly pervasive throughout
society and new technologies are continuing to emerge. The objective of this series is
to provide students, researchers, and those who enjoy self-teaching with a wide-
ranging collection of books that each focus on a relevant topic in technologies and
application of optics and photonics. These books will provide knowledge to prepare
the reader to be better able to participate in these exciting areas now and in the future.
The title of this series is Emerging Technologies in Optics and Photonics where
‘emerging’ is taken to mean ‘coming into existence,’ ‘coming into maturity,’ and
‘coming into prominence.’ IOP Publishing and I hope that you find this Series of
significant value to you and your career.
 Stigmatic Optics
Rafael G González-Acuña
Tecnológico de Monterrey, Garza Sada 2501, Monterrey 64849, Mexico

Héctor A Chaparro-Romo
Independent researcher, Alcanfores 8, Lazaro Cárdenas, Tultitlán 54916, Mexico

IOP Publishing, Bristol, UK

ª IOP Publishing Ltd 2020

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording
or otherwise, without the prior permission of the publisher, or as expressly permitted by law or
under terms agreed with the appropriate rights organization. Multiple copying is permitted in
accordance with the terms of licences issued by the Copyright Licensing Agency, the Copyright
Clearance Centre and other reproduction rights organizations.

Permission to make use of IOP Publishing content other than as set out above may be sought
at [email protected].

Rafael G González-Acuña and Héctor A Chaparro-Romo have asserted their right to be identified
as the authors of this work in accordance with sections 77 and 78 of the Copyright, Designs and
Patents Act 1988.

A supplementary Mathematica file is available at http://iopscience.iop.org/book/978-0-7503-3463-1.

ISBN 978-0-7503-3463-1 (ebook)

ISBN 978-0-7503-3461-7 (print)
ISBN 978-0-7503-3464-8 (myPrint)
ISBN 978-0-7503-3462-4 (mobi)

DOI 10.1088/978-0-7503-3463-1

Version: 20200901

IOP ebooks

British Library Cataloguing-in-Publication Data: A catalogue record for this book is available
from the British Library.

Published by IOP Publishing, wholly owned by The Institute of Physics, London

IOP Publishing, Temple Circus, Temple Way, Bristol, BS1 6HG, UK

US Office: IOP Publishing, Inc., 190 North Independence Mall West, Suite 601, Philadelphia,
PA 19106, USA
 In memory of
Jóse Angel Andrade Reyna
&
Daniela Andrade Reyna
 Contents

Preface xii
Series Editor’s foreword xiii
Acknowledgements xv
Author biographies xvii

1 The Maxwell equations 1-1

1.1 Introduction 1-1
1.2 Lorentz force 1-1
1.3 Electric flux 1-3
1.4 The Gauss law 1-4
1.5 The Gauss law for magnetism 1-5
1.6 Faraday’s law 1-5
1.7 Ampère’s law 1-6
1.8 The wave equation 1-7
1.9 The speed and propagation of light 1-8
1.10 Refraction index 1-9
1.11 Electromagnetic waves 1-9
1.11.1 One-dimensional way 1-11
1.11.2 Spherical coordinates 1-12
1.12 End notes 1-13
Further reading 1-13

2 The eikonal equation 2-1

2.1 From the wave equation, through Helmholtz equation to end 2-1
with the eikonal equation
2.2 The eikonal equation 2-3
2.3 The ray equation 2-6
2.3.1 n as constant 2-7
2.3.2 n(r ⃗ ) as a function 2-8
2.4 The Snell law from eikonal 2-8
2.5 The Fermat principle from eikonal 2-11
2.6 End notes 2-12
Further reading 2-12

3 Calculus of variations 3-1

3.1 Calculus of variations 3-1

vii
 Stigmatic Optics

3.2 The Euler equation 3-1

3.3 Newton’s second law 3-4
3.4 End notes 3-9
Further reading 3-9

4 Optics of variations 4-1

4.1 Introduction 4-1
4.2 Lagrangian and Hamiltonian optics 4-2
4.3 Law of reflection 4-5
4.4 Law of refraction 4-7
4.5 The Fermat principle and Snell’s law 4-8
4.6 Malus–Dupin’s theorem 4-8
4.7 End notes 4-11
Further reading 4-11

5 Stigmatism and stigmatic reflective surfaces 5-1

5.1 Introduction 5-1
5.2 Aberrations 5-3
5.3 Conic mirrors 5-7
5.4 Elliptic mirror 5-8
5.5 Circular mirror 5-10
5.6 Hyperbolic mirror 5-13
5.7 Parabolic mirror 5-13
5.8 End notes 5-17
Further reading 5-17

6 Stigmatic refractive surfaces: the Cartesian ovals 6-1

6.1 Introduction 6-1
6.2 Stigmatic surfaces 6-1
6.2.1 Case I: ro = ri = 0, zo → − ∞ and zi = f 6-5
6.2.2 Case II: ro = ri = 0, zo = f and zi → − ∞ 6-6
6.3 Analytical stigmatic refractive surfaces 6-7
6.3.1 Case A: ro = ri = 0, zo → − ∞ and zi = f 6-7
6.3.2 Case B: ro = ri = 0, zo = f and zi → − ∞ 6-8
6.3.3 Case C: ro = ri = 0, zo = ∓f and zi = ± f 6-10
6.3.4 Case D: ro = ri = 0, zo = −αf and zi = + f 6-12
6.3.5 Case E: ro = ri = 0, zo = αf and zi = − f 6-14

viii
 Stigmatic Optics

6.4 Conclusions 6-17

Further reading 6-18

7 The general equation of the Cartesian oval 7-1

7.1 From Ibn Sahl to Rene Descartes 7-1
7.2 A generalized problem 7-1
7.3 Mathematical model 7-2
7.4 Illustrative examples 7-6
7.5 Collimated input rays 7-6
7.6 Illustrative examples 7-11
7.7 Collimated output rays 7-13
7.8 Illustrative examples 7-14
7.9 Reflective surface 7-14
7.9.1 Parabolic mirror 7-18
7.10 Illustrative examples 7-19
7.11 End notes 7-23
Further reading 7-23

8 The stigmatic lens generated by Cartesian ovals 8-1

8.1 Introduction 8-1
8.2 Mathematical model 8-1
8.3 Examples 8-4
8.4 Collector 8-4
8.5 Examples 8-6
8.6 Collimator 8-7
8.7 Examples 8-10
8.8 Single-lens telescope with Cartesian ovals 8-10
8.9 Example 8-12
8.10 End notes 8-15
Further reading 8-15

9 The general equation of the stigmatic lenses 9-1

9.1 Introduction 9-1
9.2 Finite object finite image 9-1
9.2.1 Fermat principle 9-2
9.2.2 Snell’s law 9-3

ix
 Stigmatic Optics

9.2.3 Solution 9-6

9.2.4 The eikonal of the stigmatic lens 9-11
9.2.5 Gallery 9-13
9.3 Stigmatic aspheric collector 9-13
9.3.1 The eikonal of the stigmatic collector 9-25
9.3.2 Gallery 9-27
9.4 Stigmatic aspheric collimator 9-27
9.4.1 The eikonal of the stigmatic collimator 9-34
9.4.2 Gallery 9-36
9.5 The single-lens telescope 9-36
9.5.1 The eikonal of the single-lens telescope 9-42
9.5.2 Gallery 9-45
9.6 End notes 9-45
Further reading 9-46

10 The stigmatic lens and the Cartesian ovals 10-1

10.1 Introduction 10-1
10.2 Comparison between the different stigmatic lenses made 10-2
by Cartesian ovals
10.3 Cartesian ovals in a parametric form 10-2
10.4 Cartesian ovals in an explicit form as a first surface and 10-4
general equation of stigmatic lenses
10.5 Cartesian ovals in a parametric form as a first surface and 10-6
general equation of stigmatic lenses
10.5.1 First surface 10-6
10.5.2 Second surface 10-8
10.6 Illustrative comparison 10-14
10.7 Cartesian ovals in a parametric form for an object 10-16
at minus infinity
10.8 Cartesian ovals in an explicit form for an object 10-18
at minus infinity
10.9 Cartesian ovals in a parametric form as a first surface 10-19
and general equation of stigmatic lenses for an object
at minus infinity
10.10 Illustrative comparison 10-21
10.11 Implications 10-23
10.12 End notes 10-24
Further reading 10-24

x
 Stigmatic Optics

11 Algorithms for stigmatic design 11-1

11.1 Programs for chapter 6 11-1
11.1.1 Case: real finite object—real finite image 11-1
11.1.2 Case: real infinity object—real finite image 11-4
11.1.3 Case: real infinity object—virtual finite image 11-6
11.1.4 Case: real finite object—virtual finite image 11-8
11.1.5 Case: real finite object—real infinite image 11-10
11.1.6 Case: virtual finite object—real infinite image 11-12
11.1.7 Case: virtual finite object—virtual finite image 11-14
11.2 Programs for chapter 7 11-16
11.2.1 Case 1: real finite object—real finite image 11-16
11.2.2 Case 2: real infinity object—real finite image 11-19
11.2.3 Case 3: real infinity object—virtual finite image 11-21
11.2.4 Case 4: real finite object—virtual finite image 11-23
11.2.5 Case 5: real finite object—real infinite image 11-25
11.2.6 Case 6: virtual finite object—real infinite image 11-28
11.2.7 Case 7: virtual finite object—real finite image 11-31
11.2.8 Case 8: virtual finite object—virtual finite image 11-34
11.2.9 Case 9: real infinite object—real infinite image 11-37
11.3 Programs for chapter 8 11-39
11.3.1 Case 1: real finite object—real finite image 11-39
11.3.2 Case 2: real infinity object—real finite image 11-42
11.3.3 Case 3: real infinity object—virtual finite image 11-45
11.3.4 Case 4: real finite object—virtual finite image 11-48
11.3.5 Case 5: real finite object—real infinite image 11-51
11.3.6 Case 6: virtual finite object—real infinite image 11-54
11.3.7 Case 7: virtual finite object—real finite image 11-57
11.3.8 Case 8: virtual finite object—virtual finite image 11-60
11.3.9 Case 9: real infinite object—real infinite image 11-63
11.4 Programs for chapter 9 11-66
11.4.1 Case 1: real finite object—real finite image 11-66
11.4.2 Case 2: real infinity object—real finite image 11-69
11.4.3 Case 3: real infinity object—virtual finite image 11-72
11.4.4 Case 4: real finite object—virtual finite image 11-75
11.4.5 Case 5: real finite object—real infinite image 11-78
11.4.6 Case 6: virtual finite object—real infinite image 11-81
11.4.7 Case 7: virtual finite object—real finite image 11-83
11.4.8 Case 8: virtual finite object—virtual finite image 11-86
11.4.9 Case 9: real infinite object—real infinite image 11-89

xi
 Preface

This treatise focuses on a particular concept of geometric optics, stigmatism.

Stigmatism refers to the image property of an optical system that focuses a single
point source in object space at a single point in image space. Two of these points are
called a stigmatic pair of the optical system.
The treatise starts from the foundations of stigmatism: Maxwellʼs equations, the
eikonal equation, the ray equation, the Fermat principle and Snellʼs law. Then we
study the most important stigmatic optical systems without any paraxial or third
order approximation or without any optimization process. These systems are the
conical mirrors, the Cartesian ovals and the stigmatic lenses.
Conical mirrors are studied step by step with clear examples.
In the case of the Cartesian ovals, two paradigms are studied. The first, the
Cartesian ovals are obtained by means of a polynomial series and the second by
means of a general equation of the Cartesian oval.
For stigmatic lenses, the case is studied when the two refractive surfaces are
Cartesian ovals. Then the general equation for stigmatic lenses is obtained.
Finally, the similarities of optical systems and their nature are studied.
It is recommended to read this treatise in order.
The Authors

xii
 Series Editor’s foreword

For over a millennium, scientists have attempted to create mirrors and lenses free of
spherical aberration that is the only monochromatic axial aberration. Geometrical
imaging of an axial point object to form a perfect axial point image is known as axial
stigmatic imaging. When an optical system produces a perfect image over the entire
field-of-view, it known as a stigmatic optical system. Over time, it was learned that
spherical aberration is a constant aberration over the entirety of the image surface.
For a long time, it has been known that certain forms of lenses provided axial
stigmatic imaging. For example, when the object is at infinity, a geometrically-
perfect point image can be formed by a lens having (i) an ellipsoidal front surface
and a plane rear surface or (ii) a plane front surface and hyperbolic rear surface. A
stigmatic image for finite conjugates (magnification < 0) can be formed by using a
pair of plano-hyperbolic lenses (having focal length of f1 and f2) with the plane
surface facing one another. The magnification is simply −f2 f1. An example of a
mirror forming a stigmatic image is a parabola with the object at infinity. Although
such lenses and mirrors suffer no spherical aberration, other aberrations such as
coma and astigmatism can be bothersome. Indeed, a fast parabola can become
useless for imaging due to coma.
It is generally understood that there is not a generalized closed-form solution to
the design of a singlet lens that is axially stigmatic, i.e., one that is free of spherical
aberration. The authors of this treatise, Stigmatic Optics, elegantly attack this
challenge to develop such a generalized closed-form solution. They begin the book
by presenting Maxwell’s equations describing the behavior of electromagnetic fields
and develop the eikonal equation that provides the basis for geometric ray
propagation equation and Snell’s Law. Next is provided the necessary mathematics
needed to understand their development of the equations describing the surfaces of
lenses having the property of axial stigmatic imaging. Optical systems utilizing
Cartesian ovals are comprehensively discussed and numerous examples are pro-
vided. Subsequently, they meticulously develop the general equations for designing
axial stigmatic lenes. The resultant surface shapes can be described by a combination
of a conventional shape and an aspheric shape, or both surfaces being aspheric.
These surfaces, in general, cannot be described by the well-known polynomial
aspheric equation commonly used in lens design computer programs. The authors
skillfully explain the development of their new aspheric equations and provide
numerous examples. Throughout the book, the authors have richly included
graphics that aid in clarification of their discussions. Readers will likely appreciate
that the authors included computer code for algorithms useful in computing axial
stigmatic designs.
Stigmatic Optics is an excellent book to gain an understanding of the basics of
optical imaging and the formulation of the new aspheric shapes for achieving axial
stigmatic imaging. To continue learning about this topic, readers are encouraged to
read Analytical Lens Design by these authors, along with Julio C Gutiérrez-Vega,
which also explores the development of aspherical-shaped surface(s) that create

xiii
 Stigmatic Optics

lenses which are aplanatic, i.e., free of both spherical aberration and linear coma.
One might ask the question: ‘what is the value of this rather esoteric approach to lens
design using closed-form solutions?’. My answer is that exploring closed-form
solutions can provide further insight into creating better optical systems, although
closed-form solutions of complex optical systems seems improbable. Also, recent
advances in manufacturing free-form surfaces makes possible the creation of lenses
incorporating the authors’ new aspheric shapes.
R Barry Johnson, FInstP, FOSA, FSPIE, HonSPIE
Series Editor, Emerging Technologies in Optics and Photonics
Huntsville, Alabama

xiv
 Acknowledgements

Acknowledgements of Rafael G González-Acuña

Almighty God, creator of the Universe, this book aims to honor your glory. Thank
you Lord for giving me the intelligence, the desire, the faith and the means to
complete it. Lord, please help me to be a faithful servant of your will. Help me to
deserve the promises of your son Jesus Christ, give me a sign, and show me the
way …
I want to thank my family.
I want to thank my mother Carmen Leticia Acuña Medellín and to my father
Rogelio González Cantú to my brothers Rolando and Rogelio,
Héctor Alejandro Chaparro-Romo, Comrade, once again! we did it comrade!
time rewards!
To Professor R Barry Johnson for your support, patience and fruitful discussions.
To Ashley Gasque and Robert Trevelyan for all the support!
I would also like to thank Yoshio Catillejos, Israel Meléndez, Gustavo Medina,
Daniel Lomas, Roberto Martinez, César López, Roberto Vera, Ileana Paulette
Zambrano, Eliel Guadarrama, Miguel Rojas, Joel Guerra, Homero Pérez, Mauricio
Arroyo, Esteban Lankenau, Alejandra Guajardo, Adad Yepiz, Erick Patiño,
Alberto Silva, Michelle C Rocha, Adrian Lozano, Luis Garza, Roberto Acuña,
Rogelio Acuña, Adriana Mabel Serrano, Maria Isabel González Villarreal, Sonia
Villarreal, Dr Job Mendoza, Dr Dorilian López, Dr Servando López, Dr Raúl
Aranda, Dr Benjamin Perez-Garcia, Dr Maximino Avendaño, Dr Genaro Zavala,
Dr Carlos Hinojosa, Dr Francisco Cuevas, Dr Rafael Torres, Dr Blas Manuel
Rodríguez Lara, Professor Reinhard Klette, Professor Alois Herkommer, Professor
Russell Chipman, Professor Simon Thibault and Stephen Wolfram.
To Professor Julio C Gutiérrez-Vega and Dr Bernardino Barrientos García, my
advisors during the PhD and masters degree, respectively.
I would also like to thank several institutions: Institute of Physics, Conacyt,
Instituto Tecnológico y de Estudios Superiores de Monterrey, Wolfram Research,
Centro de Investigaciones en Optica A.C., Auckland University of Technology,
Institut für Technische Optik at Universität Stuttgart, Universidad Abierta y a
Distancia de México, Universidad Yachay Tech and Oxford Immune Algorithmics.
Thank you very much!

Acknowledgements of Héctor Alejandro Chaparro-Romo

I want to thank the IOP family widely for this opportunity, the fact that you have
believed in our own abilities and merits comforts us greatly, the feat of consolidating
a project of authentic knowledge and intertwining it with the classic always brings
with it a challenge of unknown magnitude. Through this second work, what is
rightfully true is reaffirmed. Optical design will never be the same.
I am particularly grateful to Professor Dr R Barry Johnson, editor in chief of the
series, Ashley Gasque for all the support she gave us and Robert Trevelyan for his
diligent instructions, without them this project would still be in process.

xv
 Stigmatic Optics

Beyond family and friend thanks, I want to thank prospectively all those who will
enjoy this work, either designing optical systems or consuming all the technology
that at some future time will be developed based on this knowledge.
I hope with fervor that this final result is an incentive to improve the quality of life
for all, the greatest thanks is for those who, through their mastery of knowledge,
make a moment of existence something comforting and pleasant.
I thank you, dear readers, for the patience and the decision to read our work; in
fact the number of equations and their procedures are not trivial matters, however,
with discipline and responsibility they can be mastered with such ease that their
magnitude becomes ephemeral and despicable. Among the greatest difficulties that
we have had to face, Rafael, is the fact that currently there are very few people who
enjoy this knowledge—I hope that soon you can open debate with us and with this
continue creating a model of Nature with a concrete and consistent theoretical basis.
I thank all those institutions that support the full development of their popula-
tions, all those governments that try to use reason and avoid the repression of free
thought.
This book is a tribute to our time, nowadays the role of the scientist is frequently
confused in the teaching market, I thank all those who do not give up on their
scientific thinking and much less stop creating ideas.
Father, mother, brother and Rafael, thanks again for facing the challenge as a
team, without this link many projects would not have been possible, only a just goal
makes the difficulty lessen, we again achieved it.
Finally I thank all those people who will use this knowledge to build a better
humanity, a peaceful life and a reason for existence!

xvi
 Author biographies

Rafael G González-Acuña
Rafael G González-Acuña studied industrial physics engineering at
the Tecnológico de Monterrey and studied the masterʼs degree in
optomechatronics at Centro de investigaciones en Óptica, A.C. He
is currently studying his PhD at the Tecnológico de Monterrey. His
doctoral thesis focuses on the design of free spherical aberration
lenses. He is co-author of the solution to the problem of designing
bi-aspheric singlet lenses free of spherical aberration. He is
co-author of the book Analytical Lens Design (IOP Publishing).

Héctor A Chaparro-Romo
Héctor A Chaparro-Romo obtained his bachelorʼs in electronic
engineering at the Universidad Autónoma Metropolitana, and is
currently studying for a degree in Economics at the Universidad
Nacional Autónoma de México. He is co-author of the solution to
the problem of spherical aberration in lens design, he is also
co-author of several peer-reviewed scientific articles and a book on
analytical design of optical systems. Héctor is an independent and
self-employed researcher in his home office, where he fully focuses his capabilities in
the complex field of computer networks and the Internet, As a pioneer, his main goal
is to develop http://www.biaspheric.com as a reference portal for all those who want
to learn deeply the theory of optical design that works from the rigorous analytical
paradigm. He is co-author of the book Analytical Lens Design (IOP Publishing).

xvii
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 1
The Maxwell equations

In this chapter, we give a brief review of the Maxwell equations for electromagnetic
theory, after a concise explication, we obtain the step-by-step electromagnetic wave
equation. Maxwell equations are the fundamental basis for optical theory, and
therefore to the stigmatic optics discipline.

1.1 Introduction
In this chapter, we are going to review Maxwell’s equations. The main goal is to
get the wave equation. From, the wave equation we can get the eikonal equation.
From the eikonal equation, we can derive the concept of ray and set the bases of
geometrical optics.
The general idea of this book is to take Maxwell’s equations as axioms and their
implications as theorems. In this language, the wave equation would be a theorem, a
direct consequence of Maxwell’s equations. The equation of the eikonal, under one
approximation, is an implication of the wave equation and from it, we develop the
theory of stigmatism.
The purest and most exquisite branch of geometric optics is stigmatic optics, the
branch to which this book owes its name.

1.2 Lorentz force

Let’s start with the definition of the electric field. An electric field can be described as
a vector field in which a point electric charge of value q suffers the effects of an
electrical force F⃗ given by the following equation:
F⃗ = q E,⃗ (1.1)

where E⃗ is the electric field. Electric fields can be originated, from both electrical
charges and variable magnetic fields.

doi:10.1088/978-0-7503-3463-1ch1 1-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Electric fields can be positive or negative. They are positive if they are generated
by positive charges, and negative if they are generated by negative charges. Charges
with different sign are attracted and with similar charge repel each other. Since the
electric field is a vector space it can be represented as vectors, thus it is usually
represented as vector lines. The lines emerge from positive charges and end in
negative charges, as can be seen in figures 1.1–1.3.
A magnetic field is a vector field that specifies the magnetic influence of electric
charges in relative movement and magnetised materials. A charged that is moving

Figure 1.1. Charges with different sign are attracted.

Figure 1.2. Positive charges repel each other.

Figure 1.3. Negative charges repel each other.

1-2
 Stigmatic Optics

parallel to a current of other charges experiences a force perpendicular to its own

velocity described by
F⃗ = q v⃗ × B,⃗ (1.2)

where v⃗ is the velocity of the charge and B⃗ is the magnetic field. Please notice the
cross product in equation (1.2) describes that if the charge is moving along
the magnetic field B⃗ its force will be zero.
For a particle subjected to an electric field combined with a magnetic field, the
total electromagnetic force or Lorentz force on that particle is given by the
combination of equations (1.1) and (1.2),

F⃗ = q(E⃗ + v⃗ × B⃗ ). (1.3)

The Maxwell equations entirely describe the nature of the electromagnetic fields E⃗
and B⃗ . In the following sections, we are going to describe them briefly.

1.3 Electric flux

An initial concept needed to enter Maxwell’s equations entirely is electric flux. The
electric flux, or electrostatic flux, is a scalar quantity that expresses a measure of the
electric field that passes through a defined surface, or expressed in another way, is
the measure of the number of electric field lines that penetrate a surface.
The portion of electric flux d Φ E⃗ through an infinitesimal area da is given by,

d Φ E⃗ = E⃗ · n⃗ da . (1.4)

The electric field E⃗ is multiplied by the component of the area perpendicular to

the field. n⃗ is the normal unit vector of the infinitesimal area da.
The electric flux through a surface S is therefore expressed by the surface integral,

Φ E⃗ = ∫S E⃗ · n⃗ da, (1.5)

where E⃗ is the electric field and n⃗ da is the differential surface vector that
corresponds to each infinitesimal element of the entire surface S. Please see
figure 1.4.

Figure 1.4. Flux of an electric field through a surface. On the right the normal vector n⃗ of the surface is parallel
 
to the electric field E . On the left there is inclination on the surface. Thus, there is an angle between n⃗ and E .

1-3
 Stigmatic Optics

1.4 The Gauss law

We start with Gauss’s law. Although, there are many ways to express this law and
notation differs, the integral form of the Gauss law is customarily given the
following expression,
qin
∮S E⃗ · n⃗ da = ε0
, (1.6)

where n⃗ is the normal unit vector of the closed surface S, qin is the charge inside the
closed surface S and ε0 is a constant called the permittivity free space. In the
international system of units, where force is in newtons (N), distance in meters (m),
and charge in coulombs (C),
ε0 = 8.85 × 10−12 C2 N−1 m−2. (1.7)
First, let’s pay attention of the left side of equation (1.6). The left side of this
equation is the mathematical representation of the electric flux—the number of
electric field lines—crossing into a closed surface S. In the right side the total amount
of charge contained within that surface is divided by a constant called the
permittivity of free space. Therefore, what Gauss’s law tells us is an electric charge
produces an electric field, and the flux of that field passing through any closed
surface is proportional to the total charge inside the closed surface.
Let us assume that you have a closed surface S, where the shape and size of S are
arbitrary. If there is no charge inside S, then, the electric flux is zero. If there is a
positive charge inside S, then, the electric flux through the surface is positive. But, if
you add an equal amount of negative charge, thus the total amount of charge inside
S is zero, then, the electric flux again is zero.
There is another way to express Gauss’s law using the divergence theorem. The
form is the following expression,
ρ
∇ · E⃗ = , (1.8)
ε0

where ρ is the density of charge inside and ∇ is the nabla operator,

∂ ⃗ ∂ ⃗ ∂ ⃗
∇= i + j + k (1.9)
∂x ∂y ∂z

∇ · E⃗ is the divergence of the field E⃗ . The divergence of the vector field is a scalar
computation that indicates the tendency of the field to flow away from a point.
Hence, Gauss’s law tells us that the divergence of the field E⃗ is the density of charge
divided by the permittivity free space.
Here we limit ourselves to present this form of the Gauss law because the
derivation of the divergence theorem is beyond the scope of the book. For a more
detailed analysis of the divergence theorem, the reader is invited to read the
references presented in the bibliography of this chapter.

1-4
 Stigmatic Optics

1.5 The Gauss law for magnetism

Gauss’s law for magnetism has the same structure as the Gauss law of the previous
section with the condition that there are no magnetic charges.
So, we start with the definition of magnetic flux through a surface S,

Φ B⃗ = ∫S B⃗ · n⃗ da, (1.10)

the magnetic field, B⃗ , multiplied by the component of the area perpendicular to the
field, where n⃗ is the unit normal vector of infinitesimal area da.
Therefore, over a closed surface S, Gauss’s law of magnetism is given by

∮S B⃗ · n⃗ da = 0, (1.11)

As we mentioned in the introduction, there are no magnetic charges. What

Gauss’s law of magnetism tells us is that the total magnetic flux passing through any
closed surface is zero.
Tacking the knowledge acquired in the last section, we can get the vector form of
Gauss’s law of magnetism, hence,

∇ · B⃗ = 0. (1.12)

This happens, as expected because there are no magnetic charges. Therefore, the
density of magnetic charge is zero.

1.6 Faraday’s law

Faraday’s electromagnetic induction law establishes that the electromotive force
induced in a closed circuit is directly proportional to the speed with which the
magnetic flux passing through any surface with the circuit as edge changes in time.
Thus,

∮c E⃗ · d l ⃗ = − dtd ∫S B⃗ · n⃗da, (1.13)

where E⃗ is the electric field, d l ⃗ is the infinitesimal element of the length of the circuit
represented by contour C, B⃗ is the magnetic field and S is an arbitrary surface, whose
edge is C. The right-hand rule gives the directions of contour C and n⃗da .
The electromotive force or induced voltage (represented by emf) is any cause
capable of maintaining a potential difference between two points in an open circuit
or of producing an electric current in a closed circuit.
According to the Stokes theorem, the differential form of Faraday’s law is
generally written as,

1-5
 Stigmatic Optics

∂B⃗
∇ × E⃗ = − , (1.14)
∂t

where ∇ × E⃗ is the curl of the electric field E⃗ . The curl operates on a vector field and
provides a vector result that designates the tendency of the field to circulate around a
point and the direction of the axis of greatest circulation.
What tells us the differential form of Faraday’s law is that a circulating electric
field is produced by a magnetic field that changes with time.

1.7 Ampère’s law

Ampère’s law, also called the Ampère–Maxwell law, is generally written in its
integral form as,

⎛ ⎞
∮c B⃗ · d l ⃗ = μ0⎜⎝Ienc + ε0 dtd ∫S E⃗ · n⃗da⎟⎠. (1.15)

The left side of equation (1.15) tells us about the circulation of the magnetic field
around a closed path C. On the right side, we have two elements that originate the
magnetic field. The first one is a steady current given by I enc . The other one is the
change in time of the electric flux through a surface bounded by C.
Please notice that in equation (1.15) the factor μ0 is a constant called the magnetic
permeability of free space. In the international system of units, where force is in
newtons (N) and current in amperes (A),
μ0 = 1.256 637 061 4 × 10−6 N A−2. (1.16)

Well, what equation (1.15) tells us is that an electric current or a changing electric
flux through a surface produces a circulating magnetic field around any path that
bounds that surface.
Now due the Stokes theorem, we can express the Ampère–Maxwell law as its
differential form,

⎛ ∂E⃗ ⎞
∇ × B⃗ = μ0⎜J⃗ + ε0 ⎟ , (1.17)
⎝ ∂t ⎠

The left side of the equation (1.17) is the circulating magnetic field. On the right
side are the sources of the circulating magnetic field. Notice that the first term in the
right side of equation (1.17), J⃗ is the current density vector. The second term on the
right side of the mentioned equation is the rate of change of the electric field with
time.
Therefore, what the Ampère–Maxwell law in its differential form tells us is that a
circulating magnetic field is produced by an electric current and by an electric field
that changes with time.

1-6
 Stigmatic Optics

1.8 The wave equation

We have briefly reviewed the Maxwell equations, but enough that from them, we
can obtain the wave equation. So we recall the set of Maxwell equations, equations
(1.8), (1.12) (1.14) and (1.17), as equation (1.18),
⎧ ρ
⎪∇ · E⃗ = ,
ε0
⎪
⎪∇ · B⃗ = 0,
⎪
⎪
⎨ ∂B⃗ (1.18)
⎪∇ × E⃗ = − ,
⎪ ∂t
⎪ ⎛
⎪∇ ∂E⃗ ⎞
× B⃗ = μ0⎜J⃗ + ε0 ⎟ ,
⎪
⎩ ⎝ ∂t ⎠

If we apply the curl on Faraday’s law, equation (1.14), we get,

⎛ ∂B⃗ ⎞ ∂∇ × B⃗
∇ × (∇ × E⃗ ) = ∇ × ⎜ − ⎟ = − . (1.19)
⎝ ∂t ⎠ ∂t

Now, we use the vector calculus identity expressed in equation (1.20),

∇ × (∇ × A⃗ ) = ∇(∇ · A⃗ ) − ∇2 A⃗ . (1.20)

Using the identify of equation (1.20), in equation (1.19),

∂∇ × B⃗
∇ × (∇ × E⃗ ) = ∇(∇ · E⃗ ) − ∇2 E⃗ = − . (1.21)
∂t
The last term of equation (1.21) can be replaced using Ampère’s law, equation
(1.17),
⎛ ∂E⃗ ⎞
∇ × B⃗ = μ0⎜J⃗ + ε0 ⎟ , (1.17)
⎝ ∂t ⎠

thus, replacing equation (1.17) in equation (1.21),

⎡ ⎛ ∂E⃗ ⎞⎤
∂⎢μ0⎜J⃗ + ε0 ⎟⎥
⎢⎣ ⎝ ∂t ⎠⎥⎦ (1.22)
∇ × (∇ × E⃗ ) = ∇(∇ · E⃗ ) − ∇2 E⃗ = − .
∂t
Now, with the Gauss law we can reformulate equation (1.22), thus let’s recall the
Gauss law,
ρ
∇ · E⃗ = , (1.9)
ε0

1-7
 Stigmatic Optics

replacing equation (1.8) in equation (1.22),

⎛ρ⎞ ∂J⃗ ∂ 2E⃗
∇⎜ ⎟ − ∇2 E⃗ = −μ0 − μ 0 ε0 2 . (1.23)
⎝ ε0 ⎠ ∂t ∂t

If we are working in a charge- and current-free region, ρ = 0 and J⃗ = 0, then

equation (1.23) becomes,

∂ 2E⃗
∇2 E⃗ = μ0ε0 . (1.24)
∂t 2

Equation (1.24) is called the wave equation. The wave equation is an important
second-order linear partial differential equation that describes the propagation of a
variety of waves, such as sound waves, light waves, and waves in water. It is important
in various fields such as acoustics, electromagnetism, quantum mechanics, and fluid
dynamics. The form represented in equation (1.24) is for an electric field E⃗ . But if we
apply ∇⃗× to Ampère’s law, then apply a similar procedure we can get,

∂ 2B⃗
∇2 B⃗ = μ0ε0 . (1.25)
∂t 2

Equation (1.25) is the wave equation for B⃗ . The process to get equation (1.25) is
left as an exercise to the reader.

1.9 The speed and propagation of light

The speed that a wave propagates is presented in the standard form of the wave
equation given by the following equation,
1 ∂ 2A⃗
∇2 A⃗ = 2 2 , (1.26)
v ∂t
where A⃗ is the wave and v2 is the speed of the wave. Therefore, we can take the speed
of the waves of equations (1.24) and (1.25) as,
1
= μ 0 ε0 . (1.27)
v2
Evaluating the values of the constants we get the speed of the electromagnetic waves
in the vacuum, which is given as
v = 2.9979 × 108 m s−1. (1.28)
Normally the above quantity is called speed of light and it is denoted by the letter c,
thus we set

c ≡ 2.9979 × 108 m s−1. (1.29)

Throughout the book we are going to use the notation presented in equation (1.29),

1-8
 Stigmatic Optics

1.10 Refraction index

When light is not travelling in a vacuum its speed is modified by the refraction index
of the medium in which it is travelling.
The refraction index of a homogeneous medium is constant, and it is a
dimensionless number that describes how fast light travels through the medium.
The refraction index is defined by
c
n= , (1.30)
v

where, c is the speed of light in vacuum and v is the phase velocity of light in the
medium.

1.11 Electromagnetic waves

It is time to find the solution of the wave equation of the electric field, we start by
recalling equation (1.24),
∂ 2E⃗
∇2 E⃗ = μ0ε0 . (1.24)
∂t 2
The method that we are going to apply to solve it is called variable separation.
The variable separation method refers to a procedure to find a particular complete
solution for specific problems involving partial differential equations such as a series
whose terms are the product of functions that have separate variables. It is one of the
most productive methods in mathematical physics to find solutions to physical
problems using partial differential equations.
Therefore, we are going to take the electric field as a function by the multi-
plication of a part that only depends on the position vector r ⃗ and another function
that solely depends on scalar time t.
E⃗ ( r⃗ , t ) = R⃗ ( r⃗)T (t ) (1.31)

replacing equation (1.31) in the wave equation, equation (1.24),

∂ 2[R⃗ ( r⃗)T (t )]
∇2 [R⃗ ( r⃗)T (t )] − μ0ε0 = 0. (1.32)
∂t 2
Let us pause for a moment and set a simple notation to clear the procedure,
R⃗ ( r⃗) = R⃗ , T (t ) = T . (1.33)
∂
Notice that ∇ only affects R⃗ ( r )⃗ and ∂t
only affects T (t ), therefore equation (1.32)
becomes,
∇2 (R⃗ ) 1 ∂ 2T
− μ 0 ε0 = 0. (1.34)
R⃗ T ∂t 2

1-9
 Stigmatic Optics

Since terms are being differentiated by different independent variables, they must
be equal to the same constant,
∇2 (R⃗ ) 1 ∂ 2T
= μ 0 ε0 = −k 2. (1.35)
R⃗ T ∂t 2
This leads us to a time-dependent differential equation given, by
∂ 2T k 2T
2
+ = 0, (1.36)
∂t μ 0 ε0

where,
1
= μ 0 ε0 . (1.37)
c2
The solution of equation (1.36) is given by,
T (t ) = e−ikct . (1.38)
Therefore, we can write the electric field as,
E⃗ ( r⃗ , t ) = R⃗ ( r⃗)e−ikct (1.39)

where the spatial part of equation (1.39) is given by

∇2 R⃗ ( r⃗) + k 2 R⃗ ( r⃗) = 0. (1.40)

Equation (1.40) is the Helmholtz equation. The Helmholtz equation is a partial

differential equation which is the spatial part of the wave equation.
Where k is the wave number defined as the number of radians per unit distance,
we more often use
2π
k= (1.41)
λ
where λ is the wavelength. Thus, from the definition of the refraction index,
c
n = ⟹ w = kv′ (1.42)
v
where w is the angular frequency. When light enters a medium, the wavelength is
modified as,
λc
λ= (1.43)
n
where λc is the wavelength inside the medium. The wave number is also modified as,
k = kcn′ (1.44)
where kc is the wave number inside the medium.

1-10
 Stigmatic Optics

Recapitulating, we are working with a three-dimensional vector field such that,

E⃗ ( r⃗ , t ) = E⃗ (x , y , z , t ) = Ex(x , y , z , t )eˆ x + E y(x , y , z , t )eˆ y

(1.45)
+ Ez(x , y , z , t )eˆ z .

Therefore, solving the Helmholtz equation is not the simplest task. It depends on
the dimension of the wave, whether in one, two or three dimensions, and in the
coordinate system. The coordinate system affects ∇; ∇ has different expressions for
different coordinate systems. In the next subsection, we explore some particular
solutions of the Helmholtz equation.

1.11.1 One-dimensional way

We start the one-dimensional case, we assume that there is electric filed in the z
direction.
E⃗ ( r⃗ , t ) = E⃗ z(z )e−ikct . (1.46)

Replacing equation (1.46) in the Helmholtz equation,

∂ 2E⃗ z
+ k 2 E⃗ z = 0. (1.47)
∂z 2
The solution of equation is given by
E⃗ z = Ae ikz + Be−ikz . (1.48)

Therefore, E⃗ ( r ⃗, t ) is given by

E⃗ ( r⃗ , t ) = (Ae ikz + Be−ikz )e−ikvt = Ae i (kz−kvt ) + Be i (kz+kvt ) (1.49)

simplifying,
E⃗ ( r⃗ , t ) = A cos(kz − kvt ) + B cos(kz + kvt ). (1.50)

The minus sign of the first cosine means that the wave is travelling to the right of
positive z. The plus sign of the second cosine implies that the wave is moving to the
right of negative z.
Also, notice that the time is being multiplied by the angular frequency,
w = kv. (1.51)
Notice that v is the speed inside a medium and c is the speed in vacuum.
Therefore, if we pick the wave that is traveling to positive z, E⃗ ( r ⃗, t ) is given by

E⃗ ( r⃗ , t ) = E 0 cos(kz − wt ) (1.52)

where we set A → E0. The last expression is the equation of the plane wave.

1-11
 Stigmatic Optics

1.11.2 Spherical coordinates

Now let’s pay attention to the Helmholtz equation spherical coordinates. First let’s
recall the Helmholtz equation,
∇2 E⃗ ( r⃗) + k 2 E⃗ ( r⃗) = 0. (1.53)

hence, in spherical coordinates the Helmholtz equation is expressed as,

1 ∂ ⎛ 2 ∂E⃗ ( r⃗) ⎞
⎜r ⎟ + k 2 E⃗ ( r⃗) = 0. (1.54)
r 2 ∂r ⎝ ∂r ⎠

To solve it assume that E⃗ ( r ⃗) has the following form,

E ′(r )
E⃗ ( r⃗) = eˆ r (1.55)
r
where E ′(r ) is a function of r. Thus replacing equation (1.55) in equation (1.54),

1 ∂ ⎡ 2⎛ E ′ ∂E ′ ⎞⎤ E′
⎢r ⎜ − 2 + r ⎟⎥ + k 2 =0 (1.56)
r ∂r ⎣ ⎝ r
2
∂r ⎠⎦ r
expanding,
1 ⎛ ∂E ′ ∂ 2E ′ ∂E ′ ⎞ E′
⎜−
2⎝
+r 2 + ⎟ + k2 =0 (1.57)
r ∂r ∂r ∂r ⎠ r

simplify,
1 ∂ 2E ′ E′
2
+ k2 = 0. (1.58)
r ∂r r
Notice, that is the same equation that we solved in the previous section. Therefore,
we can conclude that the solution of the wave equation in spherical coordinates has
the following form,
E
E⃗ ( r⃗ , t ) = 0 cos(kz − wt ). (1.59)
r
Notice that the amplitude of the wave decreases as r → ∞.
As an exercise to the reader, please study the Helmholtz equation in cylindrical
coordinates. The Helmholtz equation in cylindrical coordinates is the following
expression,

1 ∂ ⎡ ∂E⃗ ( r⃗) ⎤
⎢r ⎥ + k 2 E⃗ ( r⃗) = 0. (1.60)
r ∂r ⎣ ∂r ⎦

1-12
 Stigmatic Optics

1.12 End notes

In this chapter, we briefly studied Maxwell’s equations, from which we find the wave
equation. From the latter, we obtained some particular solutions and their spatial
part—the Helmholtz equation.
The Helmholtz equation will be of great help to us because through it we will find
the eikonal equation and in turn, the ray equation. These last equations lay the
foundations of geometric optics. Geometric optics is the playing field of stigmatism
which will be presented in-depth in chapter 5 and the following chapters; stigmatic
systems will be studied in detail.

Further reading
Arfken G B and Weber H J 1999 Mathematical Methods for Physicists (New York: Academic)
Boas M L 2006 Mathematical Methods in the Physical Sciences (New York: Wiley)
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Buchdahl H A 1993 An Introduction to Hamiltonian Optics (Chelmsford, MA: Courier
Corporation)
Campbell L 1882 The Life of James Clerk Maxwell (London: Macmillan)
Fleisch D 2008 A Student’s Guide to Maxwell’s equations (Cambridge: Cambridge University
Press)
Goodman J W 2005 Introduction to Fourier Optics (Greenwood Village, CO: Roberts)
Griffiths D J 2005 Introduction to Electrodynamics (Cambridge: Cambridge University Press)
Hecht E 1974 Schaum’s Outline of Optics (New York: McGraw-Hill)
Hecht E 2012 Optics (Cambridge, MA: Pearson)
Jackson J D 1999 Classical Electrodynamics (Hoboken, NJ: Wiley)
Lakshminarayanan V, Ghatak A and Thyagarajan K 2002 Lagrangian Optics (Berlin: Springer)
Lax M, Louisell W H and McKnight W B 1975 From Maxwell to paraxial wave optics Phys. Rev.
A 11 1365
Luneburg R K 1964 Mathematical Theory of Optics (Berkeley, CA: University of California
Press)
Mahon B 2004 The Man Who Changed Everything: The Life of James Clerk Maxwell (New York:
Wiley)
Maxwell J C 1990 The Scientific Letters and Papers of James Clerk Maxwell: 1846–1862
(Cambridge: Cambridge University Press)
Perko L 2013 Differential equations and Dynamical Systems vol 7 (Berlin: Springer)
Ronchi V and Barocas V 1970 The Nature of Light: An Historical Survey (Cambridge, MA:
Harvard University Press)
Zill D G 2016 Differential equations with Boundary-value Problems (Boston, MA: Cengage)

1-13
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 2
The eikonal equation

In this chapter we study the fundamental equation of geometric optics, the eikonal
equation. The eikonal equation is derived from the wave equation under the
circumstances studied in this chapter. Also in this chapter, we will study the direct
implications of the eikonal, such as the ray equation.

2.1 From the wave equation, through Helmholtz equation to end with
the eikonal equation
In the first chapter of this book, chapter 1, we had three objectives. The first one was
to make a summary of Maxwell’s equations. The second was to see that when we do
not have charges or currents, we can derive the wave equation from Maxwell’s
equations, and finally we wanted to study some particular solutions to the wave
equation. To do this, we had to obtain as an intermediate step the Helmholtz
equation, which is the spatial part of the wave equation.
In this chapter we ask ourselves: if Maxwell’s equations are our axioms, they are
true, and we do not doubt that they are; then, if the wave equation and Helmholtz’s
equation are implications that Maxwell’s equations are true, they are our theorems,
what else is true? What more effects do the wave equation and the Helmholtz
equation have and in which circumstances? The goal of this chapter is to obtain from
the wave equation and through the Helmholtz equation, the equation of the eikonal
equation. The eikonal equation is an implication of the equations already mentioned
in very particular but useful circumstances that define the geometric optics branch of
optics.
The eikonal equation is a partial differential equation with non-linearity found in
wave propagation. It is an approximated version of the wave equation.
We will study the spatial part of the wave equation, the Helmholtz equation and
given certain approximations. We will arrive at the eikonal equation.
Later we will see that from the eikonal equation, we can formulate the equation of
the ray and with it the concept of ray—a fundamental concept in stigmatic optics.

doi:10.1088/978-0-7503-3463-1ch2 2-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

It is imperative to understand these steps because they are the pillars of geometric
optics and stigmatic optics.
The eikonal equation can be seen as the intermediary of two worlds, two
paradigms, wave optics and ray optics or commonly called geometric optics.
Geometrical optics, also named ray optics, is one of the oldest sciences. It studies
the light through geometry. Euclid proposed the first premise used in geometrical
optics in his book Optics. The premise is that light propagates in straight lines,
making a connection between the paths of light and the geometry proposed in his
masterpiece, Elements. Since Euclid’s Elements focus has been on the nature
of straight lines. Elements is a mathematical and geometric treatise that consists
of thirteen books. In the books lies the foundation of Euclidean geometry.
In geometric optics, it is common to describe the propagation of light in terms of
rays. The ray is a useful abstraction to approximate the paths along which light
propagates under certain circumstances. Euclid’s premise on the propagation of
light in straight lines is only valid when light travels around a homogeneous
medium. The refraction index of a homogeneous medium is constant, and it is a
dimensionless number that describes how fast light travels through the medium. All
these properties of light described by geometrical optics are inherited by the eikonal
equation.
Geometric optics does not deal with specific optical effects, such as diffraction
and interference. Diffraction is the term that comes from the Latin diffractus, which
means broken. the etymology refers to the phenomenon by which a wave can
contour an obstacle in its propagation, moving away from the behavior of rectilinear
rays.
However, geometric optics deals with reflection and refraction, which are the two
main phenomena studied in geometrical optics. The reflection of light is the
phenomenon of returning the rays of light that fall on the surface of an object;
commonly, these objects are mirrors. Refraction is the redirection of a light ray
when it enters a medium where its speed is different. Refraction occurs when the
refractive index of the input medium is different from the refractive index of the
output medium.
Ignoring diffraction and interference is useful in practice when the wavelength is
small compared to the size of the optical elements in which the light interacts. This
paradigm is particularly useful for describing geometric aspects of images, including,
mirrors, lenses, axicons and other optical devices. Geometrical optics is the leading
theory used in optical design. The art of designing cameras, microscopes, telescopes
and optical systems in general is called optical design.
Geometric optics is a very challenging science, where most of the results are
particular cases obtained using sophisticated optimization algorithms, for example,
the design of a specific camera or telescope. The hardness of geometrical optics
happens because the equations that model the light rays and their interactions with
the different media can be quite long, complicated, and in many cases, nonlinear.
Non-linearity has led optical design to resemble art rather than a scientific discipline.
The non-linearity presented in geometric optics is also due with the eikonal equation

2-2
 Stigmatic Optics

since as we mentioned before, the eikonal equation is a partial differential equation

with non-linearity found in wave propagation.

2.2 The eikonal equation

We start our journey with a wave such that no term-dependent term t is presented.
This wave stands for the following equation,
E⃗ ( r⃗) = A⃗ ( r⃗)e ikcg(r ⃗), (2.1)

where A⃗ ( r ⃗) is the amplitude of the wave and it is multiplied by an exponential part,

something that should not be surprising according to the results obtained in the
previous chapter. This exponential part has as an exponent the product of
the complex number i, the wave number kc and a function g ( r )⃗ that depends on
the vector r ⃗ how the waveform of equation (2.1), g ( r )⃗ is proportional to wavefront.
Hence, kc is given by the following expression,
2π
kc = , (2.2)
λc
where λc is the wave length inside a homogeneous medium.
Until now we have not written or talked much about A⃗ ( r ⃗) and g ( r )⃗ ; the idea is to
study how A⃗ ( r ⃗) and g ( r )⃗ should be so that equation (2.1) will be a solution to the
Helmholtz equation. So do not be trouble by what A⃗ ( r ⃗) and g ( r )⃗ are, we will only
assume that they are functions that have continuous derivatives with respect to x, y
and z.
The first step in studying the nature of amplitude A⃗ ( r ⃗) and the function to the
wavefront proportional g ( r )⃗ is to call the Helmholtz equation,
2
∇⃗ E⃗ ( r⃗) + k 2 E⃗ ( r⃗) = 0. (2.3)

In it, we will insert equation (2.1), note that the procedure is simple but long so
that we will do it in steps. First, we will focus purely on the Laplacian that is present
in the Helmholtz equation. We begin with the derivative of E⃗ ( r ⃗) with respect to x,
and we have,
∂E⃗ ( r⃗) ∂g( r⃗) ∂A⃗ ( r⃗) ikcg(r ⃗)
= ikc A⃗ ( r⃗)e ikcg(r ⃗) + e . (2.4)
∂x ∂x ∂x
Now for the derivative respect to y we have,
∂E⃗ ( r⃗) ∂g( r⃗) ∂A⃗ ( r⃗) ikcg(r ⃗)
= ikc A⃗ ( r⃗)e ikcg(r ⃗) + e , (2.5)
∂y ∂y ∂y
and finally for the derivative with respect to z we have the same pattern, as expected,
∂E⃗ ( r⃗) ∂g( r⃗) ∂A⃗ ( r⃗) ikcg(r ⃗)
= ikc A⃗ ( r⃗)e ikcg(r ⃗) + e . (2.6)
∂z ∂z ∂z

2-3
 Stigmatic Optics

The next step is to derive E, again with respect to x, y, z, for which we will derive
equations (2.4)–(2.6) with respect to x, y, z, respectively. We start with the derivative
of equation (2.4) regarding x,

∂ 2E⃗ ( r⃗) ∂ 2A⃗ ( r⃗) ikcg(r ⃗) 2 ⃗

⎡
ikcg (r ⃗) ∂g( r ⃗)
⎤2
= e − k c A ( r ⃗ ) e ⎢ ⎥
∂x 2 ∂x 2 ⎣ ∂x ⎦
(2.7)
⎡ ∂ 2g( r⃗) ⃗ ⎤
ikcg (r )⃗ ∂A( r ⃗) ∂g( r ⃗) ⎥
+ i ⎢kc A⃗ ( r⃗)e ikcg(r ⃗) + 2 k ce ,
⎣ ∂x 2 ∂x ∂x ⎦

then, the derivative of equation (2.5) regarding y,

∂ 2E⃗ ( r⃗) ∂ 2A⃗ ( r⃗) ikcg(r ⃗) ⎡ ⎤2

2 ⃗ ikcg (r ⃗) ∂g( r ⃗)
= e − k c A ( r ⃗ ) e ⎢ ⎥
∂y 2 ∂y 2 ⎣ ∂y ⎦
(2.8)
⎡ ∂ 2g( r⃗) ⃗ ⎤
ikcg (r ⃗) ∂A( r ⃗) ∂g( r ⃗) ⎥
+ i ⎢kc A⃗ ( r⃗)e ikcg(r ⃗) + 2 k ce ,
⎣ ∂y 2 ∂y ∂y ⎦

finally, the derivative of equation (2.6) regarding z has the same pattern,

∂ 2E⃗ ( r⃗) ∂ 2A⃗ ( r⃗) ikcg(r ⃗) 2 ⃗

⎡
ikcg (r ⃗) ∂g( r ⃗)
⎤2
= e − k c A ( r ⃗ ) e ⎢ ⎥
∂z 2 ∂z 2 ⎣ ∂z ⎦
(2.9)
⎡ ∂ 2g( r⃗) ⃗ ⎤
ikcg (r )⃗ ∂A( r ⃗) ∂g( r ⃗) ⎥
+ i ⎢kc A⃗ ( r⃗)e ikcg(r ⃗) + 2 k ce ,
⎣ ∂z 2 ∂z ∂z ⎦

Now, we need to insert equations (2.7)–(2.9) in equation (2.3). This step is crucial
so please notice that we separate the real and imaginary components from equations
(2.7)–(2.9). Notice that the first real terms in equations (2.7)–(2.9) are proportional
to the Laplacian of A⃗ ( r ⃗).

∂ 2A⃗ ( r⃗) ikcg(r ⃗) ∂ 2A⃗ ( r⃗) ikcg(r ⃗) ∂ 2A⃗ ( r⃗) ikcg(r ⃗) ∇2 A⃗ ( r⃗) ⃗
e + e + e = E( r⃗). (2.10)
∂x 2 ∂y 2 ∂z 2 A⃗ ( r⃗)
The second real terms of the aforementioned equations are,
⎡ ∂g( r⃗) ⎤2 ⎡ ⎤2
ikcg (r ⃗) ∂g( r ⃗)
− kc2 A⃗ ( r⃗)e ikcg(r ⃗)⎢ ⎥ − 2 ⃗
kc A( r⃗)e ⎢ ⎥
⎣ ∂x ⎦ ⎣ ∂y ⎦
(2.11)
⎡ ∂g( r⃗) ⎤2
− kc2 A⃗ ( r⃗)e ikcg(r ⃗)⎢ = −kc2 E⃗ ( r⃗)∣∇g( r⃗)∣2 ,
⎥
⎣ ∂z ⎦

applying the same methodology to the imaginary terms and pull all together in
equation (2.3) we get the following non-easy solving expression that involves A⃗ ( r ⃗)
and g ( r )⃗ ,

2-4
 Stigmatic Optics

⎡ ∇2 A⃗ ( r⃗) ⎛ ∇A⃗ ( r⃗) · ∇g( r⃗) ⎞⎤⎥

⎢ − kc2∣∇g( r⃗)∣2 + kc2n 2 + i ⎜kc∇2 g( r⃗) + 2kc ⎟
⎢⎣ A⃗ ( r⃗) ⎝ A⃗ ( r⃗) ⎠⎥⎦ (2.12)
E⃗ ( r⃗) = 0.

From the last expression we eliminate the common factor of E⃗ ( r )⃗ , which inside
the bracket must be zero. Since it is a complex number it should be that the real part
is zero as well as the imaginary part. Let us focus on the real part of equation (2.12).
∇2 A⃗ ( r⃗)
− kc2∣∇g( r⃗)∣2 + kc2n 2 = 0. (2.13)
⃗
A( r⃗)

Equation (2.13) is still too complicated to obtain a general solution for A⃗ ( r ⃗) and
g ( r )⃗ . If we look at the imaginary part the scenario is even worse. Therefore, we need
to sacrifice one of A⃗ ( r ⃗) and g ( r )⃗ in order to just only study the consequence of the
other. If we eliminate g ( r )⃗ , we will have an answer similar to the one obtained in the
first chapter, chapter 1. So let’s explore the consequence of eliminating A⃗ ( r ⃗). We just
can’t eliminate A⃗ ( r ⃗) without justification since it is the amplitude, but if we observe if
A⃗ ( r ⃗) does not change a lot in respect to r ⃗ , then the second derivative of A⃗ ( r ⃗) with
respect to r ⃗ will be very small in comparison to A⃗ ( r ⃗), which leads to,

∇2 A⃗ ( r⃗)
A⃗ ( r⃗) ≫ ∇2 A⃗ ( r⃗), → 0. (2.14)
A⃗ ( r⃗)
This approximation is called slowly varying envelope approximation (SVEA) or
sometimes also called slowly varying amplitude approximation (SVAA). SVEA is
the assumption that the amplitude slowly varies, therefore the derivatives are small
enough that we can neglect their effect.
The slowly varying envelope approximation is often used because the resulting
equations are in many cases easier to solve than the original equations, reducing the
order of—all or some of—the highest-order partial derivatives. But the validity of
the assumptions which are made needs to be justified. In our case it is justified as
equation (2.14) holds.
Equation (2.13) under SVEA becomes,
−kc2∣∇g( r⃗)∣2 + kc2n 2 = 0, (2.15)

where kc can easily be dropped out, and we get the eikonal equation.
∣∇g( r⃗)∣ = n( r⃗). (2.16)

where g ( r )⃗ is a function of the position proportional to the wavefronts and related to

the optical path length, and n is the refractive index of the medium. The gradient of
g ( r )⃗ is directed along the normal to the surface g ( r )⃗ = constant. Therefore, the
eikonal function describes the constant-phase surfaces of a scalar wave field.

2-5
 Stigmatic Optics

Equation (2.14) is the fundamental assumption in the geometrical approach of

optics, which is that periods and wavelengths of light waves are much smaller than
the time and length scales on which the wave amplitude and the medium vary.
Under this approximation, the waves can be regarded locally as planar and
monochromatic, and consequently, the electromagnetic propagation can be sim-
plified to a model in which the light is transported along trajectories in space–time
called rays.
So far, it is natural to ask if the solutions of the wave equation that we know from
the previous chapter hold the SVEA condition of equation (2.14). As an exercise, we
take the plane wave,
E ( r⃗ , t ) = E 0 cos(kz − wt ). (2.17)
Notice that the amplitude of the plane wave is constant,
A⃗ ( r⃗) = E 0, (2.18)

therefore,
∇2 A⃗ ( r⃗)
= 0. (2.19)
A⃗ ( r⃗)
We conclude that the plane wave holds for SVEA. As an exercise to the reader check
if the spherical wave accomplishes SVEA.

2.3 The ray equation

We have obtained the equation of the eikonal. We know that the normal of g ( r )⃗ is
along the path of the light. Everything started because we took equation (2.1)
E⃗ ( r⃗) = A⃗ ( r⃗)e ikcg(r ⃗), (2.1)

and inserted the Helmholtz equation and we got the eikonal equation,
∣∇g( r⃗)∣ = n( r⃗). (2.16)
The question now is what we can infer from the eikonal and path of light? Taking
into account that the gradient of the eikonal function gives the direction of
propagation of the plane wave predicted by the eikonal equation and that, by
definition ∣∇g ( r )⃗ ∣ = n( r )⃗ , this is,
d r⃗ ∇g ( r ⃗ )
= . (2.20)
ds ∣∇g( r⃗)∣
Equivalently, taking into account that if r(⃗ s ) describes the equation of the ray
path parameterized by the arc length, then,
d r⃗
∇g ( r ⃗ ) = n = naˆ . (2.21)
ds

2-6
 Stigmatic Optics

See figure 2.1, and notice that the trajectory of light in purple is orthogonal to the
wavefront in blue. Please notice that we can write equation (2.21) as
∇ × ∇g( r⃗) = ∇ × (naˆ ) = 0. (2.22)

The curl of a gradient is always zero. Therefore, it can be shown that this equation
supports the representation,

d ⎛ d r⃗ ⎞
⎜n ⎟ = ∇n . (2.23)
ds ⎝ ds ⎠

This is the most standard form of the ray equation. The ray is the imaginary line that
represents the direction in which light propagates. The use of this model, widely
disclosed in geometric optics, simplifies the calculations thanks to SVEA. What
equation (2.16) and (2.23) tell us is that the ray path is perpendicular to the wave
front. In the next subsections we explore the cases when n is constant or a function,
which predicts equation (2.23).

2.3.1 n as constant
When in the ray equation we have n as a constant, we get
d ⎛ d r⃗ ⎞
∇n ⟹ ⎜n ⎟ = 0. (2.24)
ds ⎝ ds ⎠

Figure 2.1. The ray is the trajectory of the light under SVEA, in the picture it is colored in purple. The ray is
perpendicular to the wavefront g ( r ⃗), which is in blue.

2-7
 Stigmatic Optics

d
Applying the derivative operator ds
, we get,

d 2 r⃗
= 0. (2.25)
ds 2
If we integrate once,
d r⃗
= m⃗ (2.26)
ds
where m⃗ is constant vector. If we integrate twice,
r⃗ = m⃗ s + b⃗ (2.27)

the vector b⃗ can be seen as initial condition and m⃗ as the slope of the path ray.
The last expression is just a line. This means that when n is constant. The light
propagates in straight lines. This implication is what make SVEA so useful, since
working with straight lines is far away more mangle than working curved lines.

2.3.2 n(r ⃗ ) as a function

n( r )⃗ is a function; different trajectories may occur, giving rise to interesting
phenomena such as mirages, perfectly explainable with geometric optics and that
arise from the variation of the refractive index with temperature so that it can be
considered stratified. This is the origin, for example, of the effects that we have all
been able to see on roads on particularly hot days:
d ⎛ d r⃗ ⎞
⎜n ⎟ = ∇(n ). (2.23)
ds ⎝ ds ⎠
In this treatise, we limit ourselves to studying the light under constant refractive
index mediums, since the main topic of the book is stigmatism. Stigmatism will be
formally introduced in chapter 5.

2.4 The Snell law from eikonal

Another relevant implication of the ray equation is light passing from a medium
with refractive index n1 to another medium with refractive index n2. To show how
the ray path changes, we recall the eikonal equation in its vector form,
∇g( r⃗) = na⃗ , (2.28)
where if we integrate along a close loop in both sides, we have,

∮ ∇g(r⃗) · d l ⃗ = ∮ na⃗ · d l ⃗. (2.29)

Notice that the above integrals are equal to zero. This can be proven by the Stokes
theorem and equation (2.28).

∮∂S ∇g(r⃗) · d l ⃗ = ∫ ∫S ∇ × ∇g(r⃗) · dS = 0, (2.30)

2-8
 Stigmatic Optics

where ∂S is the border of the surface S. It is zero because the curl of a gradient is zero
∇ × ∇g ( r )⃗ = 0. Therefore, as a consequence we have,

∮ na⃗ · d l ⃗ = 0. (2.31)

Now we are in a position to answer the question presented in the begin of this
section. How does light behave when it passes from a medium with refractive index
n1 to another medium with refractive index n2? Please notice in figure 2.2, that there
is an entrance ray with direction a1⃗ and output a2⃗ . Therefore, if we compute the
closed integral over path presented in figure 2.2, taking the result of equation (2.31),
n2a2⃗ · H e1⃗ − n1a1⃗ · H e1⃗ = 0. (2.32)
Notice that we set h → 0. Manipulating equation (2.32), we have,
e1⃗ · (n2a2⃗ − n1a1⃗ ) = 0. (2.33)
Notice that the normal vector has the same direction as the unit vector of vertical
components, n⃗ = e⃗2 . Therefore,
n⃗ × (n2a2⃗ − n1a1⃗ ) = 0. (2.34)
From the last equation, we have two implications. The first one is
n2a2⃗ = n1a1⃗ , (2.35)
for this it is necessary that a1⃗ = a2⃗ , which implies that,
n2 = n1. (2.36)

Incident Ray-Light

Optical Surface

Normal Plane

Refracted
Ray-Light

Figure 2.2. Diagram of the refraction of a ray when it passes from one medium with refraction index n1 to
another of refraction index n2.

2-9
 Stigmatic Optics

This result means that the light does not pass into a change of refraction index.
The second implication is expressed in the following equation,
n⃗ × n2a2⃗ = n⃗ × n1a1⃗ , (2.37)
using the definition of the cross product.

n1 sin θ1 = n2 sin θ2. (2.38)

The last expression is called the Snell law. The Snell law is a formula used to
calculate the angle of refraction of light when crossing the separation surface
between two means of propagation of light with different refractive index. The name
comes from its discoverer, the Dutch mathematician Willebrord Snell van Royen
(1580–1626), although it was first discovered by Ibn Sahl in the year 984 AD.
Another way to express the second implication is that the n⃗ is parallel to
(n2a2⃗ − n1a1⃗ )
n⃗ × (n2a2⃗ − n1a1⃗ ) = 0, (2.39)
which implies that (n2a2⃗ − n1a1⃗ ) is along with n⃗ . Therefore, we have
(n2a2⃗ − n1a1⃗ ) = pn⃗ , (2.40)
where p is a constant. If we solve for it, we have,
p = (n2a2⃗ · n⃗ − n1a1⃗ · n⃗). (2.41)
Replacing equation (2.41) in equation (2.40) and dividing by n2,
n1 ⎛ n ⎞
a2⃗ = a1⃗ + ⎜a2⃗ · n⃗ − 1 a1⃗ · n⃗⎟n,⃗ (2.42)
n2 ⎝ n2 ⎠

and using the definition of the dot product we get,

n1 ⎛ n ⎞
a2⃗ = a1⃗ + ⎜cos θ2 − 1 cos θ1⎟n⃗ . (2.43)
n2 ⎝ n2 ⎠

In a more general way we can express equation (2.43) as,

n1 ⎛ n1 ⎞
a2⃗ = a1⃗ + ⎜ ∣n ⃗∣∣a1⃗ ∣cos θ1 − ∣n ⃗∣∣a2⃗ ∣cos θ2⎟n ⃗ . (2.44)
n2 ⎝ n2 ⎠

Looking at figure 2.2, we can see the vertical component of the refracted ray a2⃗ is
given by
n⃗ · a2⃗ = ∣n⃗∣∣a2⃗ ∣cos θ2. (2.45)
Notice that since they are unit vectors ∣n⃗∣ = ∣a2⃗ ∣ ≡ 1,
n⃗ · a2⃗ = cos θ2, (2.46)

2-10
 Stigmatic Optics

so let’s square it,

n12
(n⃗ · a2⃗ )2 = cos θ 22 = 1 − sin θ 22 = 1 − sin θ12. (2.47)
n 22
From the definition of the cross product we know,
n⃗ × a1⃗ = ∣n⃗∣∣a1⃗ ∣sin θ1(e1⃗ × e⃗2), (2.48)
also, ∣n⃗∣ = ∣a1⃗ ∣ = 1.
n⃗ × a1⃗ = sin θ1, (2.49)
powering to the square,
(n⃗ × a1⃗ )2 = sin θ12. (2.50)

Replacing the above expression in ( −n⃗ · a2⃗ )2 = 1 − (n12 /n 22 )sin θ12 , we get,

n12
n⃗ · a2⃗ = 1− (n⃗ × a1⃗ )2 . (2.51)
n 22

It is easy to see that the other terms of equation (2.44) are,

n1 n
[a1⃗ + ∣n⃗∣∣a1⃗ ∣cos θ1n⃗] = 1 [a1⃗ − (n⃗ · a1⃗ )n⃗]. (2.52)
n2 n2
Therefore, Snell’s law can be written as,

n1 n2
a2⃗ = [a1⃗ − (n⃗ · a1⃗ )n⃗] − n⃗ 1 − 12 (n⃗ × a1⃗ )2 for a2⃗ , a1⃗ , n⃗ ∈ 2. (2.53)
n2 n2

Equation (2.53) will be very useful in chapters 9 and 10 to express the stigmatic
lens equation.

2.5 The Fermat principle from eikonal

Something very interesting about any theory is that it obeys the principle of least
action. When I refer to this theory it is the equation of the eikonal, the ray equation
and all their implications. The principle of least action postulates that, for systems of
classical physics, the temporal evolution of all physical systems occurred in such a
way that an amount called ‘action’ tended to be the least possible or least energy
path. The optics version of the minimum principle is called the Fermat principle.
Fermat’s principle says that the path taken between two points by a ray of light is
the path that can be passed in the shortest time. Fermat’s principle can be stated as:

The optical length of the path followed by light between two fixed different
points is the global minima. The optical length is the physical length multiplied
by the refractive index of the medium

2-11
 Stigmatic Optics

Please pay attention to the word global minimum. Remember that the global
minimum is the smallest overall value of a set. So, imagine that we have a set, such
that its elements are all the possible optical paths from one point to another. These
paths have their respective optical length. What the Fermat principle says is that the
only physically valid path of our set is the one that has the smallest value of an
optical path length.
Mathematically, Fermat’s principle can be described as the time T a point of the
ray needs to cover a path between the points A and B, given by,
t1 t1 B
1 c ds 1
T= ∫t 0
dt =
c
∫t 0 v dt
dt =
c
∫A nds . (2.54)

Remember that c is the speed of light in vacuum, ds an infinitesimal displacement

along the ray, v = ds /dt the speed of light in a medium and n = c/v the refractive
index of that medium, t0 is the starting time, the ray is in A, t1 is the arrival time at
point B. The optical path length of a ray from a A to B is defined following integral
B
S= ∫A nds, (2.55)

related to the travel time by S = cT . The optical path length is a purely geometrical
quantity since time is not considered in its calculation. The global minimum in the
light travel time between two points A and B is equivalent to the global minimum of
the optical path length between A and B.
In the next chapter we are going to study the variational concepts behind the
Fermat principle. The in chapter 4 we are going to deduce the eikonal equation, ray
equation and their implications from the Fermat principle.

2.6 End notes

In this chapter, we obtained the basis for ray optics, the eikonal equation, and the
ray equation. From the ray equation, we obtained that at a constant refractive index,
average light travels in a straight line. From the eikonal, we obtained Snell’s law that
does not say how light changes the trajectory when it is in an interface of constant
refractive media.
Finally, we mention that all these acts obey the Fermat principle, which we briefly
describe. In the next chapter, we will see the mathematical foundations of the
Fermat principle.

Further reading
Arfken G B and Weber H J 1999 Mathematical Methods for Physicists (New York: Academic)
Boas M L 2006 Mathematical Methods in the Physical Sciences (New York: Wiley)
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Buchdahl H A 1993 An Introduction to Hamiltonian Optics (Chelmsford, MA: Courier
Corporation)

2-12
 Stigmatic Optics

Cao S and Greenhalgh S 1994 Finite-difference solution of the eikonal equation using an efficient,
first-arrival, wavefront tracking scheme Geophysics 59 632–43
Currier R and Herman M F 1985 Numerical comparison of generalized surface hopping, classical
analog, and self-consistent eikonal approximations for nonadiabatic scattering J. Chem.
Phys. 82 4509–16
Dacorogna B, Glowinski R and Pan T-W 2003 Numerical methods for the solution of a system of
eikonal equations with Dirichlet boundary conditions C. R. Math. 336 511–8
Griffiths D J 2005 Introduction to Electrodynamics (Cambridge: Cambridge University Press)
Hecht E 1974 Schaum’s Outline of Optics (New York: McGraw-Hill)
Hecht E 2012 Optics (Cambridge, MA: Pearson)
Hoffnagle J A and Shealy D L 2011 Refracting the k-function: Stavroudis’s solution to the eikonal
equation for multielement optical systems J. Opt. Soc. Am. A 28 1312–21
Huang L, Shu C-W and Zhang M 2008 Numerical boundary conditions for the fast sweeping high
order WENO methods for solving the Eikonal equation J. Comput. Math. 26 336–46
Hysing S-R and Turek S 2005 The eikonal equation: numerical efficiency versus algorithmic
complexity on quadrilateral grids Proc. of ALGORITMY vol 22
Jackson J D 1999 Classical Electrodynamics (Hoboken, NJ: Wiley)
Lakshminarayanan V, Ghatak A and Thyagarajan K 2002 Lagrangian Optics (Berlin: Springer)
Lax M, Louisell W H and McKnight W B 1975 From Maxwell to paraxial wave optics Phys. Rev.
A 11 1365
Luneburg R K 1964 Mathematical Theory of Optics (Berkeley, CA: University of California
Press)
Pegis R J 1961 I The modern development of Hamiltonian optics Prog. Opt. 1 1–29
Qian J, Zhang Y-T and Zhao H-K 2007 Fast sweeping methods for eikonal equations on
triangular meshes SIAM J. Numer. Anal. 45 83–107
Ronchi V and Barocas V 1970 The Nature of Light: An Historical Survey (Cambridge, MA:
Harvard University Press) pp 12–288
Spira A and Kimmel R 2004 An efficient solution to the eikonal equation on parametric manifolds
Interfaces Free Bound 6 315–27
Stavroudis O 2012 The Optics of Rays, Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 1995 The k function in geometrical optics and its relationship to the archetypal
wave front and the caustic surface J. Opt. Soc. Am. A 12 1010–6
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Hurtado-Ramos J B 2000 Maxwell equations and the k function J. Opt. Soc.
Am. A 17 1469–74
Vavryčuk V 2012 On numerically solving the complex eikonal equation using real ray-tracing
methods: A comparison with the exact analytical solution Geophysics 77 T109–16
Zhao H 2005 A fast sweeping method for eikonal equations Math. Comput. 74 603–27

2-13
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 3
Calculus of variations

In this chapter we will study the mathematical foundations that lie behind the
principle of least action that nature follows, the calculation of variations. The whole
theory of geometric optics can be formulated using the calculation of variations,
since from this the Fermat principle is deduced. As an example of the robustness of
the calculation of variations, we will obtain Newton’s second law from pure
variational concepts.

3.1 Calculus of variations

Calculus of variations studies the methods that allow finding stationary values of a
functional. The functional term is applied to certain functions. A functional is
a function that takes functions as its argument; that is, a function whose domain is a
set of functions. Stationary values include maximum values, minimum values, and
values of inflexion with a horizontal tangent.
Since a functional represents the mathematical model of a physical problem, the
application of variations methods is essential in areas of knowledge such as
theoretical physics, optics, Lagrangian mechanics, quantum mechanics, in engineer-
ing, etc.
The fundamental equation of calculus of variations is Euler’s equation. This
equation relates a functional with its stationary value. In this chapter, we will study
and obtain Euler’s equation. Later, we will demonstrate the robustness of the
calculus of variations by getting Newton’s second law from the variations principles.

3.2 The Euler equation

The problem is to find the y such that it makes stationary the following the integral,
x2
I= ∫x1
F (x , y , y′)dx (3.1)

doi:10.1088/978-0-7503-3463-1ch3 3-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

where F is a given functional, a function whose domain is a set of functions. y(x )

makes I a stationary value, y(x ) transforms I to a maximum or minimum. The
integral I can be seen as arc length integral. An arc length integral, also called curve
rectification, is the measure of the distance or path travelled along a curve or linear
dimension. But, we set I as the most general as possible. In our case, we are
interested in the shortest path, which is the light follows. Therefore, we are looking
at y(x ) which makes I a minimum.
Now, we consider a set of varied curves given by equation (3.1),
Y (x ) = y(x ) + ϵν(x ). (3.2)
Let ν(x ) be a function of x such that it is zero at x1 and x2, in other means
ν(x1) = ν(x2 ) = 0. Also, let ν(x ) have a continuous second derivative in the interval
(x1, x2 ). Finally, let ϵ be a completely arbitrary parameter, see figure 3.1. We assume
that Y (x ) is a set of single-valued curves. From the set of all of the single-valued
curves Y (x ), we are interested in only picking the curve that makes the following
integral a minimum.

Figure 3.1. Top: two paths given by the curves y(x ) and Y (x ) = y(x ) + ϵν(x ). y(x ) is such that I is a minimum
and Y (x ) is deflected from y(x ) by an amount ν(x ). Bottom: the deflection ν(x ) is zero at the initial and final
points.

3-2
 Stigmatic Optics

x2
I (ϵ ) = ∫x 1
F (x , Y , Y ′)dx . (3.3)

As is expressed in equation (3.3), we can see I as a function of ϵ, I (ϵ ). The ideal is

that if we set ϵ = 0, we have Y = y(x ) the desired extremal. In other words, what we
want to have is,
dI (ϵ )
= 0. (3.4)
dϵ ϵ=0

Differentiating equation (3.3) with respect to ϵ,

dI x2 ⎛ ∂F ∂Y ∂F ∂Y ⎞
dϵ
= ∫x 1
⎜
⎝ ∂Y ∂ϵ
+ ⎟dx.
∂Y ′ ∂ϵ′ ⎠
(3.5)

Notice that,
∂Y ∂Y ′
= ν(x ), = ν′(x ). (3.6)
∂ϵ ∂ϵ
replacing equation (3.6) in equation (3.5),
dI x2 ⎡ ∂F ∂F ⎤
dϵ
= ∫x
1
⎢⎣ ν(x ) +
∂Y ∂Y ′
ν′(x )⎥dx .
⎦
(3.7)

If we evaluate ϵ = 0, Y ∣ϵ=0 = y and equation (3.7) become,

⎛ dI ⎞ x2 ⎡ ∂F ∂F ⎤
⎜ ⎟ = ∫x ⎢ ν (x ) + ν′(x )⎥dx = 0. (3.8)
⎝ dϵ ⎠ϵ=0 1 ⎣ ∂y ∂y′ ⎦

Integrating the first term of the right side of equation (3.8),

d ⎛ ∂F ⎞
x2
x2 x2
∂F ∂F
∫x
1 ∂y′
ν′(x )dx =
∂y′
ν (x ) − ∫x 1
⎜ ⎟ν(x )dx .
dx ⎝ ∂y′ ⎠
(3.9)
x 1

Remember that we choose ν(x ) such that ν(x1) = ν(x2 ) = 0. Therefore, the first term
of the right side of equation (3.9) is zero. Replacing equation (3.9) in equation (3.8)
we have,
⎛ dI ⎞ x2 ⎡ ∂F d ∂F ⎤
⎜ ⎟ = ∫x ⎢ − ⎥ν(x )dx = 0. (3.10)
⎝ dϵ ⎠ϵ=0 1 ⎣ ∂y dx ∂y′ ⎦

There are two options here, to choose ν(x ) as zero or, what is inside the brackets of
equation (3.10) to be zero. If ν(x ) is zero, we gained nothing since ν(x ) = 0 does not
relate F and a y such that F is a minimum. Because in ν(x ) = 0 there is not even a
presence of F and y. Therefore, we choose the second option, and we get Euler’s
equation,

3-3
 Stigmatic Optics

∂F d ∂F
− = 0. (3.11)
∂y dx ∂y′

Every problem in the calculus of variations is solved by setting up the integral which
is to be stationary, addressing what the function F is, replacing it in the Euler
equation, and computing the solution of the differential equation obtained from the
Euler equation.
The idea is to use the Euler equation to see which is the path that light uses under
certain conditions.

3.3 Newton’s second law

In the previous chapter, chapter 2, we mentioned that historically the principle of
least action postulated, for systems of classical physics, makes that the temporal
evolution of the physical system as a whole occurr in such a way that a quantity
called action tended to be the least possible. We also mentioned that the principle
of least action in optics is known as the Fermat principle.
The calculus of variations and the principle of least action and the calculus of
variations are not only present in optics. To demonstrate the power of calculus of
variations and the principle of least action, in this section, we will deduce Newton’s
second law of calculus of variations and the principle of least action.
So, let’s assume that at time t1 a particle is in position x1, and at time t2 the same
particle is now in position x2. Which is the path of least action? We can consider all
paths. Some examples of paths can be seen in figure 3.2.
If we calculate the difference between kinetic energy and potential energy at each
point on the path, then adding all of these contributions together to get a final
number, we find the path for which this final number is the least. Then, we will know
that this is the path that the particle really takes.
We can formalize this by defining something called action S. The action is defined
as the time integral from t1 to t2 of the difference between the kinetic energy and the
potential energy,

Figure 3.2. Several paths from x1 to x2, starting at time t1 and finishing at time t2.

3-4
 Stigmatic Optics

t2
S= ∫t1
(K − P )dt . (3.12)

So the idea is to find a path where S is the minimum. But for the moment we are
going to explore the differences in single variable functions. Because we are going to
use the differences in single variable functions in the deduction of the path such that
S is the minimum.
If we consider a function f (x ) of a single variable x, then one way of knowing
whether we are located at the x value which corresponds to the minimum of f (x ) is
by moving x a little,
x → x + ϵ, (3.13)
so,
f (x ) → f (x + ϵ ), (3.14)
therefore, using Taylor series,
df d 2f
f (x + ϵ ) = f (x ) + ϵ + ϵ2 2 + ⋯ (3.15)
dx dx
then, the change of f is given by
Δf = f (x + ϵ ) − f (x ) (3.16)

df d 2f
Δf = f (x ) + ϵ + ϵ 2 2 + ⋯ − f (x ) (3.17)
dx dx

df d 2f
Δf = ϵ + ϵ2 2 + ⋯ (3.18)
dx dx
So, at minimum value, in other means df /dx = 0, the change of f is given by
d 2f
Δf = ϵ 2 +⋯ (3.19)
dx 2
So, we see at the minimum if we make a small change in our x position then our
function changes at second order in the small step changes, so Δf is proportional to ϵ 2 .

Δf = f (x + ϵ ) − f (x ) = 0. (3.20)

This means that Δf is zero at first order.

Now, it is important to realise that we are not dealing with the function of a single
variable. We are trying to find the minimum of a function, of a function. The task is
trying to find a single number to represent the point where it is least rather than in
our case where what we’re trying to do is to find an entire path for which action is
least, so we need to use a slightly different approach. So for example, we imagine the

3-5
 Stigmatic Optics

actual path being represented by the function x(t ). x(t ) represents the particle
moving from position x1 and t1 to x2 and t2, see figure 3.3.
What we want to do is look at some small deviation away from the actual path.
We hope that if we look at the change in the action as the result of that slight
deviation away from the actual path. Then, that change in the action should be zero
to first order in the change of the path.
More explicitly, if we deviate the actual path by some amount like the one
presented in figure 3.4, the amount changed is ν(t ). So, the actual path is x(t ) and the
deviated path is x(t ) + ν(t ).
Now we recall the action S, equation (3.12),
t2
S (x ) = ∫t1
(K − P )dt , (3.21)

if we are working in one dimension we have,

t2 ⎡ 1 ⎛ dx ⎞2 ⎤
S= ∫t ⎢ m⎜ ⎟ − ν(x )⎥dt , (3.22)
1 ⎣ 2 ⎝ dt ⎠ ⎦

where 12 m( dx
dt
)2 is the kinetic and ν(x ) is the potential energy.
What we are going to do is to look at how the action changes as a result of the
small deviation ν(t ) in the path x(t ). Then, we are going to look at the difference
between the actual path and the deviated path action.
If x is the path of minimum action then ΔS is zero at first order,
δS = S (x + ν ) − S (x ) = 0, (3.23)
thus,
t2 ⎡ 1 ⎛ d (x + ν ) ⎞2 ⎤
S (x + ν ) = ∫t ⎢ m⎜ ⎟ − V (x + ν )⎥dt , (3.24)
1 ⎣2 ⎝ dt ⎠ ⎦

where since the derivative is a linear operator, we have,

Figure 3.3. A particle moving from x1 in t1 to x2 in t2.

3-6
 Stigmatic Optics

Figure 3.4. The deflection ν(t ) in the path x(t ).

t2 ⎡ 1 ⎛ dx dν ⎞2 ⎤
S (x + ν ) = ∫t ⎢ m⎜ + ⎟ − V (x + ν )⎥dt . (3.25)
1 ⎣ 2 ⎝ dt dt ⎠ ⎦

Let’s focus on the squared term inside the integral of equation (3.25), expanding it,
⎛ dx dν ⎞2 ⎛ dx ⎞2 ⎛ dν ⎞2 dx dν
⎜ + ⎟ =⎜ ⎟ +⎜ ⎟ +2 , (3.26)
⎝ dt dt ⎠ ⎝ dt ⎠ ⎝ dt ⎠ dt dt
then, let’s focus on the second term inside the integral of equation (3.25), expanding
it using Taylor series,
dV ν 2 d 2V
V (x + ν ) = ν (x ) + ν + +⋯ (3.27)
dt 2! dt 2
we will ignore the higher terms,
dV
V (x + ν ) = ν (x ) + ν . (3.28)
dt
replacing equations (3.26) and (3.28) in equation (3.25),
t2 ⎡ 1 ⎛ dx ⎞2 1 ⎛ dx dν ⎞ dV ⎤
S (x + ν ) = ∫t ⎢ m⎜ ⎟ − ν(x ) + m⎜2 ⎟ − ν ⎥dt . (3.29)
1 ⎣ 2 ⎝ dt ⎠ 2 ⎝ dt dt ⎠ dx ⎦

Now, if we say that x is the path of minimum action, then ΔS is zero at first order.
Thus,
δS = S (x + ν ) − S (x ) = 0, (3.30)
replacing equations (3.22) and (3.29) in equation (3.30),
t2 ⎡ ⎛ dx dν ⎞ dV ⎤
δS = ∫t ⎢m⎜ ⎟ − ν ⎥dt = 0, (3.31)
1 ⎣ ⎝ dt dt ⎠ dx ⎦

3-7
 Stigmatic Optics

Now, look at equation (3.31) and notice that we have

d ⎛ dx ⎞ d 2x dx dn
⎜ ν⎟ = 2 ν + (3.32)
⎝
dt dt ⎠ dt dt dt
From it we can get an expression of
dx dn
,
dt dt

dx dn d ⎛ dx ⎞ d 2x
= ⎜ ν⎟ − 2 ν , (3.33)
dt dt dt ⎝ dt ⎠ dt
integrating from t1 to t2 equation (3.33),
t2
dx dn t2
d ⎛ dx ⎞ t2
d 2x
∫t dt = ∫t ⎜ ν⎟dt − ∫t νdt (3.34)
1 dt dt 1 dt ⎝ dt ⎠ 1 dt 2
Notice that the first term of equation (3.34) is zero by the fundamental theorem of
calculus and because we choose ν such that,
ν(t1) = ν(t2 ) = 0, (3.35)
thus,
d ⎛ dx ⎞
t2 t2
dx
ν =0⇒ ∫t ⎜ ν⎟dt = 0, (3.36)
dt t1 1 dt ⎝ dt ⎠

therefore, equation (3.34) can be expressed as,

t2 t2
dx dn d 2x
∫t 1 dt dt
dt = − ∫t 1 dt 2
νdt, (3.37)

replacing equation (3.37) in equation (3.31),

t2 ⎡ d 2x dV ⎤
δS = ∫t1
⎢−m 2 −
⎣ dt
⎥νdt = 0,
dx ⎦
(3.38)

Notice that what is inside the brackets of the bove equation should be zero, thus,
dV d 2x
− =m 2, (3.39)
dx dt
where the first term of the last equation is the force, since the force is the negative
derivative of potential energy and the second term is the mass multiplied by the
acceleration, thus,
F = ma. (3.40)
Remember we are in one dimension, thus the quantities of the last equation are
scalars rather than vectors.

3-8
 Stigmatic Optics

3.4 End notes

In this chapter, we studied the basic concepts of calculus of variations. Starting from
them, we find the Euler equation that relates its functional to its stationary value.
Finally, as a demonstration of the robustness and depth of the least action principle,
we obtained Newton’s second law using the calculus of variations method.
In the next chapter, we will use the calculus of variations techniques for our
purposes, studying the nature of light and stigmatism. The latter will be presented
concretely until chapter 5.

Further reading
Boas M L 2006 Mathematical Methods in the Physical Sciences (New York: Wiley)
Buchdahl 1993 An Introduction to Hamiltonian Optics (Chelmsford, MA: Courier Corporation)
del Castillo G F T 2018 An Introduction to Hamiltonian Mechanics (Berlin: Springer)
Elsgolc L D 2012 Calculus of Variations (Chelmsford, MA: Courier Corporation)
Gelfand I M et al 2000 Calculus of Variations (Chelmsford, MA: Courier Corporation)
Goldstein H, Poole C and Safko J 2002 Classical Mechanics (Reading, MA: Addison-Wesley)
Lakshminarayanan V, Ghatak A and Thyagarajan K 2002 Lagrangian Optics (Berlin: Springer)
Luneburg R K 1964 Mathematical Theory of Optics (Berkeley, CA: University of California
Press)
Marion J B 2013 Classical Dynamics of Particles and Systems (New York: Academic)
Morrey C B Jr 2009 Multiple Integrals in the Calculus of Variations (Berlin: Springer)
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)

3-9
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 4
Optics of variations

A study on the foundations of geometric optics with a purely variational perspective

is presented. Starting from the variational calculation and the Fermat principle, we
obtain the Lagrangian for light, then with the Euler equation we deduce the ray
equation. With the Fermat principle the laws of reflection and refraction are
obtained. Finally, the Malus–Dupin theorem is studied.

4.1 Introduction
In the first two chapters of this treatise, we studied Maxwell’s equations, and from
them, we arrived at the eikonal equation and the ray equation. All this formalism
used gave us a broader picture of how light can behave if we take the approx-
imations that justify the existence of the eikonal equation. Right at the end of
chapter 2, we mentioned that everything we find could be described under a single
principle, the Fermat principle. This principle, in turn, is based on the richness and
elegance of the principles of variational calculus.
In the previous chapter, chapter 3, we explored variational principles. We
discovered that using the Euler equation; we can solve many problems posed by
variational calculus. We even deduce Newton’s second law from the variational
calculus. The latter was only for deductive reasons, but this is a treatise of optics and
accurately a treatise of stigmatic optics. Although we will define stigmatism in the
next chapter, it is essential that in this study, you study the variational aspects of
light, see figure 4.1
So, let’s compute the optical path length of a light ray from one point to another;
the path length is given by,
p2
L= ∫p1
n(x , y , z )ds = 0, (4.1)

doi:10.1088/978-0-7503-3463-1ch4 4-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Figure 4.1. A path of light from point p1 to p2.

where L is the optical path length from the point p1 to the point p2 for a given path.
n(x , y, z ) is the refraction index along the path and s is a parameter of the arc length
of the path. Thus, ds is the infinitesimal arc length.
As we have mentioned before, the principle of least action transposed to optics
is the principle of Fermat. The Fermat principle tells us that light will travel the
least time path; this expressed in terms of variations can be seen as the following
integral,
p2
δL = δ ∫p 1
n(x , y , z )ds = 0, (4.2)

where δL can be seen as the variation on L. The light will travel in the path such that
the above integral is a minimum.
In this chapter, we are going to study this integral, and from it, we are going to
deduce everything that we know, from the behaviour of light: the eikonal, the ray
equation, Snell’s law and the perpendicularity of the rays and waves.

4.2 Lagrangian and Hamiltonian optics

From equation (4.2), we can write the Fermat principle as,
p2
δl = δ ∫p1
L(x , y , z , x′ , y′ , z′)dz = 0. (4.3)

L(x , y , z, x′, y′, z′) is called the Lagrangian, and it is the functional that receives as
input parameters the possible paths of light. The above integral being equal to zero
means that the path chosen by the light is the minimum. ds is the infinitesimal arc
length given by

ds = dz 1 + x ′2 + y ′2 . (4.4)
dx
We use dz since we choose z to be the integration variable. Thus, x′ = dz
and
dx
y′ = dz
. Therefore, we can rewrite equation (4.3) as,

4-2
 Stigmatic Optics

p2
δl = δ ∫p1
L(x , y , x′ , y′ , z )dz = 0, (4.5)

where the optical Lagrangian is,

L(x , y , x′ , y′ , z ) = n(x , y , z ) 1 + x ′2 + y ′2 . (4.6)

Notice that this form of the Lagrangian does not depend on z′ because we set z as the
independent variable. We choose it by tradition since typically z is set to be the
optical axis. Now the Lagrangian has five input parameters that can be divided into
three categories. The first one is x and its derivative x′, the second one is y, and its
derivative y′, and the third is the independent variable z. With this in mind, for this
system we can get two Euler equations, the first one for x and its derivative x′,
d ⎛ ∂L ⎞ ∂L
⎜ ⎟ − = 0, (4.7)
dz ⎝ ∂x′ ⎠ ∂x
and another one for y and its derivative y′

d ⎛ ∂L ⎞ ∂L
⎜ ⎟− = 0. (4.8)
dz ⎝ ∂y′ ⎠ ∂y

Notice that the role of z is similar to the variable of time in chapter 3, where we
found Newton’s second law.
Using equation (4.6), we can rewrite equation (4.7) as,
⎛ ⎞
d ⎜ nx ̇ ⎟= ∂n
⎜ ⎟ 1 + x 2̇ + y 2̇ , (4.9)
dz ⎝ 1 + x 2̇ + y 2̇ ⎠ ∂x

and equation (4.8) as,

⎛ ⎞
d ⎜ ny ̇ ⎟= ∂n
1 + x 2̇ + y 2̇ . (4.10)
dz ⎜⎝ 1 + x 2̇ + y 2̇ ⎟
⎠ ∂y

Now manipulating equation (4.9), we get,

⎛ ⎞
1 d ⎜ nx ̇ ⎟ = ∂n .
⎜ ⎟ ∂x (4.11)
1 + x 2̇ + y 2̇ dz ⎝ 1 + x 2̇ + y 2̇ ⎠

Doing the same process, but now in equation (4.10) we get,

⎛ ⎞
1 d ⎜ ny ̇ ⎟ = ∂n . (4.12)
2 dz ⎜ 2 ⎟
2
1 + ẋ + y ̇ ⎝ 1 + x ̇ + y ̇ ⎠ ∂y
2

Replacing equation (4.6) in equation (4.11),

4-3
 Stigmatic Optics

⎛ dx ⎞
d ⎜ dz ⎟ ∂n
n
⎜ ⎟= . (4.13)
ds ⎜ ds ⎟ ∂x
⎝ dz ⎠

The above expression is reduced to,

d ⎛ dx ⎞ ∂n
⎜n ⎟ = , (4.14)
ds ⎝ ds ⎠ ∂x
and we apply the same procedure on equation (4.12),
d ⎛ dy ⎞ ∂n
⎜n ⎟ = . (4.15)
ds ⎝ ds ⎠ ∂y
For the z direction we can also have a similar equation, since because we could set
ds = dz 1 + y ′2 + z ′2 where now the derivatives, inside the square root, are with
respect to x. Thus, we have,
d ⎛ dz ⎞ ∂n
⎜n ⎟ = , (4.16)
ds ⎝ ds ⎠ ∂z
combining equations (4.14)–(4.16), in vector form we have,

d ⎛ d r⃗ ⎞
⎜n ⎟ = ∇n , (4.17)
ds ⎝ ds ⎠

where r ⃗ = x eˆ 1 + y eˆ 1 + z eˆ 1. Equation (4.17) is the ray equation the same ray

equation that we found on chapter 2. Different origins, same result. Absolutely
astonishing and beautiful.
Well, there is more, Lagrangian mechanics has an alternative version, which is
also based on the calculus of variations, and it is called Hamiltonian mechanics.
Hamilton studied the Lagrangian mechanics in terms of what he called generalized
momenta. The generalized momentum in direction x is,
∂L
px = , (4.18)
∂x′
and the generalized momentum in direction y is,
∂L
py = . (4.19)
∂y′
Computing the derivative on equation (4.18), px can be written as,
x′ dx
px = n =n , (4.20)
1 + x ′2 + y ′2 ds

and computing the derivative of equation (4.19), for py we get,

4-4
 Stigmatic Optics

y′ dy
py = n =n . (4.21)
1 + x ′2 + y ′2 ds

Equations (4.20) and (4.21) are the generalized momenta. From them it is easy to see
that generalized momentum in a given direction is the refraction index multiplied by
the cosine director of the path of light in that given direction.
Making an analogy of the Hamiltonian in classical mechanics, the Hamiltonian
for the ray is given by
H (x , y , px , py , z ) = px x′ + py y′ − L. (4.22)

Replacing equations (4.20), (4.21) and (4.6), we have,

H (x , y , px , py , z ) = − n 2 − px2 − py2 , (4.23)

therefore, the equations of motion are the following,

⎧ ∂qx ∂H
⎪ =
⎪ ∂z ∂px
⎪ ∂qy ∂H
⎪ =
⎪ ∂z ∂py
⎨ , (4.24)
⎪ ∂px = −
∂H
⎪ ∂z ∂qx
⎪
⎪ ∂py ∂H
⎪ ∂z = −
⎩ ∂qy

where we take x → qx and y → qy .

Given the coordinates point (qx , qy ) and the generalized momenta (px , py ) it is
possible to solve the equations of motion for the ray position.
Lagrangian optics and Hamiltonian optics differ in certain aspects of notation but
without encapsulating the same behaviour of light, which is the behaviour of light
under SVEA, as the ray equation. In the next section, we will see more tangible
applications of the Fermat principle. The Fermat principle is the fundamental basis
of the concept called stigmatism and of all known stigmatic systems.

4.3 Law of reflection

In chapter 2, we find Snell’s law for refraction starting from implications of the
eikonal equation. Well as the name of this expression suggests, such an equation is
for when light is refracted, in the case of how much light is reflected. Reflection is the
change in the direction of a wavefront at an interface between two different media so
that the wavefront returns into the medium from which it originated.
Fermat’s principle predicts how the path of light will behave when it is reflected.
Observe figure 4.2; it shows an indecent ray of light on a surface. The surface reflects

4-5
 Stigmatic Optics

Incident Ray-Light
Reflected Ray-Light

Optical Surface

Normal Plane

Figure 4.2. Diagram of the reflection of light.

the ray. We can see if the ray starts from point (x1, y1), arrives in a straight line to the
point where it is reflected in the surface and then reaches point (x2, y2 ). Taking this
into account we can write the time that light takes from point (x1, y1) to point (x2, y2 )
as the following expression,
1⎡ ⎤
T (x ) = (x − x1)2 + y12 + (x2 − x )2 + y22 ⎦ , (4.25)
c⎣
where x is the place on the plane on the horizontal direction where the light strikes.
Now, we know that T (x ) is correct in the way for places with constant refractive
index the light moves in straight lines, all this from the ray equation.
But we do not know yet how the light reflects. The point (x , 0) is crucial here. The
value of x will tell us the slope of the straight lines. From the Fermat principle, we
know that the light moves in the least time. Thus, we can derive with respect to x and
equal to zero,
∂T (x )
= 0, (4.26)
∂x
computing the derivative,
x − x1 x2 − x
= . (4.27)
(x − x1)2 + y12 (x2 − x )2 + y22

From figure 4.2, we can see that the last equation can be re-formulated as,
sin θ1 = sin θ2. (4.28)
The last expression leads to,
θ1 = θ2. (4.29)

4-6
 Stigmatic Optics

Figure 4.2 can be seen as a simple diagram of a reflection taking place on a flat
mirror, the incident angle θ1 is equal to the reflected angle θ2 , in other words θ1 = θ2 .
We can show that it is a real minimum by computing the second derivative,
∂ 2T 1 − sin2 θ1 1 − sin2 θ2
c = +
∂x 2 (x − x1)2 + y12 (x2 − x )2 + y22
⎡ ⎤ (4.30)
= (1 − sin θ2 )⎢ ⎥.
2 1 1
+
⎢ (x − x ) 2 + y 2 (x2 − x )2 + y22 ⎥⎦
⎣ 1 1

Since sin θ22 ⩽ 1, the above expression is positive for all values of x, which means we
have an absolute minimum.

4.4 Law of refraction

Refraction occurs when light travels through a region that has a changing index of
refraction. The simplest case of refraction occurs when there is an interface between
a uniform medium with an index of refraction n1 and another medium with an index
of refraction n2. In such situations, Snell’s law, also called Law of Refraction,
describes the resulting deflection of the light ray. We get Snell’s law from the eikonal
equation in chapter 2. Now, in this section, we deduce it from Fermat’s principle.
In figure 4.3 we can see a simple diagram of a reflection taking place on a flat
interface between two mediums with constant refraction indexes. From the figure we
can put the trajectory of light as,

R(x ) = n1 d12 + x 2 + n2 d 22 + (L − x )2 , (4.31)

where, d1 is the height of the initial position of the light ray. x is the horizontal
distance from the initial position of the light ray to the origin of the coordinate
system. L is the horizontal distance from the initial position to the final position. d2 is
the height of the final position of the light ray.
To take the path which the light completes with the least time, we deviate the
equation with respect to x.
dR(x ) x (L − x )
= n1 − n2 . (4.32)
dx 2
d1 + x 2 2
d 2 + (L − x ) 2

From figure 4.3, we can see that precisely sin θ1 is given by,
x
sin θ1 = . (4.33)
d12 + x 2

Also, from the figure, we see that sin θ2 is

L−x
sin θ2 = , (4.34)
d 22 + (L − x ) 2

4-7
 Stigmatic Optics

Incident Ray-Light

Optical Surface

Normal Plane

Refracted
Ray-Light

Figure 4.3. Diagram of the refraction of light.

therefore, replacing equations (4.32) and (4.33) in equation (4.32),

n1 sin θ1 = n2 sin θ2. (4.35)

When the light rays are in a homogeneous medium, their paths are straight lines.
When the light changes from one homogeneous medium to another medium, the
refraction (deviation) is expressed by the above equation, which is Snell’s law for
refraction in angle notations.

4.5 The Fermat principle and Snell’s law

Fermat’s principle states that light travels along the path that takes the least time.
Consider the refraction index as a measure of the speed of light in a material. Then
you can use Fermat’s principle to derive the same angular relationship between the
incident and refracted rays as Snell’s law. Fermat’s principle and Snell’s law share
information in common, but they are not the same.

4.6 Malus–Dupin’s theorem

The Malus–Dupin theorem is one of the fundamental theorems of geometric optics
that relate to wave optics. The Malus–Dupin theorem tells us: if on each ray emitted
by a source, we travel the same optical paths, then the points that delimit them form
a surface normal to all rays. We call this surface wavefront. It coincides with the
wavefront given by the oscillatory theory under SVEA. When deduced from

4-8
 Stigmatic Optics

the Fermat principle, it is valid despite the number of reflections or refractions that
the ray may undergo before reaching its destination. The proof is made from the
Fermat principle.
We take two distinct infinitesimally separated paths, [AB ] and [AB′], where A is
the focus and B and B′ are the arrival points separated by equal optical paths. Then.
we define the respective optical paths as,
B
LB = ∫A n( r⃗)ds , (4.36)

and,
B′
LB ′ = ∫A n( r⃗′)ds′ , (4.37)

r ⃗ and r ⃗′ being the respective position vectors, ds and ds′ the respective differentials of
space and n( r )⃗ the refractive index. Notice that we admit that the refractive index is
differentiable.
Now, for the derivation we will use first-order Taylor series development, so we
recall it
f ( r⃗) = f ( r0⃗ ) + ∇f ( r0⃗ ) · ( r⃗ − r0⃗ ). (4.38)
The equation of the path of a light ray (deduced from the Fermat principle see
chapter 2) is
d
(n u⃗) = ∇n , (4.39)
ds
hence,
(n u⃗) = ∇nds . (4.40)

The relationship r′⃗ = r ⃗ + ε b⃗ of which dr⃗ ′ = dr⃗ + ε db⃗ can be seen in figure 4.4.
We admit that the refractive index admits a Taylor series development of order 1.
Then, we obtain that
n(r′⃗) = n( r⃗) + ∇n( r⃗) · (r′⃗ − r⃗) = n( r⃗) + ∇n( r⃗) · ε b⃗ . (4.41)

On the other hand, we will do the same with the module of the position vector:
r′⃗ = r⃗ + ∇ r⃗ · (r′⃗ − r⃗) (4.42)

Figure 4.4. A diagram of the relationship r with r′⃗.

4-9
 Stigmatic Optics

∂
= r⃗ + r e ⃗r · ( r ⃗′ − r ⃗) (4.43)
∂r

= r⃗ + u⃗ · (r′⃗ − r⃗) (4.44)

= r⃗ + u⃗ · r′⃗ − u⃗ · r⃗ (4.45)

= r⃗ + u⃗ · r⃗′ − r⃗ (4.46)

= u⃗ · r′⃗ . (4.47)
Notice that r ⃗ = e⃗r = u⃗ . Thus,
ds′ = dr⃗ ′ = u⃗ · dr⃗ ′ = u⃗ · dr⃗ + u⃗ · ε db⃗ = ds + ε u⃗ · db.
⃗ (4.48)
ds′ is replaced in the optical path, equation (4.37),
B′ B′
LB ′ = ∫A n(r′⃗)ds′ = ∫A [n( r⃗) + ∇n · ε b⃗](ds + ε u⃗ · db⃗ ), (4.49)

expanding,
B′ B′ B′ B′
LB ′ = ∫A n(r )⃗ ds + ∫A ∇n · εdb⃗ + ∫A n(r )⃗ ε u⃗ · db⃗ + ∫A ∇n · ε b⃗ε u⃗ · db⃗ . (4.50)

We remove the order elements from ε 2 , from equation (4.50)

B′ B′ B′
LB ′ = ∫A n( r⃗)ds + ∫A ∇n · ε db⃗ + ∫A n( r⃗)ε u⃗ · db⃗ . (4.51)

We calculate ΔL , equation (4.37) less equation (4.36),

B′ B′ B′ B
ΔL = ∫A n( r⃗)ds + ∫A ∇n · ε b⃗ds + ∫A n( r⃗)ε u⃗ · db⃗ − ∫A n( r⃗)ds . (4.52)

We factor the terms by powers of ε,

B′ B ⎡ B′ B′ ⎤
ΔL = ∫A n( r⃗)ds − ∫A n( r⃗)ds + ε⎢
⎣
∫A ∇n · b⃗ds + ∫A n( r⃗)u⃗ · db⃗ .⎥.
⎦
(4.53)

By the equation of the trajectories we have that d (n u⃗ · b⃗) = d (n u⃗) · b⃗ + n u⃗ · db⃗ =

∇n · b⃗ds + n u⃗ · db⃗ , so,
B′ B ⎡ B′ ⎤
ΔL =
A
∫ n( r⃗)ds −
A
∫
n( r⃗)ds + ε⎢
⎣ A
d (n b⃗ · u⃗)⎥ .
⎦
∫ (4.54)

If we assume that we have chosen B ≡ B′ then,

B′ B′ ⎡ B′ ⎤
ΔL = ∫A n( r⃗)ds − ∫A n( r⃗)ds + ε⎢
⎣
∫A d (n b⃗ · u⃗)⎥ = 0,
⎦
(4.55)

turns to,

4-10
 Stigmatic Optics

⎡ B′ ⎤
ΔL = ε⎢
⎣
∫A d (nb⃗ · u⃗)⎥ = 0.
⎦
(4.56)

Thus,
B′
∫A d (nb⃗ · u⃗) = 0, (4.57)

which implies,
n( rB⃗ ′)b⃗( rB⃗ ′) · u⃗( rB⃗ ′) − n( rA⃗ )b⃗( rA⃗ ) · u⃗( rA⃗ ) = 0, (4.58)

where rB⃗ ′ is the vector r ⃗ at point B ≡ B′ and rA⃗ is the vector r ⃗ at A. Since point A is
the focus, the separation is always null, consequently,
n( rB⃗ ′)b⃗( rB⃗ ′) · u⃗( rB⃗ ′) = 0, (4.59)

therefore, the first term of equation should be zero as well,

b⃗( rB⃗ ′) · u⃗( rB⃗ ′) = 0, (4.60)

the only way that the last expression is zero is because both vectors are perpendicular,
b⃗( rB⃗ ′) ⊥ u⃗( rB⃗ ′). (4.61)

We have r ⃗′ = r ⃗ + ε b⃗ , so ε b⃗(r) joins the points on the surface, from which the formed
surface is orthogonal to each ray. We can justify that the points form a surface for
continuity.

4.7 End notes

In this chapter, we focused on the variational behaviour of light. From it, we
successfully found the ray equation from the Euler equation.
Then, we formalized our notation of the Fermat principle with examples, and we
found Snell’s law of reflection and refraction.
Then, we studied the Lagrangian formalism under the Fermat principles. This
procedure helps us to state and prove the Malus–Dupin theorem.

Further reading
Arfken G B and Weber H J 1999 Mathematical Methods for Physicists (New York: Academic)
Boas M L 2006 Mathematical Methods in the Physical Sciences (New York: Wiley)
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation
Interference and Diffraction of Light (Amsterdam: Elsevier)
Buchdahl H A 1993 An Introduction to Hamiltonian Optics (Chelmsford, MA: Courier
Corporation)
Cao S and Greenhalgh S 1994 Finite-difference solution of the eikonal equation using an efficient,
first-arrival, wavefront tracking scheme Geophysics 59 632–43
Currier R and Herman M F 1985 Numerical comparison of generalized surface hopping, classical
analog, and self-consistent eikonal approximations for nonadiabatic scattering J. Chem.
Phys. 82 4509–16

4-11
 Stigmatic Optics

Dacorogna B, Glowinski R and Pan T-W 2003 Numerical methods for the solution of a system of
eikonal equations with Dirichlet boundary conditions C. R. Math. 336 511–8
Elsgolc L D 2012 Calculus of Variations (Chelmsford, MA: Courier Corporation)
Gelfand I M et al 2000 Calculus of Variations (Chelmsford, MA: Courier Corporation)
Goldstein H, Poole C and Safko J 2002 Classical Mechanics (Reading, MA: Addison-Wesley)
Griffiths D J 2005 Introduction to Electrodynamics (Cambridge: Cambridge University Press)
Hecht E 1974 Schaumas Outline of Optics (New York: McGraw-Hill)
Hecht E 2012 Optics (Cambridge, MA: Pearson)
Hoffnagle J A and Shealy D L 2011 Refracting the k-function: Stavroudis’s solution to the eikonal
equation for multielement optical systems J. Opt. Soc. Am. A 28 1312–21
Huang L, Shu C-W and Zhang M 2008 Numerical boundary conditions for the fast sweeping high
order WENO methods for solving the Eikonal equation J. Comput. Math. 26 336–46
Hysing S-R and Turek S 2005 The eikonal equation: numerical efficiency versus algorithmic
complexity on quadrilateral grids Proc. of ALGORITMY vol 22
Jackson J D 1999 Classical Electrodynamics (Hoboken, NJ: Wiley)
Lakshminarayanan V, Ghatak A and Thyagarajan K 2002 Lagrangian Optics (Berlin: Springer)
Lax M, Louisell W H and McKnight W B 1975 From Maxwell to paraxial wave optics Phys. Rev.
A 11 1365
Luneburg R K 1964 Mathematical Theory of Optics (Berkeley, CA: University of California
Press)
Marion J B 2013 Classical Dynamics of Particles and Systems (New York: Academic)
Morrey C B Jr 2009 Multiple Integrals in the Calculus of Variations (Berlin: Springer)
Pegis R J 1961 I The modern development of Hamiltonian optics Progress in Optics vol 1
(Amsterdam: Elsevier) 1–29
Qian J, Zhang Y-T and Zhao H-K 2007 Fast sweeping methods for eikonal equations on
triangular meshes SIAM J. Numer. Anal. 45 83–107
Ronchi V and Barocas V 1970 The Nature of Light: An Historical Survey (Cambridge, MA:
Harvard University Press) pp 12–288
Spira A and Kimmel R 2004 An efficient solution to the eikonal equation on parametric manifolds
Interfaces Free Bound. 6 315–27
Stavroudis O 2012 The Optics of Rays, Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 1995 The k function in geometrical optics and its relationship to the archetypal
wave front and the caustic surface J. Opt. Soc. Am. A 12 1010–6
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Hurtado-Ramos J B 2000 Maxwell equations and the k function J. Opt. Soc.
Am. A 17 1469–74
Vavryčuk V 2012 On numerically solving the complex eikonal equation using real ray-tracing
methods: A comparison with the exact analytical solution Geophysics 77 T109–16
Zhao H 2005 A fast sweeping method for eikonal equations Math. Comput. 74 603–27

4-12
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 5
Stigmatism and stigmatic reflective surfaces

Having laid the foundations of geometric optics, the equations of the eikonal and
the ray. We can focus on the sub-branches of geometric optics, optical design,
aberration theory, and stigmatism. In this chapter we address the aforementioned
topics and in addition we study the simplest stigmatic systems, conical mirrors.

5.1 Introduction
Imagine that you have a solid connected body with a refractive index different than
the environment where the body is located. This body is a lens if it has the function
of focusing or dispersing rays through the refraction that arises from the difference
between the refraction index of the mentioned body and the medium. Lenses are
fundamental elements of study in geometric optics.
If the body has the function of redirecting the rays such that there is no dispersion
or focus on them, then the element in question is a prism. If the body does not have a
function, then it is just a translucent rock.
We have mentioned that lenses have the function of focusing light or scattering it.
It is also true that mirrors have such features.
The law of reflection and the law of refraction depend on the normal of the
surface. Therefore the shape of the lens/mirror is essential to fulfilling its prede-
termined function.
When it comes to focusing the light, what is wanted, in principle, is that the rays
that come from a point object converge onto a point image. Therefore, what is
wanted is to have stigmatic lenses and stigmatic mirrors. Stigmatism refers to the
image-formation property of an optical system which focuses a point object into a
point image. Such points are called a stigmatic pair of the optical system.

doi:10.1088/978-0-7503-3463-1ch5 5-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Stigmatic lenses and stigmatic mirrors need to have a very particular shape. In the
case of mirrors, the reflective surfaces that form the stigmatic mirrors are the conic
sections1.
• For mirrors with parabolic surfaces, parallel rays that hit the mirror produce
reflected rays that converge onto a common focus2.
• For mirrors with spherical surfaces, all rays that emerge from a point object
located at a finite distance from the reflecting surface are reflected in the same
point. If and only if the object is situated in the centre of the circumference.
• For mirrors with elliptic surfaces, all rays that emerge from a point object are
reflected in another point, the point image.
• For mirrors with hyperbolic surfaces, all rays that arise from a point object
are reflected in a single virtual point image.

In this chapter we are going to study the conical mirrors aforementioned. The
beauty of conical mirrors lies in the intrinsic geometric properties of conic sections.
Other curved surfaces may also focus light, but not in a single point. The stigmatic
refractive surface is the Cartesian oval, which is a fourth-order function. In other
means, the Cartesian oval is a surface such that all the rays that emerge from a point
object are focused on a single image point, once they are refracted, see figure 5.1.
The conic mirrors and the Cartesian oval are results of interest that have been
preserved and will be preserved over time by their analytic nature. The Cartesian
ovals will be studied in chapters 6–8.
In the case of stigmatic lenses, it took more than two thousand years to have a
general equation that describes their surfaces. The general equation that describes
the stigmatic lenses is not trivial. The aforementioned equation will be reviewed in
chapter 9.

-20 -10 10 20

-5

Figure 5.1. Cartesian oval.

1
Conic sections are all those obtained by cutting a cone with a plane. The Greek mathematician Apollonius of
Perge (262–190 BC) was the first to study in detail the conic sections. Apollonius classified the conics in four
types: ellipses, hyperbolas, circles, and parables.
2
Diocles (240 BC–ca. 180 BC), in his work Burning Mirrors, was the first person that reported this property of
the parabolic mirror.

5-2
 Stigmatic Optics

At this point, it is convenient to define what is stigmatic optics. It is the design of an

optical system based exclusively on the premises of geometrical optics without counting
any paraxial approximation and without having any optimization process. The
relevance of the results given by the analytical optical design is that they are preserved
over time, since they are general and not particular cases. Examples of analytical optical
design results are conical mirrors, Cartesian oval and stigmatic lenses.
Before we present the derivation of the shapes of the conic mirror we will
introduce the concept of aberrations.

5.2 Aberrations
When a system is not stigmatic for all points of the object, then the system has
aberrations. In this section, we will show the terminology implemented throughout
this treatise on optical aberrations.
If the optical system has aberration, the point object is projected in a region in the
image space instead of at a single point. The nature of the region of space where
the image is formed depends on the type of aberration. The optical aberrations of the
optical system distort the image formed by the optical system.
Aberrations fall into two classes: chromatic and monochromatic. The variation of
a lens’s refractive index concerning the wavelength causes the chromatic aberrations.
The geometry of the optical system causes the monochromatic aberrations. In
general, these occur both when light is reflected and refracted, so the reflection law,
Snell’s law, and Fermat’s principle are involved in the phenomenon. They have
information about the geometry of the imaging system. Therefore, the shape of the
optical system is crucial to for monochromatic aberrations to vanish. The five basic
types of monochromatic aberrations are spherical aberration, coma, astigmatism,
field curvature, and image distortion.

Spherical aberration
Spherical aberration is the phenomenon that exists in an optical system when a point
object located on the optical axis does not have a stigmatic correspondence with a
point image. In other words, the rays that leave the point object on the optical axis
do not converge on a point image on the optical axis. It is called spherical aberration
because the spherical lens has this phenomenon. There are lenses called aspherical
because their shape is different from the sphere. In most cases, their main goal is to
reduce spherical aberration. In the following chapters, we will see that a stigmatic
lens is aspherical.
In figure 5.2 there is an example of spherical aberration generated by a spherical
surface with constant refraction index n along with the material.
Another example of spherical aberration is presented in figure 5.3. This time the
surface is parabolic.

Coma
Coma aberration in an optical system refers to the aberration suffered by the image
of a point object outside the axis. Coma makes the image appear distorted, with a

5-3
 Stigmatic Optics

15

r
-40 -20 20 40 60 80

-15

Figure 5.2. Spherical aberration in an spherical surface.

15

r
-40 -20 20 40 60 80

-15

Figure 5.3. Spherical aberration in an parabolic surface.

tail, like a coma or a comet. In other words, an optical system with coma has no
stigmatic relationship between a point object outside the optical axis and a point
image, since this image is not a point but a region, figures 5.4 and 5.5.
As another way to explain how the coma looks, in figure 5.6 we show an interface
with rays that come from an off-axis object, then the rays cross the surface of index
refraction n, and they are refracted. The refraction causes an inversion of some of the
rays, as can be seen in the same figure. This inversion in the photos looks like a coma
or a comet.

Astigmatism
The astigmatism of a point object takes place when two perpendicular planes have
different image points. In figure 5.7 we show a lens with astigmatism.
The distortion in a forming image system is measured with a rectilinear
projection. The rectilinear projection is passed through the system, a projection in
which straight lines in a scene remain straight in the image if there is no distortion in
the system. The most common distortions are in figures 5.8 and 5.9.

Field curvature
Petzval field curvature, named for Joseph Petzval, describes the optical aberration in
which a flat object normal to the optical axis is focused as a curved image. In
figure 5.10 is the field curvature in a Cartesian oval.

5-4
 Stigmatic Optics

Figure 5.4. Polynomial refractive surface with coma.

5-5
 Stigmatic Optics

Figure 5.5. Parabolic refractive surface with coma.

5-6
 Stigmatic Optics

Figure 5.6. Spherical refractive surface with coma.

5.3 Conic mirrors

Once we have introduced the concepts of stigmatism and aberrations, we can start
design stigmatic optical systems. The first optical systems that we are going to study
are the conic mirrors. Ancient Greeks reviewed the conic mirrors more than two
millennia ago. To obtain the shapes of the stigmatic mirrors, we are going to use the
Fermat principle already studied in chapter 4.

5-7
 Stigmatic Optics

Figure 5.7. An astigmatism presented in a lens.

5.4 Elliptic mirror

The Fermat principle predicts that light follows the least time path. If the refraction
index of the medium is constant, the least time path from an object placed in zo is a
straight line to the image placed in zi. This line, we can called the optical axis if zo
and zi are aligned; and the ray that follows it is the axial ray. Hence, if all the rays
follow the least time path, then the optical path of a non-axial ray is the same as the
path of the axial ray. The optical path of the axial ray is, as the Fermat principle
predicts, that the light follows the least time path. If the refraction index of the
medium is constant, the least time path from an object placed in zo is a straight line
to the image placed in zi. This line, we can called the optical axis if zo and zi are
aligned; and the ray that follows it is the axial ray. Hence, if all the rays follow the
least time path, then the optical path of a non-axial ray is the same as the path of the
axial ray. The optical path of the axial ray is,
−zo − zi , (5.1)

5-8
 Stigmatic Optics

Figure 5.8. Pincushion distortion.

while the optical path of the non-axial ray is,

(za − zi ) 2 + ra2 + ra2 + (za − zo) 2 , (5.2)

where ra is the radius of the surface za, and it works as the independent variable. The
idea is how must za be such that the mirror is stigmatic. Tacking into account what
we have deduced in the first paragraph of this chapter, we have,

−zo − zi = (za − zi ) 2 + ra2 + ra2 + (za − zo) 2 , (5.3)

where in the last expression we have −zo − zi in the left side because the coordinate
system is placed at the vertex of the mirror.
Solving for za,
⎡ ⎤
1⎢ zozi (zo + zi ) 2(zozi − ra2 ) ⎥
za = zo + zi ± . (5.4)
2 ⎢⎣ zozi ⎥
⎦

From the last expression, from the ± we take the positive sign, in order to have an
elliptic mirror, hence,

5-9
 Stigmatic Optics

Figure 5.9. Barrel distortion.

⎡ ⎤
1⎢ zozi (zo + zi ) 2(zozi − ra2 ) ⎥
za = zo + zi + . (5.5)
2 ⎢⎣ zozi ⎥
⎦

The last expression give us the shape of a mirror which, for a point object and a
point image located at zo and zi, respectively, the system is stigmatic.
An example of an elliptic mirror can be seen in figure 5.11. The configuration
design is captured in the caption of the image.

5.5 Circular mirror

The circle is an exceptional case of the ellipse when the two focuses of the ellipse are
placed together. Therefore, if we want a stigmatic optical system with a circular
mirror, we should take zo = zi . Replacing zo = zi in equation (5.5) we have,

za = zi2 − ra2 + zi (5.6)

where za perfectly matches with the shape of a circle.

5-10
 Stigmatic Optics

Figure 5.10. Petzval field curvature in a Cartesian oval.

5-11
 Stigmatic Optics

10

-30 -20 -10 0

-5

-10

10

r
-40 -30 -20 -10

-5

-10

10

-30 -20 -10 0

-5

-10

Figure 5.11. Specifications of the design: zo = −40 , zi = −25 mm , z = equation (5.5).

5-12
 Stigmatic Optics

An example of a circular mirror is figure 5.12. Please see the caption for the design
parameters.

5.6 Hyperbolic mirror

When the object is virtual, we should use equation (5.4) with the negative sign. This
procedure leads us to the hyperbolic mirror

⎡ ⎤
1⎢ zozi (zo + zi ) 2(zozi − ra2 ) ⎥
za = zo + zi − . (5.7)
2 ⎢⎣ zozi ⎥
⎦

For an example of a hyperbolic mirror, see figure 5.13.

5.7 Parabolic mirror

If the object is very far away from the image zo → ∞ and we want a stigmatic mirror
we should manipulate equation (5.3) in order to get the following expression,

ra2 + (za − zo) 2 + zo

1= , (5.8)
− (za − zi ) 2 + ra2 − zi

then, we can compute the limit when zo → ∞,

⎡ ⎤
⎢ ra2 + (za − zo) 2 + zo ⎥ za
lim =− , (5.9)
zo→−∞ ⎢ − (z − z ) 2 + r − z ⎥
2
(zi − za ) 2 + ra2 + zi
⎣ a i a i⎦

substituting equation (5.9) in equation (5.8) we have,

za
1=− , (5.10)
(zi − za ) 2 + ra2 + zi

and finally, solving for za,

ra2
za = . (5.11)
4zi

It is clear that the last expression is a parabola. An example of the parabolic

mirror can be found in figure 5.14.

5-13
 Stigmatic Optics

10

-15 -10 -5 0

-5

-10

10

r
-20 -15 -10 -5

-5

-10

10

-15 -10 -5 0

-5

-10

Figure 5.12. Specifications of the design: zo = −40 , zi = −25 mm za = zi2 − ra2 + zi .

5-14
 Stigmatic Optics

15

10

-40 -30 -20 -10

-5

-10

-15

z
15

10

r
-40 -30 -20 -10

-5

-10

-15

15

10

-40 -30 -20 -10

-5

-10

-15

Figure 5.13. Specifications of the design: zo = −40 , zi = 16 mm and za = equation (5.7).

5-15
 Stigmatic Optics

20

10

-60 -40 -20 0

-10

-20

20

10

r
-60 -40 -20

-10

-20

20

10

-60 -40 -20 0

-10

-20

ra2
Figure 5.14. Specifications of the design: zo = −∞, zi = −60 mm and za = .
4zi

5-16
 Stigmatic Optics

5.8 End notes

In this chapter, we demonstrated several essential concepts, such as stigmatism and
how the absence of stigmatism in a system forms aberrations. We presented the most
common types of aberrations and their diagrams.
Finally, using the Fermat principle, we found that the static mirrors have
conical curve shapes. For each case, we showed the deduction and an illustrative
example.

Further reading
Bass M 1995 Handbook of Optics, Volume I: Fundamentals Techniques and Design (New York:
McGraw-Hill)
Bellosta H 2002 Burning instruments: From Diocles to Ibn Sahl Arabic Sci. Philos. 12 285–303
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Braunecker B, Hentschel R and Tiziani H J 2008 Advanced Optics using Aspherical Elements vol
173 (Bellingham, WA: SPIE)
Chaves J 2016 Introduction to Nonimaging Optics 2nd edn (Boca Raton, FL: CRC Press)
Daumas M 1972 Scientific Instruments of the Seventeenth and Eighteenth Centuries and their
Makers (London: Batsford) 361 p + 142 plates
Descartes R 2012 The Geometry of Rene Descartes: With a facsimile of the First Edition
(Chelmsford, MA: Courier Corporation)
Duerr F, Benítez P, Minano J C, Meuret Y and Thienpont H 2012 Analytic design method for
optimal imaging: coupling three ray sets using two free-form lens profiles Opt. Express 20
5576–85
Glassner A S 1989 An Introduction to Ray Tracing (Amsterdam: Elsevier)
Gross H 2005 Handbook of Optical Systems, Volume 1, Fundamentals of Technical Optics
(New York: Wiley) p 848
Hogendijk J P 2002 The burning mirrors of Diocles: reflections on the methodology and purpose
of the history of pre-modern science Early Sci. Med. 7 181–97
Kingslake R and Johnson R B 2009 Lens Design Fundamentals (New York: Academic)
Lefaivre J 1951 A new approach in the analytical study of the spherical aberrations of any order
J. Opt. Soc. Am. 41 647
Lin W, Benítez P, Miñano J C, Infante J and Biot G 2011 Advances in the SMS design method
for imaging optics Optical Design and Engineering IV vol 8167 (Bellingham, WA: SPIE)
p 81670M
Luneburg R K 1964 Mathematical Theory of Optics (Berkeley, CA: University of California
Press)
Malacara-Hernández D and Malacara-Hernández Z 2017 Handbook of Optical Design (Boca
Raton, FL: CRC Press)
Miñano J C, Benítez P, Lin W, Muñoz F, Infante J and Santamaría A 2009 Overview of the SMS
design method applied to imaging optics Novel Optical Systems Design and Optimization XII
vol 7429 (Bellingham, WA: SPIE) p 74290C
Schulz G 1983 Achromatic and sharp real imaging of a point by a single aspheric lens Appl. Opt.
22 3242–8
Scott J F 2016 The Scientific Work of René Descartes: 1596-1650 (Abingdon: Routledge)

5-17
 Stigmatic Optics

Singer W, Totzeck M and Gross H 2006 Handbook of Optical Systems, Volume 2: Physical Image
Formation (New York: Wiley)
Stavroudis O 2012 The Optics of Rays, Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Feder D P 1954 Automatic computation of spot diagrams J. Opt. Soc. Am.
44 163–70
Sun H 2016 Lens Design: A Practical Guide (Boca Raton, FL: CRC Press)
Toomer G J 2012 Diocles, On Burning Mirrors: The Arabic Translation of the Lost Greek Original
vol 1 (Berlin: Springer)
Vaskas E M 1957 Note on the Wasserman-Wolf method for designing aspheric surfaces J. Opt.
Soc. Am. 47 669–70
Wassermann G D and Wolf E 1949 On the theory of aplanatic aspheric systems Proc. Phys. Soc.
Sect. B 62 2
Winston R, Miñano J C and Benitez P G 2005 Nonimaging Optics (Amsterdam: Elsevier)

5-18
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 6
Stigmatic refractive surfaces: the Cartesian ovals

The next known stigmatic surface is the Cartesian oval. In this chapter we find
different polynomial series for different circumstances of the Cartesian oval. These
polynomial series depend on whether the object or image is virtual or real. It is worth
mentioning that they are a first approximation of the Cartesian oval.

6.1 Introduction
The classic refractive optical system is based on spherical surfaces, the problem of
these surfaces is that they generate a noisy image, because they have spherical
aberration, the phenomenon studied in chapter 5. Several techniques reduce the
spherical aberrations, modifying the shape of the surface; most of them are numeric
or approximated.
The problem that we are going to solve in this chapter is the following: what is the
shape of the surface that does not introduce spherical aberration, once the ray beam
has passed through the interface? The shape is called Cartesian oval. The Cartesian
oval has several shapes according the scenario. For example if the object and the
image are real/virtual, or if one is real and the other is virtual.
This chapter is based on Valencia et al 2015 (see Further reading) where they
presented the functions that describe the Cartesian oval in different scenarios. In the
work on which this chapter is based, the authors take some well-known solutions of
Cartesian ovals and in more complex situations they present polynomial expansions
of their authorship that describe the Cartesian ovals. In this approach, it is deduced
all special cases of Cartesian ovals are from Snell’s law and not the Fermat principle.

6.2 Stigmatic surfaces

The approach we explore throughout the chapter is the Valencia et al approach or
Valencia–Calle approach. Understanding the strategy implemented here will help us
to understand the mathematical models coming from the following chapters, the
stigmatic lenses.

doi:10.1088/978-0-7503-3463-1ch6 6-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

The first assumption is that the interface is between two homogeneous optical
materials with a positive refraction index in the object space no and the image space
ni. Thus, we calculate the parameter n as,
n
n= i. (6.1)
no
Calling Snell’s law in the interface, we have,
sin θi = n sin θr , (6.2)
where the angles, θi and θr are the incident ray angle and the refracted ray angle,
respectively, see figure 6.1. Now, using the following trigonometric identity,
sin θ 2 + cos θ 2 = 1, (6.3)

sin θ = 1 − cos θ 2 , (6.4)

we can express Snell’s law, equation (6.2) as,

1 − cos θi2 = n 1 − cos θr2 , (6.5)

and squaring both sides of the Snell’s law, from equation (6.5) gives
1 − cos θi2 = n(1 − cos θr2 ). (6.6)

Now the cosine of the angle between two vectors a1⃗ and a2⃗ and the dot product of
two the vectors are related by
a1⃗ · a2⃗ = ∣a1⃗ ∣∣a2⃗ ∣ cos θ , (6.7)
taking account of this, we express the cosine of Snell’s law as,
a⃗ · n⃗
cos θi = , (6.8)
∣a⃗∣∣n⃗∣
and

Figure 6.1. Diagram of a Cartesian oval. The origin is placed in the vertex of the Cartesian oval (z, r ). The
distance from the object to the vertex is zo. The gap between the origin and the image is zi. The normal vector
of the Cartesian oval is n ⃗ , the incident ray is a ⃗ and the refracted ray is b ⃗ .

6-2
 Stigmatic Optics

−n⃗ · b⃗
cos θr = , (6.9)
∣b⃗∣∣n⃗∣
where the vector a⃗ is a unitary vector in the opposite direction to the incident ray, its
norm is expressed as ∣a⃗∣. The vector b⃗ is a unitary vector in the direction of the
refracted ray, its norm is expressed as ∣b⃗∣. Finally, the vector n⃗ is the normal vector of
the refractive surface, the surface under study and its norm is expressed as ∣n⃗∣, see
figure 6.1. So, replacing equations (6.8) and (6.9) in equation (6.6),

⎛ a⃗ · n⃗ ⎞2 ⎡ ⎛ −n⃗ · b⃗ ⎞2 ⎤
2⎢ ⎜ ⎟ ⎥.
1−⎜ ⎟ = n 1 − (6.10)
⎝ ∣a⃗∣∣n⃗∣ ⎠ ⎢⎣ ⎝ ∣b⃗∣∣n⃗∣ ⎠ ⎥⎦

Solving for n2 and simplifying we get equation (6.10)

∣b⃗∣2 (∣a⃗∣2 ∣n⃗∣2 − (a⃗ · n⃗)2 )

= n2, (6.11)
∣a⃗∣2 (∣b⃗∣2 ∣n⃗∣2 − (b⃗ · n⃗)2 )

where the vectors are given by,

a⃗ = [1, mi ], (6.12)

b⃗ = [1, mr ], (6.13)

⎡ dz ⎤
n⃗ = ⎢ , −1⎥ , (6.14)
⎣ dr ⎦

where mi is the slope of the incident ray, mr is the slope if the refracted ray and dz
dr
is
the derivative of the interface with zero spherical aberration respect to r. Therefore,
⎛ dz ⎞2
(a⃗ · n⃗)2 = ⎜ −mi + ⎟ , (6.15)
⎝ dr ⎠

⎛ dz ⎞2
(b⃗ · n⃗)2 = ⎜ −mr + ⎟ , (6.16)
⎝ dr ⎠
and

∣a⃗∣2 = 1 + mi2 , (6.17)

∣b⃗∣2 = 1 + m r2 , (6.18)

⎛ dz ⎞2
∣n⃗∣2 = 1 + ⎜ ⎟ . (6.19)
⎝ dr ⎠

6-3
 Stigmatic Optics

Replacing (a⃗ · n⃗)2 , (b⃗ · n⃗)2 , ∣a⃗∣2 , ∣b⃗∣2 and ∣n⃗∣2 , in equation (6.10) we have,
⎡ ⎡ ⎛ dz ⎞⎤ ⎛ dz ⎞2 ⎤
(1 + m r2 )⎢(1 + mi2 )⎢1 + ⎜ ⎟⎥ − ⎜ −mi + ⎟⎥
⎣ ⎣ ⎝ dr ⎠⎦ ⎝ dr ⎠ ⎦
= n2, (6.20)
⎡ ⎡ ⎛ dz ⎞⎤ ⎛ dz ⎞2⎤
(1 + mi2 )⎢(1 + m r2 )⎢1 + ⎜ ⎟⎥ − ⎜ −mr + ⎟⎥
⎣ ⎣ ⎝ dr ⎠⎦ ⎝ dr ⎠ ⎦

simplifying and reordering,

⎛ dz ⎞2 ⎛ dz ⎞2
(1 + m r2 )⎜1 + mi ⎟ = n 2(1 + mi2 )⎜1 + mr ⎟ . (6.21)
⎝ dr ⎠ ⎝ dr ⎠
dz
Now, from equation (6.21) we solve for dr
and we get,

dz mi (1 + m r2 ) − n 2mr (1 + mi2 ) ± n(mr − mi ) (1 + m r2 )(1 + mi2 )

= . (6.22)
dr n 2m r2(1 + mi2 ) − mi2(1 + m r2 )
From equation (6.21) we can explore interesting cases, like when the object is at
infinity, then the slope of the incident ray is mi → ∞, therefore,
⎛ dz ⎞2 ⎛ dz ⎞2
⟹(m r2 + 1)⎜ ⎟ = n 2⎜mr + 1⎟ , (6.23)
⎝ dr ⎠ ⎝ dr ⎠
dz
solving for dr
,

dz n
= . (6.24)
dr −nmr ± m r2 + 1

The solution of the last differential equation will give us an interface with a
surface such that when the rays, of an object placed at minus infinity, cross it will not
introduce any spherical aberration.
Anther compelling case is when the image is at infinity, which means that the
refracted rays are parallel to the optical axis. Thus, computing the limit mr → ∞ in
equation (6.21),
⎛ dz ⎞2 ⎛ dz ⎞2
⟹⎜mi + 1⎟ = n 2(mi2 + 1)⎜ ⎟ , (6.25)
⎝ dr ⎠ ⎝ dr ⎠
dz
solving for dr
,

dz 1
= . (6.26)
dr −mi ± mi2 + 1

The solution of the last expression gives us an interface with a surface such that
for a finite object, the image is at infinity.

6-4
 Stigmatic Optics

Now, let’s focus of the slopes mi and mr. First assume that the light is coming
from a point located at (ro, zo ). From that point, multiple rays emerge that touch the
surface under study, z. They cross the surface such that they are refracted, all of
them converging on a single point at (ri , zi ). Therefore, there is no spherical
aberration in the image. Taking account of this, we can write the slopes mi and
mr as,
z − zo
mi = ,
r − ro
z − zi (6.27)
mr = .
r − ri
Introducing the slopes of equation (6.27) in equation (6.21), we have,
⎡ ⎛ z − zi ⎞2 ⎤⎛ z − zo dz ⎞
2 ⎡ ⎛ z − zo ⎞2 ⎤
⎢1 + ⎜ ⎟ ⎥⎜1 + ⎟ = n 2⎢1 + ⎜ ⎟⎥
⎢⎣ ⎝ r − ri ⎠ ⎥⎦⎝ r − ro dr ⎠ ⎢⎣ ⎝ r − ro ⎠ ⎥⎦
(6.28)
⎛ z − zi dz ⎞
2
× ⎜1 + ⎟ ,
⎝ r − ri dr ⎠

simplifying and reordering the terms we can get to the following expression,
⎡ dz ⎤2 2 2
⎢⎣r − ro + (z − zo) ⎥⎦ [(r − ri ) + (z − zi ) ]
dr
= n2, (6.29)
⎡ dz ⎤2
2 2
⎢⎣r − ri + (z − zi ) ⎥⎦ [(r − ro) + (z − zo) ]
dr

from the last expression, we need to solve for dz

dr
, so we can get a differential equation
dz
of first order. Thus solving for dr
in equation (6.29),

dz (r − ro )(z − zo )[(r − ri ) 2 + (z − zi ) 2 ] − n 2 (r − ri )(z − zi )[(r − ro ) 2 + (z − zo ) 2 ]

= ,
dr n 2 (z − zi ) 2 [(r − ro ) 2 + (z − zo ) 2 ] − (z − zo ) 2 [(r − ri ) 2 + (z − zi ) 2 ]
(6.30)
n [(r − ri ) 2 + (z − zi ) 2 ][(r − ro ) 2 + (z − zo ) 2][(r − ro )(z − zi ) − (r − ri )(z − zo )]2
± .
n 2 (z − zi ) 2 [(r − ro ) 2 + (z − zo ) 2 ] − (z − zo ) 2 [(r − ri ) 2 + (z − zi ) 2 ]

Equation (6.30) is non-linear and the general solution has not been obtained yet.
In the following section we explore particular cases of equation (6.30).

6.2.1 Case I: ro = ri = 0, zo → −∞ and zi = f

In this section, we study the case when r0 = r1 = 0, which means that the object and
the image are along the optical axis. Also we have zo → −∞, which means that the
object is far away from the surface.
From equation (6.24) we can replace the slopes of equation (6.27) and obtain the
following expression,

6-5
 Stigmatic Optics

dz n
=
dr −nmr ± m r2 + 1
n (6.31)
= .
⎛ z − zi ⎞ ⎛ z − zi ⎞2
− n⎜ ⎟± ⎜ ⎟ +1
⎝ r − ri ⎠ ⎝ r − ri ⎠

Since the object is far away at infinity, we have zo to ∞. So, the rays from the
object are parallel and perfectly collimated to the optical axis and the interface under
study z (r ), therefore, equation (6.31) turns to,
⎡ ⎤
⎢ ⎥
⎢ dz n ⎥
lim ⎢ = ⎥
zo→∞ ⎢ dr ⎛ z − zi ⎞ ⎛ z − zi ⎞2 ⎥
⎢ −n⎜ ⎟± ⎜ ⎟ +1⎥ (6.32)
⎣ ⎝ r − ri ⎠ ⎝ r − ri ⎠ ⎦
dz n(r − ri )
⟹ = .
dr −n(z − zi ) ± (z − zi )2 + (r − ri )2
Now if the rays converge on the optical axis ri = 0, and zi = f , where f is the focus of
the interface, equation (6.32) turns to,
dz n(r − ri )
= ,
dr −n(z − zi ) ± (z − zi )2 + (r − ri )2
n(r )
= , (6.33)
−n(z − f ) ± (z − f )2 + (r )2
r
= .
1 2 2
f−z± (z − f ) + r
n
In section 6.3 we are going to solve the above equation. For the moment we
explore other cases.

6.2.2 Case II: ro = ri = 0, zo = f and zi → −∞

Another important case that we need to study is when the refracted rays are
collimated and paralleled to the optical axis, zi → ∞. So, we go back to equation
(6.26) and replace the slopes of equation (6.27),
dz 1
= , (6.34)
dr −mi ± mi2 + 1
1
= .
⎛ z − zo ⎞ ⎛ z − zo ⎞2 (6.35)
−⎜ ⎟± ⎜ ⎟ +1
⎝ r − ro ⎠ ⎝ r − ro ⎠

6-6
 Stigmatic Optics

Then we evaluate the limit when zi → ∞,

⎡ ⎤
⎢ ⎥
⎢ dz 1 ⎥
lim ⎢ = ⎥
zi →∞ ⎢ dr ⎛ z − zo ⎞ ⎛ z − zo ⎞2 ⎥
⎢ −⎜ ⎟± ⎜ ⎟ +1⎥ (6.36)
⎣ ⎝ r − ro ⎠ ⎝ r − ro ⎠ ⎦
dz r − ro
⟹ = .
dr −(z − zo) ± n (z − zo)2 + (r − ro)2
Again, let’s assume that the rays converge at the optical axis, which means ri = 0.
Also, we can call zi = f , where f is the focus of the interface. Thus, equation (6.36)
becomes,
dz r − ro
= ,
dr −(z − zo) ± n (z − zo)2 + (r − ro)2
r
= , (6.37)
− (z − f ) ± n (z − f ) 2 + (r ) 2
r
= .
f − z ± n (z − f ) 2 + r 2
In the next section we are going to solve the above differential equation and other
particular cases of equation (6.30).

6.3 Analytical stigmatic refractive surfaces

In this section, we are going to focus on the special cases of equation (6.30). One by
one we are going to solve them.

6.3.1 Case A: ro = ri = 0, zo → −∞ and zi = f

In this section we recover, case I of section 6.2.1. Thus, ro = ri = 0, zo → −∞ and
zi = f , when f > 0 the focus is real and when f < 0, the focus is virtual. So we
recover equation (6.33),
dz r
= .
dr 1 (6.33)
f−z± (z − f ) 2 + r 2
n
The boundary condition is that the Cartesian oval is placed at the origin of the
coordinate system,
z(0) = 0. (6.38)
The four possible solutions of equation (6.33) are the following expressions:

n(n + 1)f − sign(f ) (n + 1)n 2[(n + 1)f 2 + ( −n + 1)r 2 ]

z= , (6.39)
n2 − 1

6-7
 Stigmatic Optics

n(n − 1)f − sign(f ) (n − 1)n 2[(n + 1)f 2 + ( −n + 1)r 2 ] (6.40)

z= ,
n2 − 1

n(n + 1)f − sign(f ) (n + 1)n 2[(n + 1)f 2 + ( −n − 1)r 2 ] (6.41)

z= ,
n2 − 1

n(n − 1)f − sign(f ) (n − 1)n 2[(n − 1)f 2 + ( −n − 1)r 2 ] (6.42)

z= .
n2 − 1
In practice, using the ray tracing techniques, we found that the solution is given by

n⎡⎣(n − 1)f − sign(f ) (n − 1)[(n − 1)f 2 − (n + 1)r 2 ] ⎤⎦

z= . (6.43)
n2 − 1
Examples of Cartesian ovals of equation (6.43) can be seen in figures 6.2 and 6.3.

6.3.2 Case B: ro = ri = 0, zo = f and zi → −∞

Now, here we return to case II of section 6.2.2. Thus, ro = ri = 0, zf and zi → −∞, so
we recall equation (6.37),
dz r
= . (6.37)
dr f − z ± n (z − f ) 2 + r 2

Again we need to solve the equation taking the boundary condition when the
Cartesian oval is placed at the origin,
z(0) = 0. (6.44)

Mathematically speaking there are four solutions for equation (6.37), which can
be written as:

(n − 1)f + sign(f ) (n − 1)[(n + 1)f 2 + (n + 1)r 2 ]

z= , (6.45)
n2 − 1

(n + 1)f − sign(f ) (n + 1)[(n + 1)f 2 + (n − 1)r 2 ]

z= , (6.46)
n2 − 1

(n − 1)f − sign(f ) (n − 1)[(n − 1)f 2 + (n + 1)r 2 ]

z= , (6.47)
n2 − 1

(n + 1)f + sign(f ) (n − 1)[(n + 1)f 2 + (n − 1)r 2 ]

z= . (6.48)
n2 − 1

6-8
 Stigmatic Optics

Figure 6.2. Design specifications: n = 2, zo → −∞, zi = 40 mm and z = equation (6.43).

Figure 6.3. Design specifications: n = 2, zo → −∞, zi = − 40 mm and z = equation (6.43).

6-9
 Stigmatic Optics

To find which of the four are valid, we plot the ray tracing, we find that the solution
of equation (6.37) we want is,
(n − 1)f − sign(f ) (n − 1)[(n − 1)f 2 + (n + 1)r 2 ] (6.49)
z= .
n2 − 1
An example of the solution can be seen in figures 6.4 and 6.5.

6.3.3 Case C: ro = ri = 0, zo = ∓f and zi = ±f

For this case, we assume that the object is finite and the image as well. But in this case,
both distances are the same value f but a different sign. So then in this case we have,
ro = ri = 0, zo = −f and zi = +f . Taking these values in equation (6.30) we have,
dz (r − ro)(z − zo)[(r − ri ) 2 + (z − zi ) 2] − n 2(r − ri )(z − zi )[(r − ro) 2 + (z − zo) 2]
=
dr n 2(z − zi ) 2[(r − ro) 2 + (z − zo) 2] − (z − zo) 2[(r − ri ) 2 + (z − zi ) 2]
(6.30)
n [(r − ri ) 2 + (z − zi ) 2][(r − ro) 2 + (z − zo) 2][(r − ro)(z − zi ) − (r − ri )(z − zo)] 2
± .
n 2(z − zi ) 2[(r − ro) 2 + (z − zo) 2] − (z − zo) 2[(r − ri ) 2 + (z − zi ) 2]

Thus, we set ro = ri = 0, zo = −f and zi = f in equation (6.30),

dz (r )(z + f )[(r )2 + (z − f )2 ] − n 2(r )(z − f )[(r )2 + (z + f )2 ]
=
dr n 2(z − f )2 [(r )2 + (z + f )2 ] − (z + f )2 [(r )2 + (z − f )2 ]
(6.50)
n [(r )2 + (z − f )2 ][(r )2 + (z + f )2 ][(r )(z − f ) − (r )(z + f )]2
± ,
n 2(z − f )2 [(r )2 + (z + f )2 ] − (z + f )2 [(r )2 + (z − f )2 ]

Figure 6.4. Design specifications: n = 2, zo = −40 mm , zi → ∞ and z = equation (6.49).

6-10
 Stigmatic Optics

Figure 6.5. Design specifications: n = 2, zo = 40 mm , zi → ∞ and z = equation (6.49).

manipulating,
dz r 3(z + f ) + r(z + f )(z − f )2 − n 2r(z − f )[r 2 + (z + f )2 ]
=
dr n 2(z − f )2 [r 2 + (z + f )2 ] − (z + f )2 [r 2 + (z − f )2 ]
(6.51)
±2fnr [r 2 + (z − f )2 ][r 2 + (z + f )2 ]
+ .
n 2(z − f )2 [r 2 + (z + f )2 ] − (z + f )2 [r 2 + (z − f )2 ]
The solution of equation (6.51) is obtained by series, which corresponds to,
r2 r4 2r 6 5r 8 2r10 14f 12
z = c2 + c4 3 + c6 + c 8 + c11 + c12 ⋯
2f 8f 32f 5 128f 7 215f 9 2048f 11
∞ (6.52)
Ikr 2k
= ∑ c2k (2f )2k−1
.
k=1

The coefficients are given by

⎧ n±1
⎪ c2 = ,
⎪ n∓1
⎪ n∓1
⎪ c4 = n ± 1 ,
⎪ 2
⎪ c = (n ± 1)(n + 6n ± 1) ,
⎪
⎪
6
(n ∓ 1)3
⎨ (6.53)
⎪ c8 = n ± 1 ,
⎪ n∓1
⎪ 4 3 2
⎪ c10 = (n ± 1)(7n ± 124n + 122n ± 124n + 7) ,
⎪ (n ∓ 1)5
⎪ 4 3 2
⎪c12 = (n ± 1)(3n ∓ 44n − 46n ± 44n + 3) .
⎪
⎩ (n ± 1) 5

6-11
 Stigmatic Optics

Using the ray tracing to verify the signs of the coefficients, we find that the coefficient
must be the following,
⎧ n+1
⎪ c2 = ,
⎪ n−1
⎪ n+1
⎪ c4 =
n−1
,
⎪
⎪ c6 (n + 1)(n 2 + 6n + 1)
= ,
⎪ (n − 1)3
⎨ (6.54)
⎪ c8 n+1
= ,
⎪ n−1
⎪ (n + 1)(7n 4 + 124n3 + 122n 2 + 124n + 7)
⎪ c10 = ,
⎪ (n − 1)5
⎪ (n + 1)(3n 4 − 44n3 − 46n 2 − 44n + 3)
⎪c12 = .
⎩ (n − 1)5

An example of a Cartesian oval of this section can be seen in figure 6.6.

6.3.4 Case D: ro = ri = 0, zo = −αf and zi = +f

In this case, we assume that the object has a finite gap between the surface, and the
image also has finite position respect to the surface. But both distances are not the
same. We take the distance from the object to the surface is −αf and the distance
from the surface to the image is f. So, the special case D is when ro = ri = 0, zo = −αf
and zi = +f . Taking this values in equation (6.30) we have,

Figure 6.6. Design specifications: n = 2, f = 40 mm, zo = −f , zi = f and z = equation (6.52).

6-12
 Stigmatic Optics

dz (r − ro)(z − zo)[(r − ri ) 2 + (z − zi ) 2] − n 2(r − ri )(z − zi )[(r − ro) 2 + (z − zo) 2]

=
dr n 2(z − zi ) 2[(r − ro) 2 + (z − zo) 2] − (z − zo) 2[(r − ri ) 2 + (z − zi ) 2]
(6.30)
n [(r − ri ) 2 + (z − zi ) 2][(r − ro) 2 + (z − zo) 2][(r − ro)(z − zi ) − (r − ri )(z − zo)] 2
± ,
n 2(z − zi ) 2[(r − ro) 2 + (z − zo) 2] − (z − zo) 2[(r − ri ) 2 + (z − zi ) 2]

replacing ro = ri = 0, zo = −αf and zi = +f in the last equation,

dz (r )(z + αf )[(r )2 + (z − f )2 ] − n 2(r )(z − f )[(r )2 + (z + αf )2 ]

=
dr n 2(z − f )2 [(r )2 + (z + αf )2 ] − (z + αf )2 [(r )2 + (z − f )2 ]
(6.55)
n [(r )2 + (z − f )2 ][(r )2 + (z + αf )2 ][(r )(z − f ) − (r )(z + αf )]2
± ,
n 2(z − f )2 [(r )2 + (z + αf )2 ] − (z + αf )2 [(r )2 + (z − f )2 ]

reordering terms,

dz r 3(z + αf ) + r(z + αf )(z − f )2 − n 2r(z − f )[r 2 + (z + αf )2 ]

=
dr n 2(z − f )2 [r 2 + (z + αf )2 ] − (z + αf )2 [r 2 + (z − f )2 ]
(6.56)
±fnr(α + 1) (r 2 + (z − f )2 ][r 2 + (z + αf )2 ]
+ .
n 2(z − f )2 [r 2 + (z + αf )2 ] − (z + αf )2 [r 2 + (z − f )2 ]

The solution of equation (6.56) is obtained by series, which corresponds to

r2 r4 2r 6 5r 8 2r10 14f 12
z = c2 + c4 3 + c6 + c 8 + c11 + c12 ⋯
2f 8f 32f 5 128f 7 215f 9 2048f 11
∞ (6.57)
I r 2k
= ∑ c2k (2fk)2k−1 .
k=1

Mathematically, the coefficients are given by

⎧ αn ± 1
⎪ c2 = ,
⎪ α (n ∓ 1)
⎪ α 3n 2 ± (α 3 + 2α 2 − 2α − 1)n − 1
⎪c4 = ,
⎪ α 3(n ∓ 1) 2
⎪ 5 3 5 4 3 2 2 5 4 3 2
⎪ c6 = α n ± (2α + 3α − 3α + α + 3α + 1)n + (α + 3α + α − 3α + 3α + 2)n ± 1 ,
⎪ 5
α (n ∓ 1) 3
⎨ (6.58)
⎪ 1 7 4 7 6 5 4 3 2 3
⎪ c 8 = α7(n ± 1) 4 (α n ± (3α + 4α − 4α + 2α + 2α − 4α − 4α + 1)n ),
⎪
⎪ 1 7 6 4 3 2
⎪+ α7(n ± 1) 4 (3α + 8α − 8α + 8α − 8α − 3)n
⎪
⎪ 1 7 6 5 4 3 2
⎪± α7(n ± 1) 4 [(α + 4α + 4α − 2α − 2α + 4α − 4α − 3)n − 1)
⎩

6-13
 Stigmatic Optics

The optical valid coefficients are,

⎧ αn + 1
⎪ c2 = ,
⎪ α (n − 1)
⎪ α 3n 2 + (α 3 + 2α 2 − 2α − 1)n − 1
⎪c4 = ,
⎪ α 3(n − 1) 2
⎪ 5 3 5 4 3 2 2 5 4 3 2
⎪ c6 = α n + (2α + 3α − 3α + α + 3α + 1)n + (α + 3α + α − 3α + 3α + 2)n + 1 ,
⎪ 5
α (n − 1) 3
⎨ (6.59)
⎪ 1 7n 4 + (3α7 + 4α 6 − 4α 5 + 2α 4 + 2α 3 − 4α 2 − 4α + 1)n 3 )
c =
⎪ 8 α7(n − 1) 4 ( α
⎪
⎪ 1 7 6 4 3 2
⎪+ α7(n − 1) 4 (3α + 8α − 8α + 8α − 8α − 3)n
⎪
⎪ 1 7 6 5 4 3 2
⎪+ α7(n − 1) 4 [(α + 4α + 4α − 2α − 2α + 4α − 4α − 3)n − 1).
⎩

Examples of Cartesian ovals of case D can be seen in figures 6.7 and 6.8.

6.3.5 Case E: ro = ri = 0, zo = αf and zi = −f

This special case is almost similar to the previous special cases. But now we assume
that zo = αf , so the technical data is given by ro = ri = 0, zo = αf and zi = −f .
Recalling, equation (6.30),

dz (r − ro)(z − zo)[(r − ri ) 2 + (z − zi ) 2] − n 2(r − ri )(z − zi )[(r − ro) 2 + (z − zo) 2]

=
dr n 2(z − zi ) 2[(r − ro) 2 + (z − zo) 2] − (z − zo) 2[(r − ri ) 2 + (z − zi ) 2]
(6.30)
n [(r − ri ) 2 + (z − zi ) 2][(r − ro) 2 + (z − zo) 2][(r − ro)(z − zi ) − (r − ri )(z − zo)] 2
± ,
n 2(z − zi ) 2[(r − ro) 2 + (z − zo) 2] − (z − zo) 2[(r − ri ) 2 + (z − zi ) 2]

Figure 6.7. Design specifications: n = 2, α = 1.5, f = 40 mm, zo = −αf , zi = f and z = equation (6.57).

6-14
 Stigmatic Optics

Figure 6.8. Design specifications: n = 2, α = 1.5, f = 40 mm, zo = −αf , zi = −f and z = equation (6.57).

6-15
 Stigmatic Optics

replacing, ro = ri = 0, zo = αf and zi = −f ,

dz (r )(z − αf )[(r )2 + (z + f )2 ] − n 2(r )(z + f )[(r )2 + (z − αf )2 ]

=
dr n 2(z + f )2 [(r )2 + (z − αf )2 ] − (z − αf )2 [(r )2 + (z + f )2 ]
(6.60)
n [(r )2 + (z + f )2 ][(r )2 + (z − αf )2 ][(r )(z + f ) − (r )(z − αf )]2
± ,
n 2(z + f )2 [(r )2 + (z − αf )2 ] − (z − αf )2 [(r )2 + (z + f )2 ]

simplifying,

dz r 3(z − αf ) + r(z − αf )(z + f )2 − n 2r(z + f )(r 2 + (z − αf )2 ]

=
dr n 2(z + f )2 [r 2 + (z − αf )2 ] − (z − αf )2 [r 2 + (z + f )2 ]
(6.61)
±fnr(α + 1) [r 2 + (z + f )2 ][r 2 + (z − αf )2 ]
+ .
n 2(z + f )2 [r 2 + (z − αf )2 ] − (z − αf )2 [r 2 + (z + f )2 ]

As usual, the solution of equation (6.61) is obtained by series,

r2 r4 2r 6 5r 8 2r10 14f 12
z = c2 + c4 3 + c6 + c 8 + c11 + c12 ⋯
2f 8f 32f 5 128f 7 215f 9 2048f 11
∞ (6.62)
I r 2k
= ∑ c2k (2fk)2k−1 .
k=1

The coefficients are given by

⎧ αn ± 1
⎪ c2 = − ,
⎪ α (n ∓ 1)
⎪ α 3n 2 ± (α 3 + 2α 2 − 2α − 1)n − 1
⎪c4 = − ,
⎪ α 3(n ∓ 1) 2
⎪ 5 3 5 4 3 2 2 5 4 3 2
⎪ c6 = − α n ± (2α + 3α − 3α + α + 3α + 1)n + (α + 3α + α − 3α + 3α + 2)n ± 1 ,
⎪ 5
α (n ∓ 1) 3
⎨ (6.63)
⎪ 1 7n 4 ± (3α 7 + 4α 6 − 4α 5 + 2α 4 + 2α 3 − 4α 2 − 4α + 1)n3),
c
⎪ 8 = − (α
⎪ α 7(n ± 1) 4
⎪ 1
⎪− 7 (3α 7 + 8α 6 − 8α 4 + 8α 3 − 8α − 3)n 2 ,
⎪ α (n ± 1) 4
⎪ 1
⎪− 7 [(α 7 + 4α 6 + 4α 5 − 2α 4 − 2α 3 + 4α 2 − 4α − 3)n − 1).
⎩ α (n ± 1) 4

6-16
 Stigmatic Optics

Figure 6.9. Design specifications: n = 2, α = 1.5, f = 40 mm, zo = −αf , zi = −f and z = equation (6.57).

Using the ray tracing we verify the correct coefficients,

⎧ αn + 1
⎪ c2 = − ,
⎪ α (n − 1)
⎪ α 3n 2 + (α 3 + 2α 2 − 2α − 1)n − 1
⎪c4 = − ,
⎪ α 3(n − 1) 2
⎪ 5 3 5 4 3 2 2 5 4 3 2
⎪ c6 = − α n + (2α + 3α − 3α + α + 3α + 1)n + (α + 3α + α − 3α + 3α + 2)n + 1 ,
⎪ α 5(n − 1)3
⎨ (6.64)
⎪ 1 7n 4 + (3α 7 + 4α 6 − 4α 5 + 2α 4 + 2α 3 − 4α 2 − 4α + 1)n3)
⎪ c 8 = − (α
⎪ α 7(n − 1) 4
⎪ 1
⎪− 7 4
(3α 7 + 8α 6 − 8α 4 + 8α 3 − 8α − 3)n 2
⎪ α (n − 1)
⎪ 1
⎪− 7 [(α 7 + 4α 6 + 4α 5 − 2α 4 − 2α 3 + 4α 2 − 4α − 3)n − 1).
⎩ α (n − 1) 4

For an example of case E, please see figure 6.9.

6.4 Conclusions
In this chapter, we studied interfaces that do not generate spherical aberration,
which means all the refracted rays converge in a single point, to construct these

6-17
 Stigmatic Optics

surfaces first we explored the phenomenon described by Snell’s law at the surface
under study. The first step is to express Snell’s law in the form of the vectors involved
in the phenomenon instead of the angles. Once Snell’s ruling is in its vector form, we
solve for the derivative of the surface under study, dz /dr . Then we find a non-linear
differential equation, which in some cases can be solved. We study each instance and
present the solution as well as the plot of the solution.
Studying these interfaces is a tremendous help for understanding the solution of
singlet lenses with zero spherical aberration. In the next chapters, we will focus on
similar methods but know we will have two surfaces, not one. But it was essential to
set the basis of this strategy to understand the more complex problems in the
following chapters.

Further reading
Avendaño-Alejo M, Román-Hernández E, Castañeda L and Moreno-Oliva V I 2017 Analytic
conic constants to reduce the spherical aberration of a single lens used in collimated light
Appl. Opt. 56 6244–54
Bass M 1995 Handbook of Optics, Volume I: Fundamentals Techniques and Design (New York:
McGraw-Hill)
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Braunecker B, Hentschel R and Tiziani H J 2008 Advanced Optics Using Aspherical Elements vol
173 (Bellingham, WA: SPIE)
Castillo-Santiago G, Avendaño-Alejo M, Díaz-Uribe R and Castañeda L 2014 Analytic aspheric
coefficients to reduce the spherical aberration of lens elements used in collimated light Appl.
Opt. 53 4939–46
Chaves J 2016 Introduction to Nonimaging Optics 2nd edn (Boca Raton, FL: CRC Press)
Estrada J C V, Calle Á H B and Hernández D M 2013 Explicit representations of all refractive
optical interfaces without spherical aberration J. Opt. Soc. Am. A 30 1814–24
Glassner A S 1989 An Introduction to Ray Tracing (Amsterdam: Elsevier)
González-Acuña R G and Chaparro-Romo H A 2018 General formula for bi-aspheric singlet lens
design free of spherical aberration Appl. Opt. 57 9341–5
González-Acuña R G and Guitiérrez-Vega J C 2018 Generalization of the axicon shape: the
gaxicon J. Opt. Soc. Am. A 35 1915–8
González-Acuña R G and Gutiérrez-Vega J C 2019 Analytic formulation of a refractive-reflective
telescope free of spherical aberration Opt. Eng. 58 085105
González Acuña R G and Gutiérrez-Vega J C 2019 General formula of the refractive telescope
design free spherical aberration Novel Optical Systems, Methods, and Applications XXII vol
11105 ed C F Hahlweg and J R Mulley (Bellingham, WA: SPIE) pp 162–6
González Acuña R G and Gutiérrez-Vega J C 2019 General formula to design freeform
collimator lens free of spherical aberration and astigmatism Novel Optical Systems,
Methods, and Applications XXII vol 11105 (Bellingham, WA: SPIE) p 111050A
González-Acuña R G, Avendaño-Alejo M and Gutiérrez-Vega J C 2019a Singlet lens for
generating aberration-free patterns on deformed surfaces J. Opt. Soc. Am. A 36 925–9
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2019b General formula to
design freeform singlet free of spherical aberration and astigmatism Appl. Opt. 58 1010–5

6-18
 Stigmatic Optics

González-Acuña R G, Chaparro-Romo H A and Gutíerrez-Vega J C 2019c Single lens telescope

(arXiv: 1903.11129)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020a Analytic aplanatic
singlet lens: setting and design for three-point objects and images in the meridional plane Opt.
Eng. 59 055104
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020b Analytical Lens
Design (Bristol: IOP Publishing)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020c Analytic solution of
the eikonal for a stigmatic singlet lens Phys. Scr. 95 085201
González-Acuña R G and Gutiérrez-Vega J C 2019a General formula to eliminate spherical
aberration produced by an arbitrary number of lenses Opt. Eng. 58 085106
González-Acuña R G and Gutiérrez-Vega J C 2019b Analytic formulation of a refractive-
reflective telescope free of spherical aberration Opt. Eng. 58 085105
González-Acuña R G and Gutiérrez-Vega J C 2019c General formula for aspheric collimator lens
design free of spherical aberration Current Developments in Lens Design and Optical
Engineering XX vol 11104 ed R B Johnson, V N Mahajan and S Thibault (Bellingham,
WA: SPIE) pp 181–4
Kingslake R and Johnson R B 2009 Lens Design Fundamentals (New York: Academic)
Lefaivre J 1951 A new approach in the analytical study of the spherical aberrations of any order
J. Opt. Soc. Am. 41 647
Luneburg R K and Herzberger M 1964 Mathematical Theory of Optics (Berkeley, CA: University
of California Press)
Malacara D 1965 Two lenses to collimate red laser light Appl. Opt. 4 1652–4
Malacara-Hernández D and Malacara-Hernández Z 2016 Handbook of Optical Design (Boca
Raton, FL: CRC Press)
González-Acu na R G and Gutiérrez-Vega J C 2020 Analytic design of a spherochromatic singlet
J. Opt. Soc. Am. A 37 149–53
Schulz G 1983 Achromatic and sharp real imaging of a point by a single aspheric lens Appl. Opt.
22 3242–8
Silva-Lora A and Torres R 2020 Explicit Cartesian oval as a superconic surface for stigmatic
imaging optical systems with real or virtual source or image Proc. R. Soc. A 476
Silva-Lora A and Torres R 2020 Superconical aplanatic ovoid singlet lenses J. Opt. Soc. Am. A37
1155–65
Stavroudis O 2012 The Optics of Rays, Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Feder D P 1954 Automatic computation of spot diagrams J. Opt. Soc. Am.
44 163–70
Sun H 2016 Lens Design: A Practical Guide (Boca Raton, FL: CRC Press)
Valencia-Estrada J C and Malacara-Doblado D 2014 Parastigmatic corneal surfaces Appl. Opt.
53 3438–47
Valencia-Estrada J C, Flores-Hernández R B and Malacara-Hernández D 2015 Singlet lenses free
of all orders of spherical aberration Proc. R. Soc. A 471 20140608
Vaskas E M 1957 Note on the Wasserman-Wolf method for designing aspheric surfaces J. Opt.
Soc. Am. 47 669–70

6-19
 Stigmatic Optics

Wassermann G D and Wolf E 1949 On the theory of aplanatic aspheric systems Proc. Phys. Soc.
Sect. B 62 2
Winston R, Miñano J C and Benitez P G 2005 On the theory of aplanatic aspheric systems
Nonimaging Optics (Amsterdam: Elsevier)
Wolf E 1948 On the designing of aspheric surfaces Proc. Phys. Soc. 61 494
Wolf E and Preddy W S 1947 On the determination of aspheric profiles Proc. Phys. Soc. 59 704
Yang T, Jin G-F and Zhu J 2017 Automated design of freeform imaging systems Light Sci. Appl.
6 e17081

6-20
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 7
The general equation of the Cartesian oval

In this chapter, we will obtain the general equation of the Cartesian oval. Unlike in
the previous chapter, the presented equation is not a polynomial approximation, but
a closed expression. The equation is presented is general because it supports that the
object is real or virtual and that the image is real or virtual.

7.1 From Ibn Sahl to Rene Descartes

Cartesian ovals have been studied for over a thousand years. Although they are
named after French mathematician Rene Descartes, they were actually first
considered by Ibn Shal in 984, where today is Basra, Iraq.
Ibn Sahl comprehended very well the optics of antique Greece, but he went much
further to the unexplored field of refraction. His treatise, On the Burning Instruments,
is so unusual for its time that it makes Ibn Sahl the first mathematician compre-
hended to have studied stigmatic lens design. In his time, the 10th century, the main
study was based on catoptrics. Ibn Sahl studied burning mirrors, both parabolic and
ellipsoidal; he considered hyperbolic plano-convex lenses and hyperbolic biconvex
lenses. All these are displayed in On the Burning Instruments. But his most
notable accomplishment is the refraction law (Snell’s law) long before Snell himself.
Sahl was the first to discover the stigmatic refractive surface, today known as the
Cartesian oval. Sahl suggested a stigmatic lens comprised of two Cartesian ovals
where the rays refracted in are collimated along the optical axis. Sahl could not find
a general equation of the Cartesian ovals.

7.2 A generalized problem

In the previous chapter, we presented the Valencia–Calle method of Cartesian ovals.
In general, the technique is simple when the object and image are at minus, plus
infinite, respectively. Otherwise, if the object and the image are finite, we obtain from
Snell’s law differential equations that are quite complicated to solve.

doi:10.1088/978-0-7503-3463-1ch7 7-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Of all the representations that we have of Cartesian ovals when the object and the
image are finite, we always have long equations, whose deduction procedures are not
trivial.
Recently, Alberto Silva-Lora and Rafael Torres published a parametric equation
of the Cartesian ovals in a closed-form (see Further reading). Note that in the
previous chapter, many of the results proposed by the Valencia–Calle method end
with a non-closed polynomial approach.
The Alberto Silva-Lora and Rafael Torres method of obtaining the Cartesian
ovals will be studied in this chapter under the name of Silva–Torres ovals or the
Silva–Torres method.
What makes the Silva–Torres method different from the other procedures is
Silva–Torres has a closed solution that contains all the cases of Cartesian ovals.
Also, the Silva–Torres method has the instances of the stigmatic mirrors that we
studied in chapter 5.
7.3 Mathematical model
In this section, we will obtain the mathematical derivation of the Cartesian ovals by
the Silva–Torres method. Silva–Torres’s approach differs from the Valencia–Calle
model in that it focuses on the optical path, the Fermat principle rather than Snell’s
law. Therefore, it avoids differential equations.
The first thing to note is that the Fermat principle predicts that the Cartesian oval
is the refractive surface such that the optical path of the axial ray is the same path for
all other rays. This is because the axial ray has the minimum optical path length
between the object point located at zo and the image point located at zi, as shown in
figure 7.1.
Therefore, we can match the optical path of any non-axial ray with that of the
axial ray, which gives us the following expression,

ni (z − zi ) 2 + x 2 + y 2 + no (z − zo) 2 + x 2 + y 2 = ni zi − nozo. (7.1)

Figure 7.1. Diagram of a Cartesian oval. The origin is placed in the vertex of the Cartesian oval (z, r ). The
distance from the object to the vertex is zo. The gap between the origin and the image is zi. The distance from
the origin to a given point of the Cartesian oval ρ.

7-2
 Stigmatic Optics

Note that no is the refractive index of the medium where the object is located and
ni is the corresponding refractive index where the image is.
In our three-dimensional model, x, y and z are points in the three-dimensional
space of the Cartesian oval where the non-axial ray passes. See figure 7.1; we take ρ
as a distance from the origin of the coordinate system to x, y and z. The origin is
placed at the vertex of the Cartesian oval.
The next algebraic steps are quite long, so it is good to assign the following
parameters,

A≡ (z − zo) 2 + x 2 + y 2 , (7.2)

the other square root is abbreviated as,

B≡ (z − zi ) 2 + x 2 + y 2 , (7.3)

and finally the optical path of the axial ray is a constant given by,
k ≡ ni zi − nozo. (7.4)
Taking into account the new notation, equation (7.1) has the following form,
noA + ni B = k. (7.5)
Before looking for the Cartesian oval equation, our closest goal is to get rid of the
square roots, so we square equation (7.5) and manipulate it,
A2 n o2 + B 2n i2 − k 2 = 2ABni no. (7.6)

In the previous equation, the term that has the square roots is on the right side, to
eliminate these roots we square again on two sides of the equation mentioned above,
2 2
(A2 n o + B 2n i − k 2 ) 2 = (2ABni no) 2. (7.7)

So, we expand equation (7.7), the interesting thing is that equation (7.7) no longer
has square roots, the result is the following,
n i4x 4 + n o4x 4 − 2n i2n o2x 4 + 2y 2n i4x 2 + 2z 2n i4x 2 + 2y 2n o4x 2 + 2z 2n o4x 2
− 4y 2n i2n o2x 2 − 4z 2n i2n o2x 2 − 4n i2n o2zi2x 2 − 4n i2n o2zo2x 2 − 4zn i4zi x 2
+ 4zn i2n o2zi x 2 − 4zn o4zox 2 + 4zn i2n o2zox 2 + 4ni n o3zi zox 2 + 4n i3nozi zox 2
+ y 4n i4 + z 4n i4 + 2y 2z 2n i4 + y 4n o4 + z 4n o4 + 2y 2z 2n o4 − 2y 4n i2n o2 − 2z 4n i2n o2
− 4y 2z 2n i2n o2 + 4z 2n i4zi2 − 4y 2n i2n o2zi2 − 4z 2n i2n o2zi2 + 4z 2n o4zo2 − 4y 2n i2n o2zo2 (7.8)
− 4z 2n i2n o2zo2 − 8zni n o3zi zo2 + 8zn i2n o2zi zo2 − 4z 3n i4zi − 4y 2zn i4zi + 4z 3n i2n o2zi
+ 4y 2zn i2n o2zi − 4z 3n o4zo − 4y 2zn o4zo + 4z 3n i2n o2zo + 4y 2zn i2n o2zo + 8zn i2n o2zi2zo
− 8zn i3nozi2zo + 4y 2ni n o3zi zo + 4z 2ni n o3zi zo − 8z 2n i2n o2zi zo + 4y 2n i3nozi zo
+ 4z 2n i3nozi zo = 0.

7-3
 Stigmatic Optics

In mathematics and theoretical physics, it can become an art in how we simplify

or reorder equations for our convenience. The following equation is a rearrangement
of equation (7.8); the reason for ordering it in this way is in the following steps,
[(n i2 − n o2 )(x 2 + y 2 + z 2 ) − 2z(n i2zi − n o2zo)] 2
(7.9)
− 4ni no(ni zi − nozo)[(x 2 + y 2 + z 2 )(zi no − ni zo) + 2zzi zo(ni − no)] = 0.

Since ρ is the vector that starts from the origin and ends at points x, y and z, we have
that ρ2 = x 2 + y 2 + z 2 . Substituting ρ2 = x 2 + y 2 + z 2 in equation (7.9) we obtain
equation (7.10),
[ρ 2 (n i2 − n o2 ) − 2z(n i2zi − n o2zo)] 2
(7.10)
− 4ni no(ni zi − nozo)[ρ 2 (zi no − ni zo) + 2zzi zo(ni − no)] = 0,

and expanding the first term of the previous equation we have,

ρ 4 (n i2 − n o2 ) 2 + 4z 2(n i2zi − n o2zo) 2 − 4ρ 2 z(n i2 − n o2 )(n i2zi − n o2zo)
(7.11)
− 4ni no(ni zi − nozo)[ρ 2 (zi no − ni zo) + 2zzi zo(ni − no)] = 0.

From the previous expression we want to replace the following algebraic identities,
(ni − no)2 (ni + no)2 = −2n i2n o2 + n i4 + n o4, (7.12)

and,
(ni − no)(ni + no) = n i2 − n o2 . (7.13)

Therefore, substituting equations (7.12) and (7.13) in equation (7.10) we have

equation (7.14),
2
4z 2(n i2zi − n o2zo) − 2z(2ρ 2 (ni − no)(ni + no)(n i2zi − n o2zo)
+ 4ni zi nozo(ni − no)(ni zi − nozo)) (7.14)
+ ρ 2 [4ρ 2 ni no(ni − no) 2(ni + no) 2(ni zi − nozo)] = 0.

We can simplify equation (7.14) by dividing it by the parameter D expressed by

equation (7.15).
D ≡ 4ni zi nozo(ni − no)(ni zi − nozo). (7.15)
Dividing equation (7.14) by equation (7.15), we have,
c0Kz 2 − (1 + b1ρ 2 )z + (c0 + c1ρ 2 )ρ 2 = 0, (7.16)
where,
2 2 2
(n i zi − n o zo) (7.17)
K≡ ,
ni no(ni zi − nozo)(ni zo − nozi )

7-4
 Stigmatic Optics

ni zo − nozi
c0 ≡ , (7.18)
zi zo(ni − no)

(ni − no)(ni + no)2

c1 ≡ , (7.19)
4ni nozi zo(ni zi − nozo)
and,
(ni + no)(n i2zi − n o2zi )
b1 ≡ . (7.20)
2ni nozi zo(ni zi − nozo)
Equation (7.16) is a quadratic equation whose solution is the following
expression,
1
z=
c0K
(1 + b1ρ 2 ± 1 + (2b1 − c02K )ρ 2 + (b12 − c0c1K )ρ 4 . ) (7.21)

Equation (7.21) can be simplified taking into account that c0c1K is equal to b12 as
demonstrated by the procedure expressed in equation (7.22).

⎡ n z − nozi ⎤⎡ (ni − no)(ni + no) 2 ⎤ ⎡ 2 2

(ni zi − n o zo)
2 ⎤
c0c1K = ⎢ i o ⎥⎢ ⎥⎢ ⎥
⎣ zizo(ni − no) ⎦⎣ 4ni nozizo(ni zi − nozo) ⎦ ⎢⎣ ni no(ni zi − nozo)(ni zo − nozi ) ⎥⎦
(7.22)
2
(n + n ) 2 n 2z − n o2zi )
= i 2 2 o2 (2 i i = b12 .
4ni n o zi zo (ni zi − nozo) 2

Therefore, substituting b12 = c0c1K in equation (7.21) we have,

1
z=
c0K
(
1 + b1ρ 2 ± 1 + (2b1 − c02K )ρ 2 . ) (7.23)

In equation (7.23), multiplying the square root, we have a ± this comes from the
fact that the solution of a second-order equation has two roots. Both solutions are
mathematically valid. But the only interesting solution for this is the one where the
vertex of the Cartesian oval passes through the origin, as shown in figure 7.1.
1
z=
c0K
(
1 + b1ρ 2 − 1 + (2b1 − c02K )ρ 2 . ) (7.24)

In the original publication of this method, Alberto Silva-Lora and Rafael

Torres presented the Cartesian oval with the root in the denominator. So we
multiply both sides of equation (7.24) by (1 + b ρ
1
2
+ )
1 + (2b1 − c02K )ρ2 /

(1 + b ρ
1
2
+ 1 + (2b1 − c02K )ρ2 , )

7-5
 Stigmatic Optics

1
z=
c0K
(
1 + b1ρ 2 − 1 + (2b1 − c02K )ρ 2 )
×
(1 + b ρ 1
2
+ ).
1 + (2b1 − c02K )ρ 2 (7.25)

(1 + b ρ 1
2
+ 1 + (2b − c K )ρ )
1
2
0
2

Simplifying equation (7.25), we obtain the Silva–Torres equation,

ρ 2 (c0 + c1ρ 2 )
z= . (7.26)
ρ 2 (2b1 − c02K ) + 1 + b1ρ 2 + 1

Equation (7.26) is the most important equation in this chapter. It contains

information on all cases of Cartesian ovals. It means that you can have real or
virtual object/images and finite or infinite object/images. With equation (7.26) you
can even get the conical mirrors.
In the following section, we show the potential of this equation. But first, it is
important to note that equation (7.26) is the sagittal part of the Cartesian oval. To
obtain the radial part of the Cartesian oval, from figure 7.1 we know that the
radius is,

r = sgn(ρ) ρ 2 − z 2 . (7.27)

Equation (7.27) is the radial part of the Cartesian oval described by the Silva–
Torres model.

7.4 Illustrative examples

In the following examples, we show several Cartesian ovals generated by the Silva–
Torres method by directly using equations (7.26) and (7.27). In this section, we
discuss all relevant cases, when the object/image is real or virtual. The specifications
of each design are shown in the captions for each figure. The figures are 7.2–7.6.

7.5 Collimated input rays

Now, we study the case when the object is very far from oval Cartesian, zo → ∞. In
this case, the Cartesian oval receives collimated rays along the optical axis. To
obtain a Cartesian oval in this way, we need to evaluate the limit when zo → ∞ in all
the parameters within equations (7.26) and (7.27).
We start with the parameter K, we compute the limit when zo → ∞ in equation
(7.17),
⎡ 2 2 2 ⎤
(n i zi − n o zo)
lim (K ) = lim ⎢ ⎥. (7.28)
zo→−∞ zo→−∞ ⎢
⎣ ni no(ni zi − nozo)(ni zo − nozi ) ⎥⎦

7-6
 Stigmatic Optics

Figure 7.2. Specifications of the design: no = 1, ni = 1.5, zo = −40 mm , zi = 40 mm , z = equation (7.26) and
r = equation (7.27).

Evaluating the limit expressed in the previous equation we have,

n o2
lim (K ) = − . (7.29)
zo→−∞ n2

7-7
 Stigmatic Optics

Figure 7.3. Specifications of the design: no = 1, ni = 1.5, zo = −30 mm , zi = 40 mm , z = equation (7.26) and
r = equation (7.27).

7-8
 Stigmatic Optics

Figure 7.4. Specifications of the design: no = 1, ni = 1.5, zo = −40 mm , zi = −15 mm , z = equation (7.26) and
r = equation (7.27).

As a next act, we apply the limit when zo → −∞ in c0, equation (7.18),

⎡ n z − nozi ⎤
lim (c0) = lim ⎢ i o ⎥, (7.30)
zo→−∞ zo→−∞ ⎣ zi zo(ni − no ) ⎦

computing the limit in equation (7.30), we get,

n
lim (c0) = . (7.31)
zo→−∞ n(zi − τ ) + no(τ − zi )

7-9
 Stigmatic Optics

Figure 7.5. Specifications of the design: no = 1, ni = 1.5, zo = 40 mm , zi = 25 mm , z = equation (7.26) and

r = equation (7.27).

Then we compute limit zo → −∞ in c1, equation (7.19),

⎡ (n − no)(ni + no)2 ⎤
lim (c1) = lim ⎢ i ⎥ = 0. (7.32)
zo→−∞ zo→−∞ ⎣ 4ni nozi zo(ni zi − nozo ) ⎦

The same happens when we compute limit zo → −∞ in b1, equation (7.20)

⎡ (ni + no)(n 2zi − n 2zi ) ⎤
lim (b1) = lim ⎢ i o
⎥ = 0. (7.33)
zo→−∞ zo→−∞ ⎣ 2ni nozi zo(ni zi − nozo ) ⎦

Therefore, the sagitta of the Cartesian oval proposed by the Silva–Torres method is
given by
⎛ ⎞
⎜ lim c0⎟ρ 2
⎝zo→−∞ ⎠
lim (z ) = , (7.34)
zo→−∞ ⎛ ⎞2 ⎛ ⎞
1 − ⎜ lim c0⎟ ⎜ lim K ⎟ρ 2 + 1
⎝zo→−∞ ⎠ ⎝zo→−∞ ⎠

and the radius is given by

lim (r ) = sgn(ρ) ρ 2 − lim (z 2 ) . (7.35)

zo→−∞ zo→−∞

7-10
 Stigmatic Optics

Figure 7.6. Specifications of the design: no = 1, ni = 1.5, zo = 25 mm , zi = −75 mm z = equation (7.26) and
r = equation (7.27).

7.6 Illustrative examples

In this section, we will show an example series of Cartesian ovals with the object in
minus infinity. The image can be real or virtual. The specifications are shown in each
figure. The figures of this gallery are 7.7 and 7.8.

7-11
 Stigmatic Optics

Figure 7.7. Specifications of the design: no = 1, ni = 1.5, zo = −∞, zi = 40 mm , z = equation (7.34) and
r = equation (7.35).

7-12
 Stigmatic Optics

Figure 7.8. Specifications of the design: no = 1, ni = 1.5, zo = −∞, zi = −40 mm , z = equation (7.34) and
r = equation (7.35).

7.7 Collimated output rays

Here, we will focus on studying the Cartesian oval that has collimated rays at the
system output. For this we have to apply the limit zi → ∞ in equations (7.17)–(7.20).
This model is very similar to the model in the previous section.
Let’s start by evaluating limit zi → ∞ of K, equation (7.17),
⎡ 2 2 2 ⎤ n2
(n i zi − n o zo)
lim (K ) = lim ⎢ ⎥=− o, (7.36)
zi →∞ zi →∞ ⎢
⎣ ni no(ni zi − nozo)(ni zo − nozi ) ⎥⎦ n2

then, we have the limit zi → ∞ of c0, equation (7.18),

⎡ n z − nozi ⎤
lim (c0) = lim ⎢ i o ⎥, (7.37)
zi →∞ zi →∞ ⎣ zi zo(ni − no ) ⎦

7-13
 Stigmatic Optics

computing the limit,

no
lim (c0) = − . (7.38)
zi →∞ nzo − nozo
Then, we compute the limit zi → ∞ of c1, equation (7.19),
⎡ (n − no)(ni + no)2 ⎤
lim (c1) = lim ⎢ i ⎥ = 0, (7.39)
zi →∞ zi →∞ ⎣ 4ni nozi zo(ni zi − nozo ) ⎦

and limit zi → ∞ of b1, equation (7.20), is given by,

⎡ (n + no)(n 2zi − n 2zi ) ⎤
lim (b1) = lim ⎢ i i o
⎥ = 0. (7.40)
zi →∞ zi →∞ ⎣ 2ni nozi zo(ni zi − nozo ) ⎦

Once we compute the aforementioned limits, we can express the Cartesian oval
proposed by the Silva–Torres method for output collimated rays

⎛ ⎞
⎜ lim c0⎟ρ 2
⎝zi →∞ ⎠
lim za = , (7.41)
zi →∞ ⎛ ⎞2 ⎛ ⎞
1 − ⎜ lim c0⎟ ⎜ lim K ⎟ρ 2 + 1
⎝zi →∞ ⎠ ⎝zi →∞ ⎠

and radius

lim r = sgn(ρ) ρ 2 − lim (z 2 ) . (7.42)

zi →∞ zi →∞

7.8 Illustrative examples

Now let see some illustrative examples of Cartesian ovals with images in infinity. As
usual, the specifications are in the caption in each figure.
The figures of the gallery of the output collimated rays are 7.9–7.11.
7.9 Reflective surface
We mentioned in the introduction to this chapter that one of the most interesting
attributes of the Silva–Torres method is that it is robust enough that it has all the
information inside conical mirrors. In this section, we demonstrate the mentioned
attribute. To achieve this objective, we take two limits. First is the limit when
ni → −no and second is the limit when no → −1.
We apply both limits to equation (7.17),

⎡ ⎤ ⎡ ⎛ 2 2 2 ⎞⎤
(n i zi − n o zo)
lim ⎢ lim (K )⎥ = lim ⎢ lim ⎜⎜ ⎟⎥ , (7.43)
n o →−1⎣n i →−n o ⎦ no→−1⎢⎣ni →−no ⎝ ni no(ni zi − nozo)(ni zo − nozi ) ⎟⎠⎥⎦

7-14
 Stigmatic Optics

Figure 7.9. Specifications of the design: no = 1, ni = 1.5, zo = −40 mm , zi = ∞, z = equation (7.41) and
r = equation (7.42).

and after computing both limits we get,

⎡ ⎤ (z − zo)2
lim ⎢ lim (K )⎥ = − i . (7.44)
n o →−1⎣n i →−n o ⎦ (zi + zo)2

7-15
 Stigmatic Optics

Figure 7.10. Specifications of the design: no = 1, ni = 1.5, zo = 25 mm , zi = ∞, z = equation (7.41) and

r = equation (7.42).

Then, we compute the mentioned limits in c0, equation (7.18),

⎡ ⎤ ⎡ ⎛ n z − nozi ⎞⎤
lim ⎢ lim (c0)⎥ = lim ⎢ lim ⎜ i o ⎟⎥ . (7.45)
n o →−1⎣n i →−n o ⎦ no→−1⎣ni →−no ⎝ zi zo(ni − no) ⎠⎦

The result is the following,

⎡ ⎤ z + zo
lim ⎢ lim (c0)⎥ = i . (7.46)
n o →−1⎣n i →−n o ⎦ 2zi zo

Then, let’s focus on c1, equation (7.19),

⎡ ⎤ ⎡ ⎛ (n − no)(ni + no)2 ⎞⎤
lim ⎢ lim (c1)⎥ = lim ⎢ lim ⎜ i ⎟⎥ , (7.47)
n o →−1⎣n i →−n o ⎦ no→−1⎣ni →−no ⎝ 4ni nozi zo(ni zi − nozo) ⎠⎦

7-16
 Stigmatic Optics

Figure 7.11. Specifications of the design: no = 1, ni = 1.5, zo = −∞, zin = ∞, zi = ∞, z = equation (7.41) and
r = equation (7.42).

and the computation gives the following result,

⎡ ⎤
lim ⎢ lim (c1)⎥ = 0. (7.48)
n o →−1⎣n i →−n o ⎦
Finally, we need to apply both limits on b1, equation (7.20),
⎡ ⎤ ⎡ ⎛ (n + no)(n i2zi − n o2zi ) ⎞⎤
lim ⎢ lim (b1)⎥ = lim ⎢ lim ⎜ i ⎟⎥ . (7.49)
n o →−1⎣n i →−n o ⎦ no→−1⎢⎣ni →−no ⎝ 2ni nozi zo(ni zi − nozo) ⎠⎥⎦

After computing the limits, we can seen that b1 turns to zero, as can be seen in
equation (7.49).
⎡ ⎤
lim ⎢ lim (b1)⎥ = 0. (7.50)
n o →−1⎣n i →−n o ⎦
Therefore, the Cartesian oval proposed by the Silva–Torres method becomes a conic
mirror whose sagitta is given by equation (7.51),

⎡
lim ⎢ lim (z )⎥ =
⎤ ( )ρ zi + zo
2zizo
2

,
n o →−1⎣n i →−n o ⎦ (7.51)
1−( ) ⎡⎢⎣−
zi + zo 2
2zizo
(zi − zo)2 ⎤ 2
(zi + zo)2 ⎥
⎦
ρ +1

and its radius is given by

⎡ ⎤ ⎡ ⎤2
lim ⎢ lim (r )⎥ = sgn(ρ) ρ 2 − lim ⎢ lim (z )⎥ . (7.52)
n o →−1⎣n i →−n o ⎦ n o →−1⎣n i →−n o ⎦

7-17
 Stigmatic Optics

Equations (7.51) and (7.52) can be used to plot spherical, hyperbolic and elliptic
mirrors. In the next section we study the special case of the parabolic mirror.

7.9.1 Parabolic mirror

In order to get the shape of the parabolic mirror from the Silva–Torres method of
Cartesian ovals, it is necessary to compute the limit when zo → −∞ on equations
(7.44), (7.46), (7.51) and (7.52).
We start with equation (7.44),

⎧ ⎡ ⎤⎫ ⎛ (z − zo)2 ⎞
lim ⎨ lim ⎢ lim (K )⎥⎬ = lim ⎜ − i ⎟, (7.53)
zo→−∞⎩n o →−1⎣n i →−n o ⎦⎭ zo→−∞⎝ (zi + zo)2 ⎠

the computation of the limit when zo → −∞ gives,

⎧ ⎡ ⎤⎫
lim ⎨ lim ⎢ lim (K )⎥⎬ = − 1. (7.54)
zo→−∞⎩n o →−1⎣n i →−n o ⎦⎭

Then, let’s pay attention to equation (7.46),

⎧ ⎡ ⎤⎫ ⎛ z + zo ⎞
lim ⎨ lim ⎢ lim (c0)⎥⎬ = lim ⎜ i ⎟, (7.55)
zo→−∞⎩n o →−1⎣n i →−n o ⎦⎭ zo→−∞ ⎝ 2zi zo ⎠

applying the limits,

⎧ ⎡ ⎤⎫ 1
lim ⎨ lim ⎢ lim (c0)⎥⎬ = . (7.56)
zo→−∞⎩n o →−1⎣n i →−n o ⎦⎭ 2zi

Therefore, the parabolic mirror from the Silva–Torres perspective is given by

⎧ ⎡ ⎤⎫
lim ⎨ lim ⎢ lim (z )⎥⎬ =
( )ρ 1
2zi
2

, (7.57)
z →−∞⎩n o →−1⎣n i →−n o ⎦⎭ 1 2 2
1+( )ρ 2zi
+1

where

⎧ ⎡ ⎤⎫
lim ⎨ lim ⎢ lim (r )⎥⎬ = sgn(ρ)
r →−∞⎩n o →−1⎣n i →−n o ⎦⎭
(7.58)
⎧ ⎡ ⎤⎫2
× ρ 2 − lim ⎨ lim ⎢ lim (z )⎥⎬ .
r →−∞⎩n o →−1⎣n i →−n o ⎦⎭

For this to be successfully implemented we need to use equations (7.57) and (7.58).

7-18
 Stigmatic Optics

7.10 Illustrative examples

In this section, we present examples of all the conic mirrors using the Silva–Torres
approach. The specification of the design is captured in the caption of the images of
each example. The examples are in figures 7.12–7.15.

10

-30 -20 -10 0

-5

-10

10

-40 -30 -20 -10

-5

-10

10

-30 -20 -10 0

-5

-10

Figure 7.12. Specifications of the design: no = −ni = 1, zo = −15, zi = −40 mm , z = equation (7.51) and
r = equation (7.52).

7-19
 Stigmatic Optics

10

-30 -20 -10 0

-5

-10

10

-40 -30 -20 -10

-5

-10

10

-30 -20 -10 0

-5

-10

Figure 7.13. Specifications of the design: no = −ni = 1, zo = −40 , zi = −40 mm , z = equation (7.51) and
r = equation (7.52).

7-20
 Stigmatic Optics

15

10

-40 -30 -20 -10

-5

-10

-15

15

10

-40 -30 -20 -10

-5

-10

-15

15

10

-40 -30 -20 -10

-5

-10

-15

Figure 7.14. Specifications of the design: no = −ni = 1, zo = 15, zi = −40 mm , z = equation (7.51) and
r = equation (7.52).

7-21
 Stigmatic Optics

20

10

-60 -40 -20 0

-10

-20

20

10

r
-60 -40 -20

-10

-20

20

10

-60 -40 -20 0

-10

-20

Figure 7.15. Specifications of the design: no = −ni = 1, zo = −∞, zi = −40 mm , z = equation (7.57) and
r = equation (7.58).

7-22
 Stigmatic Optics

7.11 End notes

In this chapter, we deduce a powerful expression to get Cartesian ovals. The first
ones to get this approach were Alberto Silva-Lora and Rafael Torres. Hence, we call
this approach the Silva–Torres method. The Silva–Torres method is a closed-form
solution of a Cartesian oval. The Silva–Torres method has all in one single
expression, equation (7.26). Equation (7.26) has all the cases of the Cartesian ovals
and the conic mirrors. During the chapter, we presented examples of all the cases of
the Cartesian ovals and the conic mirrors.

Further reading
Avendaño-Alejo M, Román-Hernández E, Castañeda L and Moreno-Oliva V I 2017 Analytic
conic constants to reduce the spherical aberration of a single lens used in collimated light
Appl. Opt. 56 6244–54
Bass M 1995 Handbook of Optics, Volume I: Fundamentals Techniques and Design (New York:
McGraw-Hill)
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Braunecker B, Hentschel R and Tiziani H J 2008 Advanced Optics Using Aspherical Elements vol
173 (Bellingham, WA: SPIE)
Castillo-Santiago G, Avendaño-Alejo M, Díaz-Uribe R and Castañeda L 2014 Analytic aspheric
coefficients to reduce the spherical aberration of lens elements used in collimated light Appl.
Opt. 53 4939–46
Chaves J 2016 Introduction to Nonimaging Optics 2nd edn (Boca Raton, FL: CRC Press)
Estrada J C V, Calle Á H B and Hernández D M 2013 Explicit representations of all refractive
optical interfaces without spherical aberration J. Opt. Soc. Am. A 30 1814–24
Glassner A S 1989 An Introduction to Ray Tracing (Amsterdam: Elsevier)
González-Acuña R G and Chaparro-Romo H A 2018 General formula for bi-aspheric singlet lens
design free of spherical aberration Appl. Opt. 57 9341–5
González-Acuña R G and Guitiérrez-Vega J C 2018 Generalization of the axicon shape: the
gaxicon J. Opt. Soc. Am. A 35 1915–8
González-Acuña R G and Gutiérrez-Vega J C 2019 Analytic formulation of a refractive-reflective
telescope free of spherical aberration Opt. Eng. 58 085105
González Acuña R G and Gutiérrez-Vega J C 2019 General formula of the refractive telescope
design free spherical aberration Novel Optical Systems, Methods, and Applications XXII vol
11105 ed C F Hahlweg and J R Mulley (Bellingham, WA: SPIE) pp 162–6
González Acuña R G and Gutiérrez-Vega J C 2019 General formula to design freeform
collimator lens free of spherical aberration and astigmatism Novel Optical Systems,
Methods, and Applications XXII vol 11105 (Bellingham, WA: SPIE) p 111050A
González-Acuña R G, Avendaño-Alejo M and Gutiérrez-Vega J C 2019a Singlet lens for
generating aberration-free patterns on deformed surfaces J. Opt. Soc. Am. A 36 925–9
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2019b General formula to
design freeform singlet free of spherical aberration and astigmatism Appl. Opt. 58 1010–5

7-23
 Stigmatic Optics

González-Acuña R G, Chaparro-Romo H A and Gutíerrez-Vega J C 2019c Single lens telescope

(arXiv:1903.11129)
González-Acuña R G and Gutiérrez-Vega J C 2020 Analytic design of a spherochromatic singlet
J. Opt. Soc. Am. A 37 149–53
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020a Analytic aplanatic
singlet lens: setting and design for three-point objects and images in the meridional plane Opt.
Eng. 59 055104
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020b Analytical Lens
Design (Bristol: IOP Publishing)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020c Analytic solution of
the eikonal for a stigmatic singlet lens Phys. Scr. 95 085201
González-Acuña R G and Gutiérrez-Vega J C 2019a General formula to eliminate spherical
aberration produced by an arbitrary number of lenses Opt. Eng. 58 1–6
González-Acuña R G and Gutiérrez-Vega J C 2019b Analytic formulation of a refractive-
reflective telescope free of spherical aberration Opt. Eng. 58 1–5
González-Acuña R G and Gutiérrez-Vega J C 2019c General formula for aspheric collimator lens
design free of spherical aberration Current Developments in Lens Design and Optical
Engineering XX vol 11104 ed R B Johnson, V N Mahajan and S Thibault (Bellingham,
WA: SPIE) pp 181–4
Kingslake R and Johnson R B 2009 Lens Design Fundamentals (New York: Academic)
Lefaivre J 1951 A new approach in the analytical study of the spherical aberrations of any order
J. Opt. Soc. Am. 41 647
Luneburg R K and Herzberger M 1964 Mathematical Theory of Optics (Berkeley, CA: University
of California Press)
Malacara D 1965 Two lenses to collimate red laser light Appl. Opt. 4 1652–4
Malacara-Hernández D and Malacara-Hernández Z 2016 Handbook of Optical Design (Boca
Raton, FL: CRC Press)
Schulz G 1983 Achromatic and sharp real imaging of a point by a single aspheric lens Appl. Opt.
22 3242–8
Silva-Lora A and Torres R 2020 Explicit Cartesian oval as a superconic surface for stigmatic
imaging optical systems with real or virtual source or image Proc. R. Soc. A. 476
Silva-Lora A and Torres R 2020 Superconical aplanatic ovoid singlet lenses J. Opt. Soc. Am. A 37
1155–65
Stavroudis O 2012 The Optics of Rays, Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Feder D P 1954 Automatic computation of spot diagrams J. Opt. Soc. Am.
44 163–70
Sun H 2016 Lens Design: A Practical Guide (Boca Raton, FL: CRC Press)
Valencia-Estrada J C and Malacara-Doblado D 2014 Parastigmatic corneal surfaces Appl. Opt.
53 3438–47
Valencia-Estrada J C, Flores-Hernández R B and Malacara-Hernández D 2015 Singlet lenses free
of all orders of spherical aberration Proc. R. Soc. A 471 20140608
Vaskas E M 1957 Note on the Wasserman-Wolf method for designing aspheric surfaces J. Opt.
Soc. Am. 47 669–70

7-24
 Stigmatic Optics

Wassermann G D and Wolf E 1949 On the theory of aplanatic aspheric systems Proc. Phys. Soc.
Sect. B 62 2
Winston R, Miñano J C and Benitez P G 2005 Nonimaging Optics (Amsterdam: Elsevier)
Wolf E 1948 On the designing of aspheric surfaces Proc. Phys. Soc. 61 494
Wolf E and Preddy W S 1947 On the determination of aspheric profiles Proc. Phys. Soc. 59 704
Yang T, Jin G-F and Zhu J 2017 Automated design of freeform imaging systems Light: Sci. Appl.
6 e17081

7-25
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 8
The stigmatic lens generated by Cartesian ovals

From two stigmatic surfaces, Cartesian ovals, a stigmatic lens can be generated. In
this chapter, we will study the generation of stigmatic lenses from the Cartesian ovals
model proposed by Alberto Silva-Lora and Rafael Torres.

8.1 Introduction
In the previous chapter, we presented the rigorous model of Silva–Torres Cartesian
ovals, which is general enough to cover all cases of Cartesian ovals and conical
mirrors.
In this chapter, we now address the design of a stigmatic lens using two Silva–
Torres Cartesian ovals. The general idea is that the image formed by the first
Cartesian oval of Silva–Torres is the object of the second Cartesian oval of Silva–
Torres. The amazing thing about this strategy is that, as we mentioned, the model of
Silva–Torres Cartesian ovals covers all cases of Cartesian ovals and conical mirrors.
So, the image of the first Cartesian oval can be taken as a real or virtual object for
the second Cartesian oval.

8.2 Mathematical model

The central idea behind this mathematical model is based on two Silva–Torres
Cartesian ovals. The image produced by the first Silva–Torres Cartesian oval is
taken as the object of the second Silva–Torres Cartesian oval. See the diagram of the
model in figure 8.1. no is the refraction index of the medium that surrounds the
object. zo is the distance from the object to the vertex of the first Silva–Torres
Cartesian oval which is denoted by za. Each suffix with a is concerning the first
Silva–Torres Cartesian oval. We call the first Silva–Torres Cartesian oval just the
first surface. Notice that the coordinate system is placed in the vertex of the first
surface. n is the refraction index of the lens generated by the two Silva–Torres
Cartesian ovals. zin is the distance from the first surface to the image generated by
the first surface. The second surface takes this image as an object. That is the reason

doi:10.1088/978-0-7503-3463-1ch8 8-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Figure 8.1. Sketch of a lens designed with two Cartesian ovals as refractive surfaces. The first Cartesian oval is
(za, ra ) and the second Cartesian oval is (zb, rb ). The refraction index in the object space is no, inside the lens,
the refraction index is n, and in the image space is ni. The origin is placed in the vertex of the first Cartesian
oval. ρ is the distance from the origin to a point in the first surface. The length from the object to the first
Cartesian oval is zo; the distance from the first Cartesian oval to the image generated by the first surface is zin.
The centre depth of the lens is τ. From the origin to the image the gap is given by ze, and for the gap from the
second Cartesian oval to the image, the length is zi.

why zin has in as a suffix, in is from the inside. τ is the central thickness of the lens.
ze is the distance from the origin to the image. The second Silva–Torres Cartesian
oval is denoted by zb, and it is also called the second surface in this chapter. The
suffix b is the member mark that says that they are related to the second surface.
Taking all the considerations of the last paragraph, the mathematical model
provided in the last chapter and figure 8.1, it is easy to see that the first surface is
given by equation (8.1),

ρ 2 (c1aρ 2 + c 0a )
za ≡ (8.1)
ρ 2 (2b1a − c02aK a ) + 1 + b1a ρ 2 + 1

where its radius, equation (8.2), is,

ra ≡ sgn(ρ) ρ 2 − za(ρ)2 (8.2)

and the inside parameters are,

2 2
(n 2zin − n o zo) (8.3)
Ka ≡
nno(nzin − nozo)(nzo − nozin)
then, c0a is expressed as,
nzo − nozin
c 0a ≡ , (8.4)
zinzo(n − no)
with c1a , we have,

(n − no)(n + no)2
c1a ≡ (8.5)
4nnozinzo(nzin − nozo)

8-2
 Stigmatic Optics

and, finally b1a ,

(n + no)(n 2zin − n o2zo)

b1a ≡ . (8.6)
2nnozinzo(nzin − nozo)

Applying the same strategy as for obtaining the first surface, we can write the second
surface as,

ρ 2 (c1bρ 2 + c 0b )
zb0 ≡ (8.7)
ρ 2 (2b1b − c02bK b ) + 1 + b1bρ 2 + 1

where,
zb ≡ τ + zb0(ρ) (8.8)

where the radius is,

rb ≡ sgn(ρ) ρ 2 − zb0(ρ)2 . (8.9)

If we take no = ni , the parameters related to the second surface are

2 2
(n o (ze − τ ) − n 2(zin − τ )) (8.10)
Kb ≡ ,
nno[n(ze − τ ) − no(zin − τ )][n(zin − τ ) − no(ze − τ )]

c0b is given by,

n(ze − τ ) − no(zin − τ )
c 0b ≡ (8.11)
(n − no)(ze − τ )(zin − τ )

c1b is written as,

(n − no)(n + no)2
c1b ≡ (8.12)
4nno(ze − τ )(zin − τ )[n(zin − τ ) − no(ze − τ )]

and, finally b1b ,

(n + no)(n 2(zin − τ ) − n o2(ze − τ ))

b1b ≡ (8.13)
2nno(ze − τ )(zin − τ )[n(zin − τ ) − no(ze − τ )]

Notice that the second surface has the same structure as the first surface but it is out-
placed by the thickness central of the lens, τ.
With equations (8.1)–(8.13) we can design stigmatic lenses for countless config-
urations, where zo, zin, ze can be placed anywhere in the optical axis. Therefore, we
can design stigmatic lenses for real and virtual object/images.

8-3
 Stigmatic Optics

8.3 Examples
In this section, we present examples of stigmatic lenses designed with equations
(8.1)–(8.13). Where, the objects and images are finite but they can be real or virtual.
The specification of each design is presented in the caption of the respective figure.
The figures of this section are 8.2–8.7.

8.4 Collector
A collector lens is a lens that receives the rays from minus infinity; this means that
the image is placed at minus infinity. Therefore, if we are interested in designing a
stigmatic lens with two Cartesian ovals, we need to compute the limit when zo → −∞
over the parameters of the first surface. Notice that we only need to apply the limit
mentioned above in the first surface, because the second surface and its parameters
do not depend on zo.
We start, by computing the limit when zo → −∞ in equation (8.3),
⎡ 2 2 ⎤ n2
(n 2zin − n o zo)
lim K a = lim ⎢ ⎥=− o, (8.14)
zo→−∞ zo→−∞ ⎢
⎣ nno(nzin − nozo)(nzo − nozin) ⎥⎦ n2

then, we compute the same limit in equation (8.4),

Figure 8.2. Specifications of the design: n = 1.5, zo = −40, τ = 12, zi = 40, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.8), rb = equation (8.9), using equations (8.3)–(8.13).

8-4
 Stigmatic Optics

Figure 8.3. Specifications of the design:n = 1.5, zo = −15, τ = 10, zi = −5, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.8), rb = equation (8.9), using equations (8.3)–(8.13).

⎡ nz − nozin ⎤ n
lim c 0a = lim ⎢ o ⎥= , (8.15)
zo→−∞ zo→−∞ ⎣ zinzo(n − no ) ⎦ zin(n − no)

followed by computing the limit when zo → −∞ in equation (8.5),

⎡ (n − no)(n + no)2 ⎤
lim c1a = lim ⎢ ⎥=0 (8.16)
zo→−∞ zo→−∞ ⎣ 4nnozinzo(nzin − nozo ) ⎦

and finally, we apply the limit when zo → −∞ in equation (8.6),

⎡ (n + no)(n 2zin − n o2zo) ⎤
lim b1a = lim ⎢ ⎥ = 0. (8.17)
zo→−∞ zo→−∞ ⎣ 2nnozinzo(nzin − nozo ) ⎦

8-5
 Stigmatic Optics

Figure 8.4. Specifications of the design:n = 1.5, zo = 40, τ = 20, zi = 25, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.8), rb = equation (8.9), using equations (8.3)–(8.13).

Therefore, we can write the first surface as

⎛ ⎞
⎜ lim c 0a⎟ρ 2
⎝zo→−∞ ⎠
lim za = (8.18)
zo→−∞ ⎛ ⎞2 ⎛ ⎞
1 − ⎜ lim c 0a⎟ ⎜ lim K a⎟ρ 2 + 1
⎝zo→−∞ ⎠ ⎝zo→−∞ ⎠

and its radius as,

lim ra = sgn(ρ) ρ 2 − lim (za2 ) . (8.19)

zo→−∞ zo→−∞

Equations (8.14)–(8.19) give us the correct shape of a Cartesian oval when zo → −∞.
In the next section we will compute some example with these equations.

8.5 Examples
In the following section, we start with some examples of lenses with real objects and
real/virtual images. The parameters to design the lenses are in the captions of the
figures. The presented designs are evaluated using equations (8.7)–(8.19) without any
optimization or modification.

8-6
 Stigmatic Optics

Figure 8.5. Specifications of the design:n = 1.5, zo = 50, τ = 30, zi = −60, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.8), rb = equation (8.9), using equations (8.3)–(8.13).

In figures 8.8–8.11 are lenses of the gallery. All the rays that come from minus
infinity are a focus on the image point at ze.

8.6 Collimator
In this section, we are going to work with the inverse problem of the last part. Now
the input rays come from a point source located a finite distance with respect to the
first surface, and the output rays are collimated. Therefore, we need to compute the
limit when ze → ∞ for the parameter of the second surface.
We shall being with equation (8.10); the limit gives us,
⎡ [n o2(ze − τ ) − n 2(zin − τ )]
2 ⎤ n2
lim K b = lim ⎢ ⎥ = − o , (8.20)
ze →∞ ze →∞ ⎢
⎣ nn o[n(ze − τ ) − n o(zin − τ )][n(zin − τ ) − n o(ze − τ )] ⎥⎦ n2

8-7
 Stigmatic Optics

Figure 8.6. Specifications of the design:n = 1.5, zo = 50, τ = 30, zi = −60, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.8), rb = equation (8.9), using equations (8.3)–(8.13).

then, it is appropriated to compute the limit when ze → ∞ over equation (8.11),

⎡ n(ze − τ ) − no(zin − τ ) ⎤ n
lim c 0b = lim ⎢ ⎥= . (8.21)
ze →∞ ze →∞ ⎣ (n − no )(ze − τ )(zin − τ ) ⎦ −nτ + nzin + noτ − nozin

Now let’s apply the limit when ze → ∞ in equation (8.12),

⎡ (n − no)(n + no)2 ⎤
lim c1b = lim ⎢ ⎥=0 (8.22)
ze →∞ ze →∞ ⎣ 4nno(ze − τ )(zin − τ )[n(zin − τ ) − no(ze − τ )] ⎦

and finally we compute the mentioned limit in equation (8.13),

⎡ (n + no)[n 2(zin − τ ) − n o2(ze − τ )] ⎤
lim b1b = lim ⎢ ⎥ = 0. (8.23)
ze →∞ ze →∞ ⎣ 2nno(ze − τ )(zin − τ )[n(zin − τ ) − no(ze − τ )] ⎦

8-8
 Stigmatic Optics

Figure 8.7. Specifications of the design: zo = 10 mm, zi = −60 mm, τ = 15 mm, n = 1.5, zin = nzi, za = equation
(8.1), ra = equation (8.2), zb = equation (8.8), rb = equation 8.9, using equations (8.3), (8.4), (8.5), (8.6), (8.7),
(8.10), (8.11), (8.12 and (8.13).

As a result of the above computations we write the second surface as,

⎛ n ⎞ 2
⎜ ⎟ρ
⎝ −nτ + nzin + noτ − nozin ⎠
lim zb0 = (8.24)
ze →∞ ⎛ n ⎞2 ⎛ n o2 ⎞ 2
1−⎜ ⎟ ⎜ − ⎟ρ + 1
⎝ −nτ + nzin + noτ − nozin ⎠ ⎝ n 2 ⎠

where,

lim zb = τ + lim zb0 (8.25)

ze →∞ ze →∞

8-9
 Stigmatic Optics

Figure 8.8. Specifications of the design: n = 1.5, τ = 10, zi = 40, zin = n zi using za = equation (8.18),
ra = equation (8.19), zb = equation (8.8), rb = equation (8.9) with equations (8.7)–(8.19).

and the radius becomes,

lim rb = sgn(ρ) ρ 2 − lim (zb0)2 . (8.26)

ze →∞ ze →∞

Equations (8.1)–(8.6), (8.25), (8.24), (8.26), (8.20), (8.21), (8.22), and (8.23) are the
ones we need if we want to design a collimator stigmatic singlet lens using Cartesian
ovals. Notice we do not compute any limit over the first surface and its parameters
because ze is not presented in them.

8.7 Examples
The following figures have the ray tracing and specification of several examples of
collimator lenses. All the presented models are estimated using equations (8.1)–(8.6),
(8.20)–(8.26). The figures are 8.12 and 8.13.

8.8 Single-lens telescope with Cartesian ovals

The single-lens telescope is more a myth than a practical lens used in engineering.
The legend of the single-lens telescope comes from the paraxial optics. In paraxial

8-10
 Stigmatic Optics

Figure 8.9. Specifications of the design: n = 1.5, τ = 10, zi = 50, zin = n zi using za = equation (8.18),
ra = equation (8.19), zb = equation (8.8), rb = equation (8.9) with equations (8.7)–(8.19).

optics, all lenses can be represented by matrices. But there is not a matrix for the
single-lens telescope; there is a matrix for flat glass. However, the rays that come
collimated along the optical axis refracted by a flat glass at the output are
collimated, but not amplified. The idea of the single-lens telescope is a lens such
that the collimated rays entering the lens are refracted such that they suffer an
amplification, and in the output they are collimated. Also the myth comes from the
first telescopes, which were made from at least two lenses.
In this section, we are going partially demystify the single-lens telescope by
obtaining the Cartesian ovals of the single-lens telescope. This procedure is to show
the robustness of the method implemented during the chapter.
To design it we only use the equations (8.14)–(8.26).

8-11
 Stigmatic Optics

Figure 8.10. Specifications of the design: n = 1.5, τ = 28, zi = −20, zin = n zi using za = equation (8.18),
ra = equation (8.19), zb = equation (8.8), rb = equation (8.9) with equations (8.7)–(8.19).

8.9 Example
A possible application of the single-lens telescope is not as a telescope, but as a beam
expander (typically an array of two lenses that expand a laser beam).
The single-beam expander lens and the single-lens telescope are the same because
ray optics are invertible. The only difference is that in the singlet beam expander lens
at the output, the amplification is positive and in the single-lens telescope it is
negative. For us, a positive amplification means that the beam expands and a
negative amplification means that the beam contracts.

8-12
 Stigmatic Optics

Figure 8.11. Specifications of the design:n = 1.5, τ = 10, zi = 40, zin = n zi using −za = equation (8.18),
ra = equation (8.19), zb = equation (8.8), rb = equation (8.9) with equations (8.7)–(8.19).

Figure 8.12. Specifications of the design: n = 1.5, zo = −55, τ = 30, zi = ∞, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.25), rb = equation (8.26), using equations (8.3)–(8.6), (8.20)–(8.24). This is
a special case because the inner wavefront is flat, and it happens that the second surface is flat too.

8-13
 Stigmatic Optics

Figure 8.13. Specifications of the design: n = 1.5, zo = 16, τ = 1, zi = ∞, zin = n zi , za = equation (8.1),
ra = equation (8.2), zb = equation (8.25), rb = equation (8.26), using equations (8.3)–(8.6), (8.20)–(8.24).

30

20

10

-40 -20 20 40 60 80

-10

-20

-30

Figure 8.14. Specifications of the design: n = 1.5, τ = 25, zin = −60, za = equation (8.18), ra = equation (8.19),
zb = equation (8.25), rb = equation (8.26), using equations (8.14)–(8.17), (8.20)–(8.24).

In figures 8.14–8.16 there is a single-lens telescope. At the input and output, the
rays are collimated. The example is computed using equations (8.14)–(8.26). In the
caption of the mentioned figure is the design specifications.

8-14
 Stigmatic Optics

20

10

-40 -20 20 40 60 80

-10

-20

Figure 8.15. Specifications of the design: n = 1.5, τ = 25, zin = 60, za = equation (8.18), ra = equation (8.19),
zb = equation (8.25), rb = equation (8.26), using equations (8.14)–(8.17), (8.20)–(8.24).

20

10

-40 -20 20 40 60 80

-10

-20

Figure 8.16. Specifications of the design:n = 1.5, τ = 25, zin = ∞, za = equation (8.18), ra = equation (8.19),
zb = equation (8.25), rb = equation (8.26), using equations (8.14)–(8.17), (8.20)–(8.24).

8.10 End notes

In this chapter, we have demonstrated that stigmatic singlet lenses exist for real/virtual
objects and real/virtual images using Cartesian ovals. Also, we tested several for
several configurations. When the object is finite/infinite, and the image is finite/infinite.
In the next chapter, we are going to generalise the concept of stigmatic lenses outside
the Cartesian ovals we are going to obtain the general equation of stigmatic lenses.

Further reading
González Acuña R G and Gutiérrez-Vega J C 2019 General formula of the refractive telescope
design free spherical aberration Novel Optical Systems, Methods, and Applications XXII
vol 11105 ed C F Hahlweg and J R Mulley (Bellingham, WA: SPIE) pp 162–6
González Acuña R G and Gutiérrez-Vega J C 2019 General formula to design freeform
collimator lens free of spherical aberration and astigmatism Novel Optical Systems,
Methods, and Applications XXII vol 11105 (Bellingham, WA: SPIE) p 111050A

8-15
 Stigmatic Optics

Avendaño-Alejo M, Román-Hernández E, Castañeda L and Moreno-Oliva V I 2017 Analytic

conic constants to reduce the spherical aberration of a single lens used in collimated light
Appl. Opt. 56 6244–54
Bass M 1995 Handbook of Optics, Volume I: Fundamentals Techniques and Design (New York:
McGraw-Hill)
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Braunecker B, Hentschel R and Tiziani H J 2008 Advanced Optics Using Aspherical elements
vol 173 (Bellingham, WA: SPIE)
Castillo-Santiago G, Avendaño-Alejo M, Díaz-Uribe R and Castañeda L 2014 Analytic aspheric
coefficients to reduce the spherical aberration of lens elements used in collimated light Appl.
Opt. 53 4939–46
Chaves J 2016 Introduction to Nonimaging Optics 2nd edn (Boca Raton, FL: CRC Press)
Estrada J C V, Calle Á H B and Hernández D M 2013 Explicit representations of all refractive
optical interfaces without spherical aberration J. Opt. Soc. Am. A 30 1814–24
Glassner A S 1989 An Introduction to Ray Tracing (Amsterdam: Elsevier)
González-Acuña R G and Chaparro-Romo H A 2018 General formula for bi-aspheric singlet lens
design free of spherical aberration Appl. Opt. 57 9341–5
González-Acuña R G and Guitiérrez-Vega J C 2018 Generalization of the axicon shape: the
gaxicon J. Opt. Soc. Am. A 35 1915–8
González-Acuña R G and Gutiérrez-Vega J C 2019 Analytic formulation of a refractive-reflective
telescope free of spherical aberration Opt. Eng. 58 085105
González-Acuña R G, Avendaño-Alejo M and Gutiérrez-Vega J C 2019a Singlet lens for
generating aberration-free patterns on deformed surfaces J. Opt. Soc. Am. A 36 925–9
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2019b General formula to
design freeform singlet free of spherical aberration and astigmatism Appl. Opt. 58 1010–5
González-Acuña R G, Chaparro-Romo H A and Gutíerrez-Vega J C 1903 Single lens telescope
(arXiv: 1903.11129)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020a Analytic aplanatic
singlet lens: setting and design for three-point objects and images in the meridional plane Opt.
Eng. 59 055104
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020b Analytical Lens
Design (Bristol: IOP Publishing)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020c Analytic solution of
the eikonal for a stigmatic singlet lens Phys. Scr. 95 085201
González-Acuña R G and Gutiérrez-Vega J C 2019a General formula to eliminate spherical
aberration produced by an arbitrary number of lenses Opt. Eng. 58 085106
González-Acuña R G and Gutiérrez-Vega J C 2019b Analytic formulation of a refractive-
reflective telescope free of spherical aberration Opt. Eng. 58 085105
González-Acuña R G and Gutiérrez-Vega J C 2019c General formula for aspheric collimator lens
design free of spherical aberration Current Developments in Lens Design and Optical Engineering
XX ed R B Johnson, V N Mahajan and S Thibault (Bellingham, WA: SPIE) pp 181–4
Kingslake R and Johnson R B 2009 Lens Design Fundamentals (New York: Academic)
Lefaivre J 1951 A new approach in the analytical study of the spherical aberrations of any order
J. Opt. Soc. Am. 41 647

8-16
 Stigmatic Optics

Luneburg R K and Herzberger M 1964 Mathematical Theory of Optics (Berkeley, CA: University
of California Press)
Malacara D 1965 Two lenses to collimate red laser light Appl. Opt. 4 1652–4
Malacara-Hernández D and Malacara-Hernández Z 2016 Handbook of Optical Design (Boca
Raton, FL: CRC Press)
González-Acu na R G and Gutiérrez-Vega J C 2020 Analytic design of a spherochromatic singlet
J. Opt. Soc. Am. A 37 149–53
Schulz G 1983 Achromatic and sharp real imaging of a point by a single aspheric lens Appl. Opt.
22 3242–8
Silva-Lora A and Torres R 2020 Explicit Cartesian oval as a superconic surface for stigmatic
imaging optical systems with real or virtual source or image Proc. R. Soc. A 476
Silva-Lora A and Torres R 2020 Superconical aplanatic ovoid singlet lenses J. Opt. Soc. Am. A37
1155–65
Stavroudis O 2012 Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Feder D P 1954 Automatic Computation of Spot Diagrams J. Opt. Soc. Am.
44 163–70
Sun H Lens Design: A Practical Guide (Boca Raton, FL: CRC Press)
Valencia-Estrada J C and Malacara-Doblado D 2014 Parastigmatic corneal surfaces Appl. Opt.
53 3438–47
Valencia-Estrada J C, Flores-Hernández R B and Malacara-Hernández D 2015 Singlet lenses free
of all orders of spherical aberration Proc. R. Soc. A 471 20140608
Vaskas E M 1957 Note on the Wasserman-Wolf method for designing aspheric surfaces J. Opt.
Soc. Am. 47 669–70
Wassermann G D and Wolf E 1949 On the theory of aplanatic aspheric systems Proc. Phys. Soc.
Sect. B 62 2
Winston R, Miñano J C and Benitez P G 2005 Nonimaging Optics (Amsterdam: Elsevier)
Wolf E 1948 On the designing of aspheric surfaces Proc. Phys. Soc. 61 494
Wolf E and Preddy W S 1947 On the determination of aspheric profiles Proc. Phys. Soc. 59 704
Yang T, Jin G-F and Zhu J 2017 Automated design of freeform imaging systems Light Sci. Appl.
6 e17081

8-17
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 9
The general equation of the stigmatic lenses

In the previous chapter, we studied an on-axis stigmatic singlet lens when the two
refractive surfaces are Cartesian ovals. Here, we present the general formula of the
on-axis stigmatic singlet lens. The input of the general formula is the first surface of
the singlet lens. The first surface needs to be continuous. The output is the correcting
second surface of the lens; the second surface is such that the singlet is stigmatic.

9.1 Introduction
In the previous chapter, we found a particular design of a lens-free spherical
aberration. The lens is made up of two Silva–Torres Cartesian ovals. In this chapter
we introduce a more general expression for spherical aberration-free lenses. We will
call this expression the general equation for stigmatic lenses. Said equation receives
as input parameter a first surface. The equation gives as output a second surface such
that the pair of surfaces form a stigmatic lens.
We will first study the case when the image object distance is finite. Then we will
study when the object is in the minus infinite, when the image is in the plus infinite.
Finally, we will analyze when the object and image are at the minus and plus infinity,
respectively.

9.2 Finite object finite image

The problem to solve in this chapter is announced as: given the first surface of the
lens, the positions of the object and image, the refractive indexes in the object space,
inside the glass and in the image space and central thickness of the singlet, how must
be the second surface such that the singlet is free of spherical aberration?
The answer we are expecting is now the shape of the second surface (zb, rb ), given
the first surface (za, ra ), such that the singlet is free of spherical aberration. Where ra
is the only independent variable. Thus, zb, rb and za are functions of ra. zb is the
sagitta of the second surface and its radius is rb. The sagitta of the first surface is za

doi:10.1088/978-0-7503-3463-1ch9 9-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Figure 9.1. Diagram of an on-axis stigmatic singlet. The first and second surfaces are defined by (za, ra ) and
(zb, rb ), sequentially. The axial interval between the first surface and the object is zo, the axial diameter of the
lens is τ, and the axial length within the second surface and the image is zi. The refraction index in the object
space is no. Inside the lens the refraction index is n and in the image space is ni.

and the radius is ra. As usual, the coordinate system is set at the vertex of za. See
figure 9.1.
We pretend that the object is surrounded by a medium, which has a refractive
index of no. The image is in a medium with a refractive index of ni. The refraction
index n of the lens is constant, and the singlet is radially symmetric. At the centre, the
singlet lens has a thickness of τ. The length from the object to the first surface is zo.
The range from the second surface to the image is zi. Please see figure 9.1.

9.2.1 Fermat principle

We have noticed before that for a stigmatic design the optical path length of all rays that
emanate from a point object on the axis and finish up on a point image on the axis
requirement will be the equivalent. Thus, the optical path length of all rays is constant.
Also, from the Fermat principle, we know that the light travels the paths such that
it spends the least time.
Consequently, combining both statements, we can compare the optical path
length of the axial ray and the optical path length of the non-axial ray.
We start with the optical path length of the axial ray. It comes from −zo , advances
inside the lens and progresses a length τ and ultimately, from the other surface, it
goes a length zi to meet the point image. Please see figure 9.1. The optical path of the
axial ray is given by,

−nozo + nτ + ni zi = constant. (9.1)

The optical path length of a non-axial ray is not simple, since it depends on whether
the object/image is real or virtual.
We start with the more natural case, when the object and the image are real,
zo < 0 and zi > 0,

9-2
 Stigmatic Optics

− nozo + nτ + ni zi = no ra2 + (za − zo)2 + n (rb − ra )2 + (zb − za )2

(9.2)
+ ni rb2 + (zb − τ − zi )2 ,

when zo > 0 and zi > 0, is when the object is virtual and the image is real,

− nozo + nτ + ni zi = − no ra2 + (za − zo)2 + n (rb − ra )2 + (zb − za )2

(9.3)
+ ni rb2 + (zb − τ − zi )2 .

Real object and virtual image: zo < 0 and zi < 0,

− nozo + nτ + ni zi = no ra2 + (za − zo)2 + n (rb − ra )2 + (zb − za )2

(9.4)
− ni rb2 + (zb − τ − zi )2 .

Ultimately, if both are virtual, zo > 0 and zi < 0,

− nozo + nτ + ni zi = − no ra2 + (za − zo)2 + n (rb − ra )2 + (zb − za )2

(9.5)
− ni rb2 + (zb − τ − zi )2 ,

the four previous equations can be combined in a single expression given by

−nozo + nτ + ni zi = −no sgn(zo) ra2 + (za − zo)2 + n (rb − ra )2 + (zb − za )2

(9.6)
+ni sgn(zi ) rb2 + (zb − τ − zi )2 .

Note that the sgn(·) is the sign of their argument, thus if the argument is zero, the
function is not defined. Notice that the two unknowns are (zb, rb ), and we have only
an equation. Thus, we need another one. We are going to get if from Snell’s law.

9.2.2 Snell’s law

Now, we focus on Snell’s law, which is the equation that information has when light
crosses from one medium to another. The vector form of Snell’s law at the first
surface is,

no n2
v2⃗ = [v1⃗ − (n⃗ a · v1⃗ )n⃗ a] − n⃗ a 1 − o2 (n⃗ a × v1⃗ ) 2 for v⃗2 , v1⃗ , n⃗ a ∈  2 , (9.7)
− sgn(zo)n n

where v⃗1 is the unit vector of the incident ray, v⃗2 is unit vector of the refracted ray
and finally n⃗ a is the normal vector of the first surface. Finally −sgn(zo ) comes from
the fact that the object is real/virtual. Please see figure 9.1.

9-3
 Stigmatic Optics

The related unit vectors at the first surface are,

ra e1⃗ + (za − zo)e⃗2 (rb − ra )e1⃗ + (zb − za )e⃗2 za′e1⃗ − e⃗2
v1⃗ = , v2⃗ = , n⃗ a = , (9.8)
ra2 + (za − zo )2 (rb − ra )2 + (zb − za )2 1 + za′2

where e1⃗ is for the r direction and e2⃗ is for the z direction. The idea is to replace the
unit vectors in equation (9.7). The procedure is long, thus we first focus on (n⃗ a × v1⃗ )2 ,

[ra + (za − zo)za′]

(n⃗ a × v1⃗ ) = (e1⃗ × e⃗2), (9.9)
ra2 + (za − zo)2 1 + za′2

notice, that (e1⃗ × e2⃗ )2 = e3⃗ 2 = 1, thus,

[ra + (za − zo)za′]2

(n⃗ a × v1⃗ )2 = . (9.10)
[ra2 + (za − zo)2 ](1 + za′2 )

Thus, the square root in equation (9.7) is,

1 2 (za′e1⃗ − e⃗2) n o2[ra + (za − zo)za′]2

−n⃗ a 1 + ( n⃗ a × v⃗
1 ) = − 1 − . (9.11)
n2 1 + za′2 n 2[ra2 + (za − zo)2 ](1 + za′2 )

In equation (9.7), the term v1⃗ − (n⃗ a · v1⃗ )n⃗ a , is

⎡ ⎤
ra e1⃗ + (za − zo )e ⃗2 raza′ − (za − zo ) ⎥ (za′e1⃗ − e ⃗2) . (9.12)
v1⃗ − (n⃗ a · v1⃗ )n⃗ a = −⎢
ra2 + (za − zo )2 ⎢ r 2 + (z − z ) 2 1 + z ′ 2 ⎥ 1 + z ′ 2
⎣ a a o a ⎦ a

no
Simplifying and multiplying by −sgn(zo )n
the above expression,

no no[(za − zo)za′ + ra ]
[v1⃗ − (n⃗ a · v1⃗ )n⃗ a] = e1⃗
−sgn(zo)n −sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
(9.13)
no[ra + (za − zo)za′]za′
+ e ⃗2 .
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )

Summing equations (9.11) and (9.13), we have v⃗2 as, the output of equation (9.7),
no[(za − zo)za′ + ra ]
v⃗2 = e1⃗
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
no[ra + (za − zo)za′]za′
+ e ⃗2 (9.14)
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
(za′e1⃗ − e⃗2) n o2[ra + (za − zo)za′]2
− 1− .
1 + za′2 n 2[ra2 + (za − zo)2 ](1 + za′2 )

9-4
 Stigmatic Optics

Separating coordinates, e1⃗ and e⃗2 ,

rb − ra no[(za − zo)za′ + ra ]
=
2
(zb − za ) + (rb − ra ) 2
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
n o2[ra + (za − zo)za′]2 (9.15)
1−
n 2[ra2 + (zo − za )2 ](1 + za′ 2 )
− za′ ,
1 + za′ 2

and,
zb − za no[ra + (za − zo)za′]za′
=
2
(zb − za ) + (rb − ra ) 2
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
n o2[ra + (za − zo)za′]2 (9.16)
1−
n 2[ra2 + (zo − za )2 ](1 + za′ 2 )
+ ,
1 + za′ 2

The unknowns zb and rb are only in the left side of equations (9.15) and (9.16).
Notice that the right side of equations (9.15) and (9.16) are the cosine directors inside
the lens. Thus, ℘2r + ℘2z = 1. ℘r is the director in the direction of e1⃗ . ℘z is the director
along e2⃗ . Therefore, we have the following expression if assign ℘r and ℘z in the right
side of equations (9.15) and (9.16),
no[(za − zo)za′ + ra ]
℘r =
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
n o2[ra + (za − zo)za′]2 (9.17)
1−
n 2[ra2 + (zo − za )2 ](1 + za′ 2 )
− za′ ,
1 + za′ 2

and,
no[ra + (za − zo)za′]za′
℘z =
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
n o2[ra + (za − zo)za′]2 (9.18)
1−
n 2[ra2 + (zo − za )2 ](1 + za′ 2 )
+ .
1 + za′ 2

Now, we can simplify the left side of equations (9.15) and (9.16) with the following
parameter,

ϑ≡ (zb − za )2 + (rb − ra )2 . (9.19)

9-5
 Stigmatic Optics

ϑ is the distance traveled by a ray inside the lens. Replacing ϑ in equations (9.15) and
(9.16),
zb − za
= ℘z , (9.20)
ϑ
and,
rb − ra
= ℘r . (9.21)
ϑ
From, the previous equation we can simply solve for zb and rb,
zb = za + ϑ℘z , (9.22)

and,
rb = ra + ϑ℘r . (9.23)

The above expression is the solution that we want to get. It describes, point by point
how the second surface of the lens must be such that it is stigmatic. The only
problem in this equation is that we do not know the length of ϑ. But, we can find the
solution mixing this result with the Fermat principle. In the next section, we are
going to do it!

9.2.3 Solution
In the last section, we obtained how zb and rb must be, but in terms of ϑ. Therefore,
we need an extra equation, and that equation is the Fermat principle that relates the
optical path of an axial ray with the optical path of a non-axial ray. It is essential to
remark that we have just two unknowns zb and rb but we have a system with three
equations. Two equations are given by Snell’s law at the first surface, another one
provided by the Fermat principle. There is nothing wrong here since the equations
granted by Snell’s law are not independent. We recall equation (9.6),

− nozo + nτ + ni zi = − no sgn(zo) ra2 + (za − zo)2

+ n (rb − ra )2 + (zb − za )2 (9.6)
+ ni sgn(zi ) rb2 + (zb − τ − zi )2 .

We assign the following parameter to simplify the last expression,

zf ≡ −nozo + nτ + ni zi + sgn(zo)n 0 ra2 + (zo − za )2 (9.24)

9-6
 Stigmatic Optics

replacing equation (9.24) in equation (9.6),

zf = n (rb − ra )2 + (zb − za )2 + ni rb2 + ( −τ + zb − zi )2 (9.25)

replacing equations (9.22) and (9.23) in equation (9.25),

zf = n ϑ + ni (ra + ϑ℘r)2 + ( −τ + za − zi + ϑ℘z)2 . (9.26)

Also, we assign another parameter to clean equation (9.26),

zτ ≡ −τ + za − zi (9.27)

then, we replace equation (9.27) in equation (9.26),

zf = n ϑ + ni (ra + ϑ℘r)2 + (zτ + ϑ℘z)2 (9.28)

squaring both side of the last equation and manipulating,

(zf − n ϑ)2 = n i2[(ra + ϑ℘r)2 + (zτ + ϑ℘z)2 ] (9.29)

expanding,

z f2 − 2zf n ϑ + n 2 ϑ 2 = n i2[ra2 + 2ra ϑ℘r + ϑ 2℘ 2r + zτ2 + 2zτ ϑ℘z + ϑ 2℘ 2z ]. (9.30)

Remember that ϑ 2℘2r + ϑ 2℘2z = ϑ 2 , thus equation (9.30) is reduced to,

z f2 − 2zf n ϑ + n 2 ϑ 2 = n i2[ra2 + 2ra ϑ℘r + ϑ 2 + zτ2 + 2zτ ϑ℘z] (9.31)

collecting for ϑ,

ϑ 2(n i2 − n 2 ) + ϑ(2zf n + 2n i2ra℘r + 2n i2zτ℘z) + n i2(ra2 + zτ2 ) − z f2 = 0 (9.32)

multiplying equation (9.32) by 4(ni2 − n 2 ),

4(n i2 − n 2 )2 ϑ 2 + 4(n i2 − n 2 )(2zf n + 2n i2ra℘r + 2n i2zτ℘z)ϑ

(9.33)
= −4(n i2 − n 2 )[n i2(ra2 + zτ2 ) − z f2 ]

summing (2zf n + 2ni2ra℘r + 2ni2zτ ℘z)2 in both side of equation (9.33),

4(n i2 − n 2 )2 ϑ 2 + 4(n i2 − n 2 )(2zf n + 2n i2ra℘r + 2n i2zτ℘z)ϑ

+ (2zf n + 2n i2ra℘r + 2n i2zτ℘z)2 (9.34)
= (2zf n + 2n i2ra℘r + 2n i2zτ℘z)2 − 4(n i2 − n 2 )[n i2(ra2 + zτ2 ) − z f2 ]

9-7
 Stigmatic Optics

factoring the left side of equation (9.34),

[2(n i2 − n 2 )ϑ + (2zf n + 2n i2ra℘r + 2n i2zτ℘z)]2
(9.35)
= (2zf n + 2n i2ra℘r + 2n i2zτ℘z)2 − 4(n i2 − n 2 )[n i2(ra2 + zτ2 ) − z f2 ]

applying the square root in both sides,

2(n i2 − n 2 )ϑ + (2zf n + 2n i2ra℘r + 2n i2zτ℘z)
(9.36)
= ± (2zf n + 2n i2ra℘r + 2n i2zτ℘z)2 − 4(n i2 − n 2 )[n i2(ra2 + zτ2 ) − z f2 ]

solving for ϑ,
− 2[zf n + ni 2(ra℘r + zτ ℘z)]
ϑ =
2ni 2 − 2n 2
(9.37)
− 4(ni 2 − n 2)⎡⎣ni 2(ra2 + zτ2) − zf2⎤⎦ + 4⎡⎣zf n + ni 2(ra℘r + zτ ℘z)⎤⎦
2

+
2ni 2 − 2n 2

eliminating the number 2,

− [zf n + ni 2(ra℘r + zτ ℘z)]
ϑ =
ni 2 − n 2
(9.38)
± (ni 2 − n 2)⎡⎣ni 2(ra2 + zτ2) − zf2⎤⎦ + ⎡⎣zf n + ni 2(ra℘r + zτ ℘z)⎤⎦
2

ni 2 − n 2

with equations (9.22), (9.23) and (9.38) we can write the second surface as,
⎧ zb = za + ϑ℘z ,
⎨ (9.39)
⎩ rb = ra + ϑ℘r .

Equation (9.39) is the most important equation in the chapter. It tells us how the
second surface must be, point by point, such that we get a stigmatic lens. It is
significant to observe that equation (9.39) only works for singlet lenses such that the
rays inside them do not cross each other. In other terms, equations (9.22) and (9.23)
tell us that for a point of the first surface, there is a unique point in the second surface
for it, to get archive stigmatism.
Also, it is necessary to remark that in the expression of ϑ, equation (9.38), there is
no expression for sgn(tb ). In other words, the second surface adapts its shape
according to where the image is located; it does not matter if the image is real or
virtual.
Another essential remark about equation (9.38) is that it has a plus–minus sign ±
multiplying the square root, the plus–minus sign leading us to two solutions.
The problem is, which is the solution that works for a given case? Well, it depends
on the sign of the refraction index n and if the object/image is virtual or real. In the

9-8
 Stigmatic Optics

following sections, we are going to show several illustrative examples. But first we
show an expanded version of equation (9.39),

9-9
Stigmatic Optics

9-10
 Stigmatic Optics

The process to get equation (9.39) looks very easy if you read this chapter from
the beginning to here. The process can be easy, but the equation is not, it is gigantic
if you expand it. However, for centuries, people tried to get the general analytical
closed-form solution, but they failed. The secret is that in the whole process, we do
not use any angle. The usual form of Snell’s law complicates everything since there is
no clear relation between the angles and the optical paths. With Snell’s law in its
vector form, it is very easy to see the association between the optical paths and the
director cosines.
Note that we simply did not obtain the general analytical solution in a closed way
for the problem, we also discovered that the solution is unique. We are going to
study this statement in the following chapter.

9.2.4 The eikonal of the stigmatic lens

The eikonal of the stigmatic lens proposed in the last section is a two dimensional
piecewise function divided into three regions. Thus, we can express the eikonal
H (t , ra ) with the order pairs (t , ra ) as its input; and its output is the order pair given
by (kx(t , ra ), ky(t , ra )). Notice that t is the time. Thus, H (t , ra ) = (kx(t , ra ), ky(t , ra )).
The three regions are: from the object to the first surface, inside the lens, and from
the second surface to the image.
There is a one-to-one relationship between the first surface za and the second
surface zb. Thus, any two rays inside the lens do not cross each other, and they meet
only at the object and image points, i.e. the stigmatic pair. For that reason we can
generate the eikonal of the stigmatic lens using the angles θ1, θ2 , and θ3 shown in
figure 9.2. The angle θ1 is subtended between an input ray that strikes the first surface
at za(ra ) and the optical axis, thus θ1 is given by

Figure 9.2. Ray tracing of an on-axis stigmatic singlet. The first and second surfaces are described by (za, ra )
and (zb, rb ), respectively. The axial distance between the first surface and the object is zo, the axial thickness of
the lens is τ, and the axial distance between the second surface and the image is zi. θ1 is the angle subtended
between an input ray that strikes the first surface at za(ra ) and the optical axis. θ 2 is the angle subtended
between the optical axis and the ray travelling inside the lens. θ3 is the angle subtended by the output ray and
optical axis.

9-11
 Stigmatic Optics

⎡ ra ⎤
θ1 = arctan ⎢ ⎥. (9.40)
⎣ za(ra ) − zo ⎦

θ2 is the angle subtended by the optical axis and the ray travelling inside the lens,
thus,
⎡ r (r ) − ra ⎤
θ2 = arctan ⎢ b a ⎥. (9.41)
⎣ zb(ra ) − za(ra ) ⎦

Finally, θ3 is the angle subtended by the output ray and optical axis, thus we have,
⎡ rb(ra ) ⎤
θ3 = arctan ⎢ ⎥. (9.42)
⎣ zb(ra ) − τ − zo ⎦

With these angles in mind it is easy to construct the eikonal H (t , ra ) that follows
the Fermat principle,
H (t , ra ) = (kx(t , ra ), ky(t , ra )), (9.43)
taking no = ni = 1,

⎧ (zo + t cos θ1, t sin θ1), t < c1,

⎪
H (t , ra ) = ⎨ (v(t − c1)cos θ 2, v(t − c1)sin θ 2 ), c1 < t < c 2, (9.44)
⎪
⎩ (zb(ra ) + (t − c 2 )cos θ 3, rb(ra ) + (t − c 2 )sin θ 3), t > c 2.

where v = c /n is the speed of light inside the lens, and c is the speed of light in
vacuum (outside the lens no = ni = 1). c1 is the condition that divides the eikonal in
the first and second region. c2 is the condition that divides the eikonal in the second
and third region.
From figure 9.2, it is easy to see where c1 and c2 come from, outside the lens t < c1
za(ra ) = τ cos θ1 + zo ⟹ c1 = [za(ra ) − zo ]sec θ1, (9.45)
Thus, if c1 > t > c2 we are inside the lens. Now for c2,
zb(ra ) = v(t − c1)cos θ2 ⟹ c2 = c1 + sec θ2[zb(ra ) − za(ra )]/ v . (9.46)
If t > c2 , we are in the third region of the system. Equation (9.44) is an analytical
solution of equation (2.16), the eikonal equation when the object is real.
To plot the eikonals when the object is virtual, we slightly modify H (t , ra ) as
follows,
⎧ (zo + (t − R )cos θ1, (t − R )sin θ1), t < c1,
⎪
H (t , ra ) = ⎨ (v(t − c1)cos θ 2, v(t − c1)sin θ 2 ), c1 < t < c 2, (9.47)
⎪
⎩ (zb(ra ) + (t − c 2 )cos θ 3, rb(ra ) + (t − c 2 )sin θ 3), t > c 2.

where R = zo + nτ and the conditions become,

9-12
 Stigmatic Optics

c1 = (nτ + zo) + sec θ1[za(ra ) − zo ], (9.48)

and,
sec θ2[zb(ra ) − za(ra )]
c2 = c1 + . (9.49)
v
The angles are the same, thus θ1, θ2 , and θ3 are described by equations (9.40), (9.41)
and (9.42), respectively.
Equation (9.47) is the analytical solution of equation (2.16), the eikonal equation
when the object is virtual.

9.2.5 Gallery
In this section, we present several examples of lenses free of spherical aberration
when the object/image is real/virtual.
In the caption of the images are the specification values of each design. All the
lenses presented in this section have been designed computing directly equation
(9.39). No process of optimization has been applied.
The figures of this section are 9.3–9.16.

9.3 Stigmatic aspheric collector

The next example is to study the stigmatic aspheric collector. This case is when
zo → −∞, thus the lens collects the rays from zo. to get the collector lens; we need to
apply the limit when zo → −∞ to some parameters inside equation (9.39). The
parameters are the cosine directors ℘r , ℘z , and zf, equations (9.17), (9.18), and
(9.24). Let’s recall them.
We start with the cosine directors, equation (9.17),
no[(za − zo)za′ + ra ]
℘r =
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
n o2[ra + (za − zo)za′]2 (9.17)
1−
n 2[ra2 + (zo − za )2 ](1 + za′ 2 )
− za′ ,
1 + za′ 2

and equation (9.18),

no[ra + (za − zo)za′]za′
℘z =
−sgn(zo)n ra2 + (zo − za )2 (1 + za′ 2 )
n o2[ra + (za − zo)za′]2 (9.18)
1−
n 2[ra2 + (zo − za )2 ](1 + za′ 2 )
+ .
1 + za′ 2

9-13
 Stigmatic Optics

Figure 9.3. Top: ray tracing, centre: waves, bottom: rays and waves. Design specifications: n = 1.5,
zo = −55 mm , τ = 29 mm , zi = 30 mm , za = 0 , zb= equation (9.39) and H (t, ra )= equation (9.44).

9-14
 Stigmatic Optics

Figure 9.4. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = −55 mm ,
τ = 29 mm , zi = 30 mm , za = (29 − 292 − ra2 ), zb= equation (9.39) and H (t, ra )= equation (9.44).

9-15
 Stigmatic Optics

Figure 9.5. Top: rays in purple, centre: waves in blue, bottom: rays and waves. Design specifications: n = 1.5,
zo = −55 mm , τ = 29 mm , zi = 30 mm , za = −(29 − 292 − ra2 ), zb= equation (9.39) and H (t, ra )=
equation (9.44).

9-16
 Stigmatic Optics

Figure 9.6. Top: rays in purple, centre: waves in blue, bottom: rays and waves. Design specifications: n = 1.5,
zo = −55 mm , τ = 20 mm , zi = 80 mm , za = ra2 /50 , zb= equation (9.39) and H (t, ra )= equation (9.44).

and the parameter zf, equation (9.24),

zf ≡ no sgn (zo) (za − zo) 2 + ra2 + ni zi − nozo + nτ . (9.24)

applying the zo → −∞ in equation (9.17),

⎛ 2 ⎞
(n 2 − n o )(za′) 2 + n 2 ⎟
za′⎜⎜no − n (za′) 2 + 1 ⎟
⎝ n 2[(za′) 2 + 1] ⎠ (9.50)
lim ℘z = 2
,
zo→−∞ n[(za′) + 1]

9-17
 Stigmatic Optics

Figure 9.7. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = −55 mm ,
τ = 12 mm , zi = 80 mm , za = ra2 /50 , zb= equation (9.39) and H (t, ra )= equation (9.44).

and, computing the aforementioned limit in equation (9.18),

2
(n 2 − n o )(za′) 2 + n 2
n 2((za′) 2 + 1) no(za′) 2 (9.51)
lim ℘z = + ,
zo→−∞ (za′) 2 + 1 n(za′) 2 + n

we compute the mentoned limit on parameter zf with equation (9.24),

lim zf = −zano + ni zi + nτ . (9.52)
zo→−∞

9-18
 Stigmatic Optics

Figure 9.8. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = −55 mm ,
τ = 29 mm , zi = −55 mm , za = 0 , zb= equation (9.39) and H (t, ra )= equation (9.44).

9-19
 Stigmatic Optics

Figure 9.9. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = −55 mm ,
τ = 29 mm , zi = −55 mm , za = 29 − 292 − ra2 , zb = equation (9.39) and H (t, ra ) = equation (9.44).

9-20
 Stigmatic Optics

Figure 9.10. Design specifications: n = 1.5, zo = −55 mm , τ = 29 mm , zi = −55 mm , za = −(29 − 292 − ra2 ),
zb = equation (9.39) and H (t, ra ) = equation (9.44).

9-21
 Stigmatic Optics

Figure 9.11. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = 60 mm ,
τ = 29 mm , zi = 50 mm , za = 0 , zb = equation (9.39) and H (t, ra ) = equation (9.47).

Figure 9.12. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = 60 mm ,
τ = 29 mm , zi = 50 mm , za = (29 − 292 − ra2 ), zb = equation (9.39) and H (t, ra ) = equation (9.47).

9-22
 Stigmatic Optics

Figure 9.13. Top: rays, centre: waves, bottom: rays and waves. Design specifications: n = 1.5, zo = 60 mm ,
τ = 29 mm , zi = 50 mm , za = −(29 − 292 − ra2 ), zb = equation (9.39) and H (t, ra ) = equation (9.47).

9-23
 Stigmatic Optics

Figure 9.14. Design specifications: no = ni = 1, n = 1.5, zo = 50 mm , τ = 29 mm , zi = −25 mm , za = 0 ,

zb = equation (9.39) and H (t, ra ) = equation (9.47).

Then, when zo → −∞ becomes ϑ, equation (9.38), becomes,

⎧ ⎡ ⎤2 ⎫
−β ± β 2 + (n 2 − n i2 )⎨n i2ra2 + n i2zτ2 − ⎢ lim (zf )⎥ ⎬
⎩ ⎣zo→−∞ ⎦ ⎭ (9.53)
lim (ϑ) = ,
zo→−∞ (n i2 − n 2 )
where,
⎧⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎫
β ≡ ⎨⎢ lim (zf )⎥n + n i2ra⎢ lim (℘r)⎥ + n i2zτ⎢ lim (℘z)⎥⎬ . (9.54)
⎩⎣zo→−∞ ⎦ ⎣zo→−∞ ⎦ ⎣zo→−∞ ⎦⎭
Therefore, the second surface is given by,
⎡ ⎤⎡ ⎤
lim (zb) = za + ⎢ lim (ϑ)⎥⎢ lim (℘z)⎥ ,
zo→−∞ ⎣ zo→−∞ ⎦⎣ zo→−∞ ⎦
(9.55)
⎡ ⎤⎡ ⎤
lim (rb) = ra + ⎢ lim (ϑ)⎥⎢ lim (℘r)⎥ .
zo→−∞ ⎣zo→−∞ ⎦⎣zo→−∞ ⎦

To properly use (9.55) we need to use (9.50)–(9.54).

It is important to observe the novelty of equation (9.55). We have not made any
change in the procedure to get it; we apply a limit on equation (9.39) and it is
perfectly supported.

9-24
 Stigmatic Optics

Figure 9.15. Design specifications: no = ni = 1, n = 1.5, zo = 50 mm , τ = 29 mm , zi = −25 mm ,

za = (29 − 292 − ra2 ), zb = equation (9.39) and H (t, ra ) = equation (9.47).

In this case, equation (9.55), is only valid for real objects and virtual/real images.
Finally, we want to note that the evaluation of the limit when zo → −∞ is possible
with the vector notation proposed. In another case, using angles with the usual form
of Snell’s law will lead to several difficulties.

9.3.1 The eikonal of the stigmatic collector

In the case of the eikonal of the stigmatic collector, we need to compute some limits,
over equation (9.44) and equations (9.40), (9.41) and (9.42), respectively.
We start with θ1 equation (9.40),
⎡ ra ⎤
lim (θ1) = lim arctan ⎢ ⎥ = 0. (9.56)
zo→−∞ zo→−∞ ⎣ za(ra ) − zo ⎦
θ2 is when zo → −∞, equation (9.41),
⎡ r (r ) − ra ⎤
lim (θ2 ) = lim arctan ⎢ b a ⎥
zo→−∞ zo→−∞ ⎣ zb(ra ) − za(ra ) ⎦
⎡ lim rb(ra ) − ra ⎤ (9.57)
⎢ z →−∞ ⎥
= arctan ⎢ o .
⎢⎣ zolim zb(ra ) − za(ra ) ⎥⎥
→−∞ ⎦

9-25
 Stigmatic Optics

Figure 9.16. Design specifications: no = ni = 1, n = 1.5, zo = 50 mm , τ = 29 mm , zi = −25 mm ,

za = −(29 − 292 − ra2 ), zb = equation (9.39) and H (t, ra ) = equation (9.47).

Finally, θ3, is when zo → −∞, equation (9.42),

⎡ rb(ra ) ⎤
lim (θ3) = lim arctan ⎢ ⎥
zo→−∞ zo→−∞ ⎣ zb(ra ) − τ − zo ⎦
⎡ lim rb(ra ) ⎤ (9.58)
⎢ zo→−∞ ⎥
= arctan ⎢ .
lim z (r ) − τ − zo ⎥⎥
⎣⎢ zo→−∞ b a ⎦

Therefore, H (t , ra ) when zo → −∞, equation (9.44) becomes,

⎧[Zo + t , ra ], t < c1,
⎪
⎪ ⎡ ⎛ ⎞
⎪⎢⎣v(t − c1) cos ⎜⎝zolim θ 2⎟ ,
→−∞ ⎠
⎪
⎪ ⎛ ⎞⎤
⎪v(t − c1) sin ⎜ lim θ2⎟⎥ , c1 < t < c2 ,
lim H (t , ra ) = ⎨ ⎝zo→−∞ ⎠⎦ (9.59)
zo→−∞ ⎪⎡
⎪⎢ lim z (r ) + (t − c ) cos ⎛⎜ lim θ ⎞⎟ ,
⎪⎣zo→−∞ b a 2 3
⎝zo→−∞ ⎠
⎪
⎪ ⎛ ⎞⎤
⎪ zolim rb(ra ) + (t − c2 ) cos ⎜ lim θ3⎟⎥ , t > c2 ,
⎩ →−∞ ⎝zo→−∞ ⎠⎦

9-26
 Stigmatic Optics

where Zo is just a negative constant where the plot starts. The condition c1 becomes,
equation (9.45),
⎛ ⎞
za(ra ) = τ cos ⎜ lim θ1⎟ + Zo ⟹ c1 = [za(ra ) − Zo ], (9.60)
⎝zo→−∞ ⎠

and c2, equation (9.46),

⎛ ⎞ ⎛ ⎞ [z (r ) − za(ra )]
zb(ra ) = v(t − c1) cos ⎜ lim θ2⎟ ⟹ c2 = c1 + sec ⎜ lim θ2⎟ b a . (9.61)
⎝ zo→−∞ ⎠ ⎝ zo→−∞ ⎠ v
In the following section, we are going to show several lenses analogous to the
parabolic mirror. Equation (9.59) is the analytical solution of the eikonal equation,
equation (2.16), for the collector lens.

9.3.2 Gallery
In this section we show a gallery of stigmatic collectors. As usual in the caption is all
the information needed to reproduce the respective example. The figures of the
gallery are 9.17–9.22.

9.4 Stigmatic aspheric collimator

If the collectors at the entrance of the ray are collimated along the optical axis, the
collimator is the opposite. In the collimator the output rays are the ones collimated
along the axis.
To get the collimator we need to compute zi → ∞ in the optical path and then
solve for the unknowns zb and rb. Notice that cosine directors ℘r and ℘z are not
affected by the aforementioned limit. Thus ℘r and ℘z give equations (9.17) and
(9.18).
Here, we need to implement the limit when zi → ∞, this time with the Fermat
principle; we recall the Fermat principle of equation (9.6),

− nozo + nτ + ni zi = − no sgn(zo) ra2 + (za − zo)2

+ n (rb − ra )2 + (zb − za )2 (9.6)
+ ni sgn(zi ) rb2 + (zb − τ − zi )2 ,

replacing ϑ ≡ (rb − ra )2 + (zb − za )2 in equation (9.6),

−zono + nτ + zi ni − ni sgn(zi ) rb2 + (zb − τ − zi )2 − n ϑ =

(9.62)
−no sgn(zo) ra2 + (za − zo)2 .

Notice that on the right side of the above equation nothing depends on zi. Then, it
does not matter what the value of zi is; the right side stays equal. On the other hand,
on the left side, we have two times zi. We can evaluate the limit when zi → ∞ only in
the left side, thus,

9-27
 Stigmatic Optics

Figure 9.17. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
( )
zo → −∞, τ = 15 mm , zi = 50 mm , za = − 50 − 502 − ra2 , zb = equation (9.55) and H (t, ra ) = equation
(9.59).

9-28
 Stigmatic Optics

20

10

-60 -40 -20 20 40 60

-10

-20

20

10

-60 -40 -20 20 40 60

-10

-20

20

10

-60 -40 -20 20 40 60

-10

-20

Figure 9.18. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
( )
zo → −∞, τ = 15 mm , zi = 50 mm , za = 50 − 502 − ra2 , zb = equation (9.55) and H (t, ra ) = equation
(9.59).

⎡ ⎤
lim ⎣ −zono + nτ + zi ni − ni sgn(zi ) rb2 + (zb − τ − zi )2 − n ϑ⎦
zi →∞
⎡ ⎤ ⎡ ⎤ (9.63)
= ( −ni + n )τ − nozo + ni ⎢ lim (zb)⎥ − n⎢ lim (ϑ)⎥ .
⎣zi →∞ ⎦ ⎣zi →∞ ⎦

9-29
 Stigmatic Optics

Figure 9.19. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
zo → −∞, τ = 15 mm , zi = 50 mm , za = 0 , zb = equation (9.55) and H (t, ra ) = equation (9.59).

9-30
 Stigmatic Optics

Figure 9.20. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
zo → −∞, τ = 29 mm , zi = −55 mm , za = 0 , zb = equation (9.55) and H (t, ra ) = equation (9.59).

9-31
 Stigmatic Optics

Figure 9.21. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
zo → −∞, τ = 29 mm , zi = −55 mm , za = 29 − 292 − ra2 , zb = equation (9.55) and H (t, ra ) = equation
(9.59).

9-32
 Stigmatic Optics

Figure 9.22. Design specifications: no = ni = 1, n = 1.5, zo → −∞, τ = 29 mm , zi = −55 mm ,

za − (29 − 292 − ra2 ), zb = equation (9.55) and H (t, ra ) = equation (9.59).

Computing when zi → ∞ in equation (9.63),

⎡ ⎤ ⎡ ⎤
( −ni + n )τ − nozo + ni ⎢ lim (zb)⎥ − n⎢ lim (ϑ)⎥
⎣zi →∞ ⎦ ⎣zi →∞ ⎦ (9.64)
= −no sgn(zo) ra2 + (za − zo) ,2

9-33
 Stigmatic Optics

Let’s replace zb = za + ϑ℘z in the previous equation,

⎡ ⎤ ⎡ ⎤
( −ni + n )τ − nozo + ni ⎢ lim (za + ϑ℘z)⎥ − n⎢ lim (ϑ)⎥
⎣zi →∞ ⎦ ⎣zi →∞ ⎦ (9.65)
= −no sgn(zo) ra2 2
+ (za − zo) .

Since za and ℘z are not affected by the limit when zi → ∞, we have,

⎡ ⎤
( −ni + n )τ − nozo + ni za + ni ⎢ lim (ϑ)⎥℘z
⎣zi →∞ ⎦
(9.66)
⎡ ⎤
= −no sgn(zo) ra2 + (za − zo)2 + n⎢ lim (ϑ)⎥ .
⎣zi →∞ ⎦
⎡ ⎤
Solving for ⎢ lim (ϑ)⎥,
⎣zi →∞ ⎦

⎡ ⎤ ( −ni + n )τ − nozo + ni za + no sgn(zo) ra2 + (za − zo)2

⎢⎣ lim (ϑ)⎥⎦ = . (9.67)
zi →∞ n − ni ℘z

⎡ ⎤
Finally, the solution is given by equation (9.68), where ⎢ lim (ϑ)⎥ is taken from
⎣tb →∞ ⎦
equation (9.67) and ℘r and ℘z are obtained from equations (9.17) and (9.18),

⎧ ⎡ ⎤
⎪ lim (zb) = za + ⎢ lim (ϑ)⎥℘z ,
⎪ zi →∞ ⎣zi →∞ ⎦
⎨ (9.68)
⎪ ⎡ ⎤
⎪ lim (rb) = ra + ⎢ lim (ϑ)⎥℘r .
⎩ zi →∞ ⎣zi →∞ ⎦

Equation (9.68) describes the second surface of the lens collimator.

⎡ ⎤
To properly use equation (9.68), we need equation (9.67) for ⎢ lim (ϑ)⎥ equations
⎣zi →∞ ⎦
(9.17) and (9.18), for ℘r and ℘z respectively.

9.4.1 The eikonal of the stigmatic collimator

We now focus on the eikonal of the stigmatic collimator. The first thing to notice is
that θ1 does not depend on zi, thus

lim (θ1) = θ1. (9.40)

zo→∞

9-34
 Stigmatic Optics

θ2 is when zi → ∞, equation (9.41) becomes,

⎡ lim rb(ra ) − ra ⎤
⎡ rb(ra ) − ra ⎤ ⎢ z →∞ ⎥
lim (θ2 ) = lim arctan ⎢ ⎥ = arctan ⎢ i ⎥. (9.69)
zi →∞ zi →−∞ ⎣ zb(ra ) − za(ra ) ⎦ lim z (
⎢⎣ zi →∞ b ar ) − z ( r
a a ⎥)
⎦

Now we look for θ3, when zi → ∞, equation (9.42) becomes,

⎡ lim rb(ra ) ⎤
⎡ rb(ra ) ⎤ ⎢ zi →∞ ⎥
lim (θ3) = lim arctan ⎢ ⎥ = arctan ⎢ ⎥. (9.70)
zi →∞ zi →∞ ⎣ zb(ra ) − τ − zo ⎦ lim z (
⎢⎣ zi →∞ b a r ) − τ − zo⎥
⎦

Finally, the eikonal H (t , ra ) when zo → ∞ is given,

⎧[zo + t cos θ1, t sin θ1], t < c1,

⎪
⎪⎡ ⎛
⎜
⎞
⎟,
⎪⎣⎢v ( t − c1) cos lim θ 2
⎝zi →∞ ⎠
⎪
⎪ ⎛ ⎞⎤
⎪ v(t − c1) sin ⎜ lim θ2⎟⎥ , c1 < t < c2 ,
lim H (t , ra ) = ⎨ ⎝zi →∞ ⎠⎦ (9.71)
zi →∞ ⎪⎡
⎪⎢ lim z (r ) + (t − c ) cos ⎛⎜ lim θ ⎞⎟ ,
⎪⎣zi →∞ b a 2
⎝zi →∞ ⎠
3

⎪
⎪ ⎛ ⎞⎤
lim
⎪ zi →∞ rb ( ra ) + ( t − c 2 ) cos ⎜ lim θ 3⎟⎥ , t > c2 .
⎩ ⎝zi →∞ ⎠⎦

The condition c1 stays as, equation (9.45),

za(ra ) = t cos(θ1) + zo ⟹ c1 = [za(ra ) − zo ], (9.45)

but c2 is modified too, equation (9.46) becomes,

⎛ ⎞
lim zb(ra ) = v(t − c1) cos ⎜ lim θ2⎟
zo→−∞ ⎝ zo→−∞ ⎠
⎡ ⎤ (9.72)
⎛ ⎢ lim zb(ra ) − za(ra )⎥
⎞ ⎣zo→−∞ ⎦
⟹c2 = c1 + sec ⎜ lim θ2⎟ .
⎝zo→−∞ ⎠ v

9-35
 Stigmatic Optics

To plot the eikonals when the object is virtual we use H (t , ra ) as,

⎧[zo + (t − R ) cos θ1, (t − R ) sin θ1], t < c1,

⎪
⎪⎡ ⎛ ⎞
⎪⎢⎣v(t − c1) cos ⎜⎝zlim i →∞
θ 2⎟ ,
⎠
⎪
⎪ ⎛ ⎞⎤
⎪ v(t − c1) sin ⎜ lim θ2⎟⎥ , c1 < t < c2 ,
lim H (t , ra ) = ⎨ ⎝zi →∞ ⎠⎦ (9.73)
zi →∞ ⎪⎡
⎪⎢ lim z (r ) + (t − c ) cos ⎛⎜ lim θ ⎞⎟ ,
⎪⎣zi →∞ b a 2
⎝zi →∞ ⎠
3

⎪
⎪ ⎛ ⎞⎤
lim
⎪ zi →∞ rb( ra ) + ( t − c2 ) cos ⎜ lim θ 3⎟⎥ , t > c2 .
⎩ ⎝zi →∞ ⎠⎦

where R = zo + nτ . The c1 is given by,

c1 = (nτ + zo) + sec θ1[za(ra ) − zo ], (9.74)
and c2 is given by,
sec θ2[zb(ra ) − za(ra )]
c2 = c1 + . (9.75)
v
The angles θ1, θ2 , and θ3 are given by equations (9.40), (9.69) and (9.70),
respectively.

9.4.2 Gallery
In the next gallery there are several collimator lenses free of spherical aberration. All
the examples have been computed using equation (9.68). The details of each design
are in the caption of the figure that corresponds to it. The figures of the gallery are
9.23–9.28.

9.5 The single-lens telescope

The last case is when zo → −∞ and zi → −∞. Thus, at the input/output the rays are
collimated along the optical axis. This case we call the single-lens telescope
To get the single-lens telescope, the general equation is simple. We need to
compute the limit when zo → −∞ in the cosine directors ℘r , ℘z and in the
parameters f equations (9.17), (9.18) and, (9.24). Or respectively use equations
(9.50), (9.51) and, (9.52).
Then, we need to compute the limit zi → ∞ in the Fermat principle of equation
(9.6). But, we have already done that procedure, and only need to compute the limit
zi → ∞ in ϑ of equation (9.67). Let’s recall it and apply the mentioned limit over it,
⎡ ⎤
⎡ ⎤ ( − ni + n)τ − nozo + ni za + no sgn(zo) ra2 + (za − zo) 2
lim ⎢ lim (ϑ)⎥ = lim ⎢ ⎥ , (9.76)
zo→−∞⎣zi →∞ ⎦ zo→−∞ ⎢ n − ni ℘z ⎥
⎣ ⎦

9-36
 Stigmatic Optics

Figure 9.23. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
zo = −55 mm , τ = 29 mm , zi → ∞, za = 0 , zb = equation (9.68) and H (t, ra ) = equation (9.71).

9-37
 Stigmatic Optics

Figure 9.24. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
zo = −55 mm , τ = 29 mm , zi → ∞, za = 29 − 292 − ra2 , zb = equation (9.68) and H (t, ra ) = equation (9.71).

9-38
 Stigmatic Optics

Figure 9.25. Top: rays, centre: waves, bottom: rays and waves. Design specifications: no = ni = 1, n = 1.5,
zo = −55 mm , τ = 29 mm , zi → ∞, za = −(29 − 292 − ra2 ), zb = equation (9.68) and H (t, ra ) = equation
(9.71).

9-39
 Stigmatic Optics

Figure 9.26. Design specifications: no = ni = 1, n = 1.5, zo = 70 mm , τ = 29 mm , zi → ∞, za = 0 ,

zb = equation (9.68) and H (t, ra ) = equation (9.73).

⎡ ⎤
thus, when we apply the limit on lim ⎢ lim (ϑ)⎥. we have,
zo→−∞⎣zi →∞ ⎦
⎡ ⎤ n (z − τ ) − zano + nτ
lim ⎢ lim (ϑ)⎥ = i a .
⎣
zo→−∞ zi →∞ ⎦ ⎡ ⎤ (9.77)
n − ni ⎢ lim (℘z)⎥
⎣zo→−∞ ⎦

Thus, the second surface of the single-lens telescope is given by,

⎧ ⎡ ⎤ ⎧ ⎡ ⎤⎫⎡ ⎤
⎪ lim ⎢ lim (zb)⎥ = za + ⎨ lim ⎢ lim (ϑ)⎥⎬⎢ lim (℘z)⎥ ,
⎪ zo→−∞⎣zi →∞ ⎦ ⎩zo→−∞⎣zi →∞ ⎦⎭⎣zo→−∞ ⎦
⎨ (9.78)
⎪ ⎡ ⎤ ⎧ ⎡ ⎤⎫⎡ ⎤
⎪ z lim ⎢⎣ lim (rb)⎥⎦ = ra + ⎨ lim ⎢⎣ lim (ϑ)⎥⎦⎬⎢⎣ lim (℘r)⎥⎦ .
⎩ o→−∞ zi →∞ ⎩zo→−∞ zi →∞ ⎭ zo→−∞

Equation (9.78) corresponds to the second surface of the single-lens telescope if and
⎡ ⎤ ⎡ ⎤ ⎡ ⎤
only if the parameters ⎢ lim (℘r)⎥, ⎢ lim (℘z)⎥, and lim ⎢ lim (ϑ)⎥ are given by
⎣ta →−∞ ⎦ ⎣ta →−∞ ⎦ ⎣
ta →−∞ tb →∞ ⎦
equations (9.50), (9.51), and, (9.77), respectively.
Equation (9.78) is so simple; it is just a fraction as simple in comparison with the
other cases when at least one or both object and image are finite. It is the general
equation where every pair of surfaces that compose a single-lens telescope must have
one surface at least with this shape.

9-40
 Stigmatic Optics

Figure 9.27. Design specifications: no = ni = 1, n = 1.5, zo = 70 mm , τ = 29 mm , zi → ∞,

za = (29 − 292 − ra2 ), zb = equation (9.68) and H (t, ra ) = equation (9.73).

Figure 9.28. Design specifications: no = ni = 1, n = 1.5, zo = 70 mm , τ = 29 mm , zi → ∞

za = −(29 − 292 − ra2 ), zb = equation (9.68) and H (t, ra ) = equation (9.73).

9-41
 Stigmatic Optics

9.5.1 The eikonal of the single-lens telescope

To get the eikonal of the single-lens telescope we need to apply limits over the angles
θ1, θ2 , θ3. The limits are when zi → ∞ and zo → −∞.
Starting with θ1, in equation (9.40) we apply the limits zi → ∞ and zo → −∞ and
we get,
⎡ ⎤ ⎧ ⎡ ra ⎤⎫
lim ⎢ lim (θ1)⎥ = lim ⎨ lim arctan ⎢ ⎥⎬ = 0. (9.79)
zi →∞⎣zo→−∞ ⎦ zi →∞⎩zo→−∞ ⎣ za(ra ) − zo ⎦⎭

Now we focus on θ2 , we apply the limits zi → ∞ and zo → −∞ in equation (9.41),

⎡ ⎤ ⎧ ⎡ r (r ) − ra ⎤⎫
lim ⎢ lim (θ2 )⎥ = lim ⎨ lim arctan ⎢ b a ⎥⎬
zi →∞⎣zo→−∞ ⎦ zi →∞⎩zo→−∞ ⎣ zb(ra ) − za(ra ) ⎦⎭
⎧ ⎡ ⎤ ⎫
⎪ lim ⎢ lim rb(ra )⎥ − ra ⎪ (9.80)
⎪ zi →∞⎣zo→−∞ ⎦ ⎪
= arctan ⎨ ⎬.
⎡ ⎤
⎪ lim lim z (r ) − z (r ) ⎪
⎪ ⎢ b a ⎥ a a ⎪
⎩ zi →∞⎣zo→−∞ ⎦ ⎭

Then, it is the turn for θ3, we apply the limits zi → ∞ and zo → −∞ in equation
(9.42),
⎡ ⎤ ⎧ ⎡ rb(ra ) ⎤⎫
lim ⎢ lim (θ3)⎥ = lim ⎨ lim arctan ⎢ ⎥⎬
zi →∞⎣zo→−∞ ⎦ zi →∞⎩zo→−∞ ⎣ zb(ra ) − τ − zo ⎦⎭
⎧ ⎡ ⎤ ⎫
⎪ lim ⎢ lim rb(ra )⎥ ⎪ (9.81)
⎪ zi →∞⎣zo→−∞ ⎦ ⎪
= arctan ⎨ ⎬.
⎪ lim ⎡ lim z (r )⎤ − τ − z ⎪
⎪ ⎢ b a ⎥ o⎪
⎩ zi →∞⎣zo→−∞ ⎦ ⎭

Therefore, the eikonal of the single-lens telescope H (t , ra ) is given by,

⎧[Zo + t, ra ], t < c1,

⎪
⎪ ⎡ ⎛ ⎡ ⎤⎞
⎪ ⎢v(t − c1)cos ⎜ lim ⎢⎣ lim θ 2 ⎥⎦⎟,
⎪ ⎣ ⎝ zi →∞ zo→−∞ ⎠
⎪ ⎛ ⎡ ⎤⎞⎤
⎡ ⎤ ⎪ ⎪v(t − c1)sin ⎜ lim ⎢ lim θ 2 ⎥⎟⎥,
⎣ ⎦⎠⎦
c1 < t < c2 ,
lim ⎢ lim H (t, ra )⎥ = ⎨ ⎝ zi →∞ zo →−∞ (9.82)
zi →∞ ⎣ zo→−∞ ⎦ ⎪
⎪⎢⎡ ⎡ ⎤ ⎛ ⎡ ⎤ ⎞
⎪ ⎣zlim ⎢ lim zb(ra )⎥⎦ + (t − c2 )cos ⎜ lim ⎢⎣ lim θ 3⎥⎦⎟,
i →∞ ⎣ zo→−∞ ⎝zi →∞ zo→−∞ ⎠
⎪
⎪ ⎡ ⎤ ⎛ ⎡ ⎤⎞⎤
⎪ lim ⎢ lim rb(ra )⎥ + (t − c2 )cos ⎜ lim ⎢ lim θ 3⎥⎟⎥, t > c2 ,
⎪ zi →∞ ⎣zo→−∞ ⎦ ⎝zi →∞ ⎣zo→−∞ ⎦⎠⎦
⎩

9-42
 Stigmatic Optics

Figure 9.29. Design specifications: no = ni = 1, n = 1.5, zo = −∞, τ = 29 mm , zi → ∞, za = 29 − 292 − ra2 ,

zb = equation (9.78) and H (t, ra ) = equation (9.82).

9-43
 Stigmatic Optics

Figure 9.30. Design specifications: no = ni = 1, n = 1.5, zo = −∞, τ = 29 mm , zi → ∞,

za = −(29 − 292 − ra2 ), zb = equation (9.78) and H (t, ra ) = equation (9.82).

9-44
 Stigmatic Optics

Figure 9.31. Design specifications: no = ni = 1, n = 1.5, zo = −∞, τ = 29 mm , zi → ∞, za = 0 , zb = equation

(9.78) and H (t, ra ) = equation (9.82).

where Zo is just a negative constant where the plot starts. The condition c1 becomes,

⎛ ⎡ ⎤⎞
za(ra ) = t cos ⎜ lim ⎢ lim (θ1)⎥⎟ + Zo ⟹ c1 = [za (ra ) − Zo ], (9.83)
⎝zi →∞⎣zo→−∞ ⎦⎠

and c2,

⎡ ⎤ ⎛ ⎡ ⎤⎞
lim ⎢ lim zb(ra )⎥ = v(t − c1) cos ⎜ lim ⎢ lim (θ2 )⎥⎟⟹
zi →∞⎣zo→−∞ ⎦ ⎝zi →∞⎣zo→−∞ ⎦⎠
⎡ ⎤ (9.84)
lim ⎢ lim zb(ra )⎥ − za(ra )
⎛ ⎡ ⎤⎞ zi →∞⎣zo→−∞ ⎦
c2 = c1 + sec ⎜ lim ⎢ lim (θ2 )⎥⎟ .
⎝zi →∞⎣zo→−∞ ⎦⎠ v

In the following section, we are going to show several eikonals and single-lens
telescopes.

9.5.2 Gallery
Examples of single-lens telescopes are presented in the following gallery. Each
figure shows the input values with the respective ray tracing. The figures of the
gallery are 9.29–9.31.

9.6 End notes

In this chapter, we have demonstrated that all stigmatic singlet lenses that exist for
real/virtual objects and real/virtual images are in a single equation, equation (9.39).
We found the general equation that describes them; we also found that it is unique.

9-45
 Stigmatic Optics

We tested several for several configurations. When the object is finite/infinite, and
the image is finite/infinite. This is not only when the first surfaces are conics, but also
when the first surfaces are other continuous functions. In all the cases, we obtained
the expected results, when the rays do not cross each other.

Further reading
Born M and Wolf E 2013 Principles of Optics: Electromagnetic Theory of Propagation,
Interference and Diffraction of Light (Amsterdam: Elsevier)
Chaves J 2016 Introduction to Nonimaging Optics 2nd edn (Boca Raton, FL: CRC press)
Descartes R 1637a De la nature des lignes courbes (Leiden)
Descartes R 1637b La Géométrie (Leiden)
González-Acuña R G and Chaparro-Romo H A 2018 General formula for bi-aspheric singlet lens
design free of spherical aberration Appl. Opt. 57 9341–5
González-Acuña R G and Guitiérrez-Vega J C 2018 Generalization of the axicon shape: the
gaxicon J. Opt. Soc. Am. A 35 1915–8
González Acuña R G and Gutiérrez-Vega J C 2019 General formula of the refractive telescope
design free spherical aberration Novel Optical Systems, Methods, and Applications XXII vol
11105 ed C F Hahlweg and J R Mulley (Bellingham, WA: SPIE) pp 162–6
González Acuña R G and Gutiérrez-Vega J C 2019 General formula to design freeform
collimator lens free of spherical aberration and astigmatism Novel Optical Systems,
Methods, and Applications XXII vol 11105 (Bellingham, WA: SPIE) p 111050A
González-Acuña R G and Gutiérrez-Vega J C 2019 Analytic formulation of a refractive-reflective
telescope free of spherical aberration Opt. Eng. 58 085105
González-Acuña R G, Avendaño-Alejo M and Gutiérrez-Vega J C 2019a Singlet lens for
generating aberration-free patterns on deformed surfaces J. Opt. Soc. Am. A 36 925–9
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2019b General formula to
design freeform singlet free of spherical aberration and astigmatism Appl. Opt. 58 1010–5
González-Acuña R G, Chaparro-Romo H A and Gutíerrez-Vega J C 2019c Single lens telescope
(arXiv: 1903.11129)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020a Analytic aplanatic
singlet lens: setting and design for three-point objects and images in the meridional plane Opt.
Eng. 59 055104
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020b Analytical Lens
Design (Bristol: IOP Publishing)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020c Analytic solution of
the eikonal for a stigmatic singlet lens Phys. Scr. 95 085201
González-Acuña R G and Gutiérrez-Vega J C 2019a General formula to eliminate spherical
aberration produced by an arbitrary number of lenses Opt. Eng. 58 085106
González-Acuña R G and Gutiérrez-Vega J C 2019b General formula for aspheric collimator lens
design free of spherical aberration Current Developments in Lens Design and Optical
Engineering XXvol 11104 ed R B Johnson, V N Mahajan and S Thibault (Bellingham,
WA: SPIE) pp 181–4
Huygens C 1690 Traité de la lumière (Leiden)
Kingslake R and Johnson R B 2008 Lens Design Fundamentals (New York: Academic)
Luneburg R K and Herzberger M 1964 Mathematical Theory of Optics (Berkeley, CA: University
of California Press)

9-46
 Stigmatic Optics

González-Acuña R G and Gutiérrez-Vega J C 2020 Analytic design of a spherochromatic singlet

J. Opt. Soc. Am. A 37 149–53
Newton I 1704 Opticks, or, a Treatise of the Reflections, Refractions, Inflections & Colours of Light
(London)
Silva-Lora A and Torres R 2020 Explicit Cartesian oval as a superconic surface for stigmatic
imaging optical systems with real or virtual source or image Proc. R. Soc. A 476
Stavroudis O 2012 The Optics of Rays, Wavefronts, and Caustics vol 38 (Amsterdam: Elsevier)
Stavroudis O N 2006 The Mathematics of Geometrical and Physical Optics: The k-function and its
Ramifications (New York: Wiley)
Stavroudis O N and Feder D P 1954 Automatic computation of spot diagrams J. Opt. Soc. Am.
44 163–70
Wassermann G D and Wolf E 1949 On the theory of aplanatic aspheric systems Proc. Phys. Soc.
Sect. B 62 2
Winston R, Miñano J C and Benitez P G 2005 Nonimaging Optics (Amsterdam: Elsevier)
Wolf E 1948 On the designing of aspheric surfaces Proc. Phys. Soc. 61 494
Wolf E and Preddy W S 1947 On the determination of aspheric profiles Proc. Phys. Soc. 59 704

9-47
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 10
The stigmatic lens and the Cartesian ovals

In the last two chapters we have studied stigmatic lenses. In chapter 8 we study
stigmatic lenses generated by Cartesian ovals with the Silva–Torres equation. In
chapter 9 we deduced the general equation for stigmatic lenses and further deduced
that it is unique. In this chapter we will compare the lenses of chapters 8 and 9, with
this study we will verify the uniqueness of the designs, the general equation of
stigmatic lenses and stigmatism in general.

10.1 Introduction
Up to this point in this treatise, we have studied the origin of the foundations of
geometric optics. Starting from the eikonal, we define the ray and observe that when
all the rays of an object point converge at an image point, we have stigmatism.
Then we began to study the stigmatic systems: conical mirrors and stigmatic
refractive surfaces, Cartesian ovals. We explored different ways of describing
Cartesian ovals. With Cartesian ovals, we designed stigmatic lenses. Later in the
previous chapter, we found the general equation of the stigmatic lens.
The procedure to find the stigmatic lens tells us that the solution is unique. We
have to be precise; the equation is general because it covers all cases of lenses that are
stigmatic. Now given an initial configuration, there is only a second surface that
makes the lens stigmatic. For example given a function za, zo, τ, zi, no, n, ni there is
only a single zb, rb capable of making the system stigmatic.
Although it is evident with the procedure of the previous chapter for everyone, it
is not easy to see because the size of the equations is enormous and can cause doubts,
which is natural. In this chapter, we will generate three stigmatic lenses and compare
their output surfaces. If the surfaces are equal, it means that, as we expected, the
general equation for stigmatic lenses given a configuration gives a single solution.
The three types of astigmatic lenses start from very different procedures, and in
the end, we compare for finite objects and finite images if the second surfaces are
equal.

doi:10.1088/978-0-7503-3463-1ch10 10-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

Additionally, we will make the same comparison for when the object is very far
from the first surface of the stigmatic lens.

10.2 Comparison between the different stigmatic lenses made by

Cartesian ovals
The three systems that we are going to compare are the Silva–Torres stigmatic lenses
from chapter 8, the stigmatic lenses with the general equation of chapter 9 with a first
entrance surface of a Cartesian oval proposed by the Valencia–Calle model in
chapter 6. Finally, we will take a hybrid version of the Silva–Torres Cartesian oval
from chapter 7 and the general equation of stigmatic lenses from chapter 9.
Since all the equations presented from chapter 6 to 9 are unusually long, we will
show stigmatic lenses in sections. After the lenses are presented, the next section will
compare the last surface of the lens.
The idea is that since each model will have a Cartesian oval, it is the same, but
described in different ways; the models must give as a second surface a single
Cartesian oval.
In the section where the models are compared, no equation is given since the
notation in many of the terms is repeated and can be confusing, only the number of
the equations used and the section to which it belongs is quoted.
In this way, we would confirm that all the models are consistent and all arrive at
the same curve and that Nature for this specific problem only accepts one solution.
All roads lead to Rome.
Lastly, before presenting the comparison of the three models, the hybrid between
the Silva–Torres Cartesian oval and the general equation of stigmatic lenses will be
deduced.

10.3 Cartesian ovals in a parametric form

We start with the Silva–Torres stigmatic lenses from chapter 8. So, we first recall the
parameter of the first surface, equations (8.1)–(8.13). We start with the parameters
inside equations (8.1) and (8.8),
The first term is equation (8.3),
2 2
(n 2zin − n o zo) (8.3)
Ka = ,
nno(nzin − nozo)(nzo − nozin)

where, n is the refraction index inside the lens. zo is the distance from the object to
the vertex of the first surface. zin is the image formed by the first surface. Please see
figure 10.1.
We continue with the other parameters as,
nzo − nozin
c 0a = , (8.4)
zinzo(n − no)

10-2
 Stigmatic Optics

Figure 10.1. Diagram of a lens generated by two Cartesian ovals. The first surface is (za, ra ) and the second
Cartesian oval is (zb, rb ). The refraction index in the object space is no, inside the glass the refraction index n
and in the image space ni. The origin is placed in the vertex of the first surface. ρ is the distance from the origin
to a point in the first surface. The distance from the object to the first surface is zo; the distance from the first
surface to the image generated by the first surface is zin. The central thickness of the lens is τ. From the origin to
the image the gap is given by ze and for the distance from the second surface to the image the length is zi.

we take c1a from equation (8.5),

(n − no)(n + no)2
c1a = , (8.5)
4nnozinzo(nzin − nozo)
and, b1a from equation (8.6),

(n + no)(n 2zin − n o2zo)

b1a = . (8.6)
2nnozinzo(nzin − nozo)
With these parameters we can now properly recall the first surface, equation (8.1),

ρ 2 (c1aρ 2 + c 0a )
za = , (8.1)
ρ 2 (2b1a − c02aK a ) + 1 + b1a ρ 2 + 1

and its radius, equation (8.2),

ra = sgn(ρ) ρ 2 − za(ρ)2 . (8.2)

Remember that no in the object area, ni is the refractive index at the image point and
we take no = n1.
We continue with the parameters of the second surface, we take Kb from equation
(8.10),
2
[n o2(ze − τ ) − n 2(zin − τ )]
Kb = . (8.10)
nno[n(ze − τ ) − no(zin − τ )][n(zin − τ ) − no(ze − τ )]

Remember that ze is the distance from the origin of the coordinate system to the
image formed by the second surface and τ is the central thickness of the lens. See
figure 10.1.

10-3
 Stigmatic Optics

c0b is from equation (8.11),

n(ze − τ ) − no(zin − τ )
c 0b = , (8.11)
(n − no)(ze − τ )(zin − τ )
then, from equation (8.12), we have,

(n − no)(n + no)2
c1b = . (8.12)
4nno(ze − τ )(zin − τ )[n(zin − τ ) − no(ze − τ )]

Finally, we recall equation (8.13),

(n + no)[n 2(zin − τ ) − n o2(ze − τ )]

b1b = . (8.13)
2nno(ze − τ )(zin − τ )[n(zin − τ ) − no(ze − τ )]

The second surface is given by equation (8.7), so we recall it,

ρ 2 (c1bρ 2 + c 0b )
zb0 = , (8.7)
ρ 2 (2b1b − c02bK b ) + 1 + b1bρ 2 + 1

where,

zb = τ + zb0(ρ), (8.8)

and the radius of the second surface is by equation (8.9),

rb = sgn(ρ) ρ 2 − zb0(ρ)2 . (8.9)

We are going to use equations of this section to compare them with the stigmatic
lenses of the following sections.

10.4 Cartesian ovals in an explicit form as a first surface and general

equation of stigmatic lenses
Now it is time for the hybrid model when the first surface is given by an explicit
expression of the Cartesian oval described by the paradigm of chapter 6. Therefore,
we take the first surface as of case E in section 6.3.5, equation (6.63),

r2 r4 2r 6 5r 8 2r10 14f 12
za = c2 + c4 3 + c6 + c 8 + c11 + c12 ⋯
2f 8f 32f 5 128f 7 215f 9 2048f 11
∞ (6.63)
Ikr 2k
= ∑ c2k (2f )2k−1
,
k=1

where the coefficients are given by equation (6.65)

10-4
 Stigmatic Optics

⎧ αn + 1
⎪ c2 = − ,
⎪ α (n − 1)
⎪ α 3n 2 + (α 3 + 2α 2 − 2α − 1)n − 1
⎪c4 = − ,
⎪ α 3(n − 1) 2
⎪ 5 3 5 4 3 2 2 5 4 3 2
⎪ c6 = − α n + (2α + 3α − 3α + α + 3α + 1)n + (α + 3α + α − 3α + 3α + 2)n + 1 ,
⎪ 5
α (n − 1) 3
⎨ (6.65)
⎪ 1 7n 4 + (3α 7 + 4α 6 − 4α 5 + 2α 4 + 2α 3 − 4α 2 − 4α + 1)n3)
c
⎪ 8 = − (α
⎪ α 7(n − 1) 4
⎪ 1
⎪− 7 (3α 7 + 8α 6 − 8α 4 + 8α 3 − 8α − 3)n 2
⎪ α (n − 1) 4
⎪ 1
⎪− 7 ((α 7 + 4α 6 + 4α 5 − 2α 4 − 2α 3 + 4α 2 − 4α − 3)n − 1).
⎩ α (n − 1) 4

The expression of (zb, rb ) is given by equation (9.39),

⎧ zb = za + ϑ℘z ,
⎨ (9.39)
⎩ rb = ra + ϑ℘r .

where,

−⎡⎣zf n + n i2(ra℘r + zτ℘z)⎤⎦

(n i − n 2 )⎡⎣n i (ra + zτ ) − z f ⎤⎦ + ⎡⎣zf n + n i (ra℘r + zτ℘z)⎤⎦

± 2 2 2 2 2 2 2 (9.38)
ϑ= ,
n i2 − n 2
and n is the refraction index inside the lens. n 0 = ni are the refraction indices outside
the lens. zo = αf is the distance from the object to the vertex of the first surface. τ is
the central thickness of the lens. zi is the distance from the second surface to the
image. Please see figure 10.2.
We continue with zf by recalling equation (9.24),

zf = −nozo + nτ + ni zi + sgn(zo)n 0 ra2 + (zo − za )2 , (9.24)

and, zτ by recalling equation (9.27),

zτ = −τ + za − zi . (9.27)

Finally, ℘r , ℘z are the cosine director of the ray inside the lens,

2
n o2[ra + (za − zo)za′ ]
1− ⎡ 2 ⎤
no[(za − zo)za′ + ra ]
2
(
n ⎣ra + (zo − za ) ⎦ 1 + za′ 2
2
) (9.17)
℘r = − za′ ,
−sgn(zo)n ra2 2 2
+ (zo − za ) (1 + za′ ) 1 + za′ 2

10-5
 Stigmatic Optics

Figure 10.2. Diagram of an on-axis stigmatic singlet. The first is (za, ra ) and second surface is (zb, rb ). Here, in
this chapter we take the first surface as a polynomial expansion of a Cartesian oval. The gap from the object to
the first is zo, the axial thickness of the lens is τ, and the gap from the second surface to the image is zi. The
refraction index in the object space is no. The refraction index of the lens is n and in the image space, the
refraction index is ni. Notice that the origin is placed in the vertex of the first surface, the polynomial expansion
of a Cartesian oval.

and,
2
n o2[ra + (za − zo)za′ ]
1− ⎡ 2 ⎤
no[ra + (za − zo)za′]za′
2
(
n ⎣ra + (zo − za ) ⎦ 1 + za′ 2
2
) (9.18)
℘z = + .
−sgn(zo)n ra2 2
+ (zo − za ) (1 + za′ ) 2 1 + za′ 2

Notice that the following condition holds, ℘2r + ℘2z = 1.

These are the representative equations for the stigmatic lens of this section.

10.5 Cartesian ovals in a parametric form as a first surface and

general equation of stigmatic lenses
Since we do not deduce the expression for a Cartesian oval in a parametric form as a
first surface and general equation of stigmatic lenses in this section, we are going to
demonstrate it.
We start with the notation. n is the refraction index inside the lens. n 0 = ni are the
refraction indices outside. zo is the gap between the object and the first surface. The
central thickness of the lens is given by τ. zo is the gap between the image and the
second surface. Please see figure 10.3.

10.5.1 First surface

Now, let’s focus first on the first surface. The first surface is a Cartesian oval under
the paradigm of Silva–Torres. Therefore, it is in a parametric form where the
independent variable is ρ.

10-6
 Stigmatic Optics

Figure 10.3. An on-axis stigmatic singlet. The first surface is (za, ra ) and second surface is (zb, rb ). The gap
from the object to the first is zo, the axial thickness of the lens is τ, and the length from the second surface to the
image is zi. The refraction index around the object is no. The refraction index of the lens is n and around the
image, the refraction index is ni.

So, for Ka we recall equation (8.3),

2 2
(n 2zin − n o zo) (8.3)
Ka = ,
nno(nzin − nozo)(nzo − nozin)
for c0a we use equation (8.4),
nzo − nozin
c 0a = , (8.4)
zinzo(n − no)
then, we recall equation (8.5),
(n − no)(n + no)2
c1a = . (8.5)
4nnozinzo(nzin − nozo)
Finally, we use equation (8.6) for b1a

(n + no)(n 2zin − n o2zo)

b1a = . (8.6)
2nnozinzo(nzin − nozo)
Given the parameters inside the first surface, we can recall it as a Cartesian oval
under the paradigm of Silva–Torres with equation (8.1),
ρ 2 (c1aρ 2 + c 0a )
za = , (8.1)
ρ 2 (2b1a − c02aK a ) + 1 + b1a ρ 2 + 1

whose ratio is equation (8.2),

ra = sgn(ρ) ρ 2 − za(ρ)2 . (8.2)

10-7
 Stigmatic Optics

Notice that zin is the image formed by the first surface. Since the second surface is in
terms of a Cartesian oval under the paradigm of Silva–Torres, the general equation
of stigmatic lenses zin will not be as relevant as in the model of Silva–Torres. Hence it
is not presented in the diagram of figure 10.3.

10.5.2 Second surface

Notice that the first surface is in a parametric form, where the independent variable
is ρ. The general equation of stigmatic lenses presented in chapter 9 receives as first
surfaces explicit expressions, not parametric ones. Therefore, we need to adapt it
such that the first surface is an expression in a parametric form, where the
independent variable is ρ. In this section, we are going to deduce the adaptation
above of the general equation of stigmatic lenses.
We begin by recalling the Snell law at the first surface.

no n2
v2⃗ = [v1⃗ − (n⃗ a · v1⃗ )n⃗ a] − n⃗ a 1 + o2 (n⃗ a × v1⃗ ) 2 for v2⃗ , v1⃗ , n⃗ a ∈ 2, (10.1)
− sgn(zo)n n

where v1⃗ is the unit vector of the incident ray, v2⃗ is unit vector of the refracted ray and
finally n⃗ a is the normal vector of the first surface. Notice the term −sgn(zo ), it comes
from the fact that the object can be real or virtual. Please see figure 10.3.
The unit vectors at the first surface are,
ra e⃗1 + (za − zo)e⃗ 2 (rb − ra )e⃗1 + (zb − za )e⃗ 2 za′e⃗1 − ra′e⃗ 2
v1⃗ = , v2⃗ = , n⃗ a = , (10.2)
ra2 + (za − zo)2 (rb − ra )2 + (zb − za )2 ra′2 + za′2

where e⃗1 is for the r direction and e⃗2 is for the z direction.
Also, notice that the normal vector has been changed because now the derivatives
∂r ∂z
are in respect to ρ and not ra. Thus, ra′ = ∂ρa and za′ = ∂ρa .
We need to replace the unit vectors of equations (10.2), in equation (10.1). We
start with (n⃗ a × v1⃗ ),

[rara′ + (za − zo)za′]

(n⃗ a × v1⃗ ) = (e1⃗ × e⃗ 2), (10.3)
ra2 + (za − zo)2 ra′2 + za′2

squaring,

[rara′ + (za − zo)za′]2

(n⃗ a × v1⃗ )2 = − . (10.4)
[ra2 + (za − zo)2 ](ra′2 + za′2 )

Then the square root of equation (10.1) is,

n o2 ′ ′
2 = − (za e ⃗1 − ra e ⃗ 2) 1 − [rara′ + (za − zo)za′]2
− n⃗ a 1 + ( n ⃗ a × v1⃗ ) . (10.5)
n2 ra′2 + za′2 n 2[ra2 + (za − zo) 2](ra′2 + za′2 )

10-8
 Stigmatic Optics

Let us focus on the other terms of equation (10.1), v1⃗ − (n⃗ a · v1⃗ )n⃗ a ,
ra e1⃗ + (za − zo)e⃗ 2
v1⃗ − (n⃗ a · v1⃗ )n⃗ a =
ra2 + (za − zo)2
⎡ ⎤ (10.6)
raza′ − (za − zo)ra′ ⎥ (za′e⃗1 − ra′e⃗ 2) .
−⎢
⎢⎣ r + (z − z )2 r′2 + z ′2 ⎥⎦ r′2 + z ′2
2
a a o a a a a

Manipulating and multiplying by no /( −sgn(zo )n

no no[(za − zo)za′ + rara′]
[v1⃗ − (n⃗ a · v1⃗ )n⃗ a] = e ⃗1
−sgn(zo)n −sgn(zo)n ra2 + (zo − za )2 (1 + za′2 )
(10.7)
no(ra + (za − zo)za′)za′
+ e⃗ 2.
−sgn(zo)n ra2 + (zo − za )2 (1 + za′2 )

Summing equations (10.5) and (10.7), we have v2⃗ of equation (10.1) as we wanted,
no[(za − zo)za′ + rara′]ra′
v2⃗ = e1⃗
−sgn(zo)n ra2 + (zo − za )2 (ra′2 + za′2 )
no[rara′ + (za − zo)za′]za′
+ e⃗ 2 (10.8)
−sgn(zo)n ra2 + (zo − za )2 (ra′2 + za′2 )
(za′e1⃗ − ra′e⃗ 2) n o2[rara′ + (za − zo)za′]2
− 1− .
ra′2 + za′2 n 2[ra2 + (zo − za )2 ](ra′2 + za′2 )

Separating the coordinates of the last equation, e⃗1 and e⃗2 ,

rb − ra no[(za − zo)za′ + rara′]ra′
=
(zb − za )2 + (rb − ra )2 −sgn(zo)n ra2 + (zo − za )2 (ra′2 + za′2 )
2 (10.9)
n o2[ra ra′ + (za − zo)za′ ]
1− ⎡ 2 ⎤
2 2
(
n ⎣ra + (zo − za ) ⎦ ra′ + za′ 2
2
)
− za′ ,
ra′2 + za′2

and,
zb − za no[rara′ + (za − zo)za′]za′
=
(zb − za )2 + (rb − ra )2 −sgn(zo)n ra2 + (zo − za )2 (ra′2 + za′2 )
2 (10.10)
n o2[ra ra′ + (za − zo)za′ ]
1− ⎡ 2 ⎤
2
(
n ⎣ra + (zo − za ) ⎦ ra′ + za′ 2
2 2
)
+ ra′ ,
ra′2 + za′2

10-9
 Stigmatic Optics

Notice that in the left side of the equations, again, there are only parameters that we
know. Thus, we assign them to the new parameters ℘r and ℘z ,
2
n o2 [rara′ + (za − zo )za′]
1− ⎡ ⎤
n 2 ⎣ra2 + (zo − za )2 ⎦ ra′ 2 + za′ 2
( ) (10.11)
no[(za − zo )za′ + rara′]ra′
℘r = − za′ ,
− sgn(zo )n ra2 + (zo − za )2 (ra′2 + za′2 ) ra′2 + za′2

and,
2
n o2 [rara′ + (za − zo )za′]
1− ⎡ 2 ⎤ 2
no[rara′ + (za − zo )za′]za′
2 2
(
n ⎣ra + (zo − za ) ⎦ ra′ + za′ 2 ) (10.12)
℘z = + ra′ .
− sgn(zo )n ra2 + (zo − za )2 (ra′2 + za′2 ) ra′2 + za′2

It is important to remark that ℘r and ℘z are the cosine directors traveling inside the
lens. ℘r is the cosine director with direction r and ℘z is the cosine director with
direction z. Thus, ℘2r + ℘2z = 1.
For derivation proposes, we assign a name to the distance traveled by each ray
inside the lens as,

ϑ≡ (zb − za )2 + (rb − ra )2 . (10.13)

Replacing equation (10.13) in equations (10.11) and (10.12),

zb − za
= ℘z , (10.14)
ϑ
and,
rb − ra
= ℘r . (10.15)
ϑ
Solving for zb and rb,
zb = za + ϑ℘z , (10.16)
and,
rb = ra + ϑ℘r . (10.17)
We get the same structure of chapter 8; we just need to know what ϑ is. For that, we
will follow a similar procedure. Thus, we recall the Fermat principle. The Fermat
principle optical path length of the axial ray is equal to the optical path length of the
non-axial ray,

− nozo + nτ + ni zi = n (rb − ra )2 + (zb − za )2

+ ni rb2 + ( −τ + zb − zi )2 (10.18)
+ sgn(zo)no ra2 + (zo − za )2 .

10-10
 Stigmatic Optics

Let’s assign the following parameter to clean the last equation,

zf ≡ −nozo + nτ + ni zi + sgn(zo)n 0 ra2 + (zo − za )2 (10.19)

replacing equation (10.19) in equation (10.18),

zf = n (rb − ra )2 + (zb − za )2 + ni rb2 + ( −τ + zb − zi )2 (10.20)

replacing equations (10.17) and (10.16) in equation (10.20),

zf = n ϑ + ni (ra + ϑ℘ 2r + ( −τ + za − zi + ϑ℘z)2 (10.21)

we assign a new parameter to clean equation (10.21),

zτ ≡ −τ + za − zi (10.22)
then, we replace equation (10.22) in equation (10.21),
zf = n ϑ + ni (ra + ϑ℘r)2 + (zτ + ϑ℘z)2 (10.23)
squaring both side of equation (10.23) and manipulating,
(zf − n ϑ)2 = n i2[(ra + ϑ℘r)2 + (zτ + ϑ℘z)2 ], (10.24)
expanding,
z f2 − 2zf n ϑ + n 2 ϑ 2 = n i2[ra2 + 2ra ϑ℘r + ϑ 2℘ 2r + zτ2 + 2zτ ϑ℘z + ϑ 2℘ 2z ], (10.25)

solving for ϑ,
−⎡⎣zf n + n i2(ra℘r + zτ℘z)⎤⎦

(n i − n 2 )⎡⎣n i (ra + zτ ) − z f ⎤⎦ + ⎡⎣zf n + n i (ra℘r + zτ℘z)⎤⎦

± 2 2 2 2 2 2 2 (10.26)
ϑ= ,
n i2 − n 2
then, we get almost the same equation, the sagitta is
zb = za + ϑ℘z (10.27)
and the radius,
rb = ra + ϑ℘r . (10.28)
The difference between equations (10.27), (10.28) and equation (9.39) is that the first
ones are for the first surface in a parametric form. Equation (9.39) is for the first
surface in an explicit form.
The procedure is almost the same as the one performed in chapter 9, but now the
independent variable is ρ instead of ra.
The equations of this section, that we use in the following comparison, are
equations (10.16), (10.17), (10.19), (10.22), (10.27) and (10.28).
The following expressions are the extended version of (10.27) and (10.28). We
show them to raise awareness of the complexity and length of these expressions, but
we always refer to them as equations (10.27) and (10.28).

10-11
Stigmatic Optics

10-12
Stigmatic Optics

10-13
 Stigmatic Optics

10.6 Illustrative comparison

We have three different models of stigmatic lenses with very different procedures to
obtain the solution.
The method in section 10.3 starts from a very extensive procedure presented in
chapters 7 and 8.
In section 10.4, take the first surface from chapter 6 and the second surface
from chapter 9. How the first surface of chapter 6 is obtained compared to
chapters 7 and 8 is very different. The first comes from solving a differential
equation and the second from an algebraic manipulation of a consequence of the
Fermat principle.
It is essential to mention that all the presented equations of the first and above all
of the second surface are massive equations that at first glance are not seen to give
the same or a similar result.
In the following figures, figures 10.4–10.6, a series of lenses described by the
methods of the three sections is shown. The continuous gray line is for the refractive
surfaces of section 10.3. The dashed purple line is for the refraction surfaces

25

20

15

10

r
-10 -5 5 10

Figure 10.4. Overlapping curves are presented representing the refractive surfaces of a stigmatic lens on the
axis. The three curves that are superimposed were computed with the equations of sections 10.3–10.5. The
continuous gray line is for the refractive surfaces of section 10.3. The dashed purple line is for the refraction
surfaces computed using the equations of section 10.4. Finally, the dotted blue line represented the lines
obtained by the equations of section 10.5. Input values: no = ni = 1, n = 2, α = 1.5, f = 40 mm, zo = −αf ,
zin = f , ze = 70 mm , τ = 20 mm , zi = ze − τ .

10-14
 Stigmatic Optics

12

10

r
-10 -5 5 10

Figure 10.5. Input values: no = ni = 1, n = 2, α = 1.5, f = 50 mm, zo = −αf , zin = f , ze = 70 mm , τ = 10 mm ,

zi = ze − τ .

12

10

r
-10 -5 5 10

Figure 10.6. Input values: no = ni = 1, n = 2, α = 2 , f = 50 mm, zo = −αf , zin = f , ze = 70 mm , τ = 10 mm ,

zi = ze − τ .

computed using the equations of 10.4. Finally, the dotted blue line represented the
lines obtained by the equations of section 10.5. Note that for all examples, the gray,
dotted purple, and dotted blue lines are overwritten (figure 10.7).
This result is no accident and confirms what we predicted in chapter 9. Given the
input values, the curves of the stigmatic lenses are unique.
We have three very different paradigms, the functions that describe them are very
complex, lengthy and varied, but when they are plotted, they give the same curve.
Although in this section, we limit ourselves to the examples already mentioned,
we have compared around a hundred curves with different design specifications. In
all cases, the same curve is always obtained.
In the following four sections, we will demonstrate that this equality is not unique
in the case where objects and images are finite. We will take the object from minus
infinity and make the same comparison.
Following the same order, the next section will be for the collector proposed by
Silva–Torres, section 10.7. In section 10.8, it will be the turn for the hybrid collector

10-15
 Stigmatic Optics

15

10

r
-10 -5 5 10

Figure 10.7. Coinciding curves are shown describing the refractive surfaces of a stigmatic lens. The three
curves that are overlapped, were computed with the equations of sections 10.3–10.5. The constant grey line is
for the refractive surfaces of section 10.3. The dashed purple line is for the refraction surfaces of section 10.4.
The dotted blue line represented the lines of the equations of section 10.5. Input values: no = ni = 1, n = 1.5,
α = 1.45, f = 50 mm, zo = −αf , zin = f , ze = 70 mm , τ = 15 mm , zi = ze − τ .

lens model between the Cartesian oval described by Valencia–Calle and the general
equation of stigmatic lenses. Section 10.9 presents the collector lens described by the
hybrid method that makes up the first surface of a Silva–Torres Cartesian oval and
the general equation for stigmatic lenses. Finally, in section 10.10, there is another
illustrative comparison.

10.7 Cartesian ovals in a parametric form for an object at minus

infinity
We begin the new comparison with the model of Silva–Torres of a stigmatic lens for
zo → −∞.
Therefore, we take lim K a from equation (8.14),
zo→−∞

n o2
lim K a = − . (8.14)
zo→−∞ n2
To get lim c0a we recall equation (8.15),
zo→−∞
n
lim c 0a = . (8.15)
zo→−∞ zin(n − no)
lim c1a is given by equation (8.16),
zo→−∞

lim c1a = 0 (8.16)

zo→−∞

10-16
 Stigmatic Optics

and lim b1a is taken from equation (8.17),

zo→−∞

lim b1a = 0. (8.17)

zo→−∞

The sagitta when lim is described with equation (8.17),

zo→−∞

⎛ ⎞
⎜ lim c 0a⎟ρ 2
⎝zo→−∞ ⎠
lim za = (8.18)
zo→−∞ ⎛ ⎞2 ⎛ ⎞
1 − ⎜ lim c 0a⎟ ⎜ lim K a⎟ρ 2 + 1
⎝zo→−∞ ⎠ ⎝zo→−∞ ⎠
and the radius, equation (8.19),

lim ra = sgn(ρ) ρ 2 − lim (za2 ) . (8.19)

zo→−∞ zo→−∞

The parameters of the second surface are the following,

2
[n o2(zi − τ ) − n 2(zin − τ )]
Kb = (8.10)
nno[n(zi − τ ) − no(zin − τ )][n(zin − τ ) − no(zi − τ )]

n(zi − τ ) − no(zin − τ )
c 0b = . (8.11)
(n − no)(zi − τ )(zin − τ )
We recall equation (8.12),
(n − no)(n + no)2
c1b = . (8.12)
4nno(zi − τ )(zin − τ )[n(zin − τ ) − no(zi − τ )]
Finally, lim b1b is recalled with equation (8.13),
zo→−∞

(n + no)(n 2(zin − τ ) − n o2(zi − τ ))

b1b = . (8.13)
2nno(zi − τ )(zin − τ )[n(zin − τ ) − no(zi − τ )]
The second surface of the stigmatic lens proposed by Silva–Torres is,
ρ 2 (c1bρ 2 + c 0b )
zb0 = (8.7)
ρ 2 (2b1b − c02bK b ) + 1 + b1bρ 2 + 1

where,
zb = τ + zb0(ρ), (8.8)
and,
rb = sgn(ρ) ρ 2 − zb0(ρ)2 . (8.9)
The equations of this section, equations (8.7)–(8.19), are the ones used in section 10.10.

10-17
 Stigmatic Optics

10.8 Cartesian ovals in an explicit form for an object at minus

infinity
Now we introduce the Cartesian oval when zo → −∞ is in its explicit form, with
equation (6.44).
n⎡⎣(n − 1)f − sign(f ) (n − 1)n 2[(n − 1)f 2 − (n + 1)r 2 ] ⎤⎦
za = . (6.44)
n2 − 1
Given the Cartesian oval when the limit zo → −∞ is in its explicit form, the second
surface is described by equation (9.55),
⎧ ⎡ ⎤⎡ ⎤
⎪
⎪ lim (zb) = za + ⎣⎢lim (ϑ)⎦⎥⎣⎢lim (℘z)⎦⎥ ,
⎨ (9.55)
⎪ lim (r ) = r + ⎡lim (ϑ)⎤⎡lim (℘ )⎤ ,
⎪ b a
⎣⎢ ⎦⎥⎣⎢ ⎥
r ⎦
⎩

where, lim (ϑ) is described by equation (9.53),

zo→−∞

⎧ ⎡ ⎤2 ⎫
−β ± β 2 + (n 2 − n i2 )⎨n i2ra2 + n i2zτ2 − ⎢ lim (zf )⎥ ⎬
⎩ ⎣zo→−∞ ⎦ ⎭ (9.53)
lim (ϑ) = 2
,
zo→−∞ (n i − n 2 )
where,
⎧⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎫
β ≡ ⎨⎢ lim (zf )⎥n + n i2ra⎢ lim (℘r)⎥ + n i2zτ⎢ lim (℘z)⎥⎬ , (9.54)
⎩⎣zo→−∞ ⎦ ⎣zo→−∞ ⎦ ⎣zo→−∞ ⎦⎭

and,
lim zf = −zano + ni zi + nτ . (9.52)
zo→−∞

Finally, to get the cosine directors when lim , we recall equations (9.50) and (9.51),
zo→−∞

⎛ ⎞
′⎜ ′ 2
za no − n (za ) + 1
(n 2 − n o2 )(za′) + n 2
2
⎟
⎜ ⎡ 2
n 2⎢(za′) + 1⎥
⎤ ⎟ (9.50)
⎝ ⎣ ⎦ ⎠
lim ℘z = 2
,
zo→−∞ n[(za′) + 1]
and,

(n2 − n o2)(za′)2 + n2
( 2
n 2 (za′) + 1 ) no(za′)2 (9.51)
lim ℘z = + .
zo→−∞ (za′)2 + 1 n(za′)2 + n

The above equations are the ones implemented in the comparison of section 10.10.

10-18
 Stigmatic Optics

10.9 Cartesian ovals in a parametric form as a first surface and

general equation of stigmatic lenses for an object at minus
infinity
Finally, we arrive at our last model, when the first surface is a Cartesian oval of
Silva–Torres and the second surface is the general equation of stigmatic lenses.
We start by recalling lim K a with equation (8.14),
zo→−∞

n o2
lim K a = − . (8.14)
zo→−∞ n2
lim c0a is with equation (8.15),
zo→−∞
n
lim c 0a = (8.15)
zo→−∞ zin(n − no)
if we recall equation (8.16), we recall lim c1a ,
zo→−∞

lim c1a = 0 (8.16)

zo→−∞

and,
lim b1a = 0. (8.17)
zo→−∞

The sagitta of the first surface when zo → −∞ is

⎛ ⎞
⎜ lim c 0a⎟ρ 2
⎝zo→−∞ ⎠
lim za = (8.18)
zo→−∞ ⎛ ⎞2 ⎛ ⎞
1 − ⎜ lim c 0a⎟ ⎜ lim K a⎟ρ 2 + 1
⎝zo→−∞ ⎠ ⎝zo→−∞ ⎠

and its radius is,

lim ra = sgn(ρ) ρ 2 − lim (za2 ) . (8.19)
zo→−∞ zo→−∞

Now, the second surface is when limit zo → −∞ is applied over equations (10.27)
and (10.28),
⎧ ⎡ ⎤⎡ ⎤
⎪
⎪ lim(zb) = lim(za ) + ⎣⎢lim(ϑ)⎦⎥⎣⎢lim(℘z)⎦⎥ ,
⎨ (10.29)
⎪ lim(r ) = lim(r ) + ⎡lim(ϑ)⎤⎡lim(℘ )⎤ ,
⎪ b a
⎣⎢ ⎦⎥⎣⎢ ⎥
r ⎦
⎩
where, lim (ϑ) is the limit when over equation (10.26),
zo→−∞

β− α + β2
lim (ϑ) = . (10.30)
zo→−∞ n 2 − n i2

10-19
 Stigmatic Optics

α is,
⎧ ⎡ ⎤ ⎫
α ≡ (n 2 − n i2 )⎨n i2⎢ lim (ra2 )⎥ + zτ2 − lim (z f2 )⎬ . (10.31)
⎩ ⎣zo→−∞ ⎦ zo→−∞ ⎭

β is,
⎡ ⎤
β ≡ lim (zf )n + n i2⎢ lim (ra ) lim (℘r) + zτ lim (℘z)⎥ (10.32)
zo→−∞ ⎣ zo→−∞ zo→−∞ zo→−∞ ⎦

zf when zo → −∞ is given by
lim zf = lim (nτ + ni zi − noza ) (10.33)
zo→−∞ zo→−∞

zτ when zo → −∞ is given by
lim zτ = lim ( −τ + za − zi ). (10.34)
zo→−∞ zo→−∞

Finally, when zo → −∞,the cosine directors of equations (10.11) and (10.12) are
given by
⎧
⎪
⎪ no[(za − zo)za′ + rara′]ra′
lim ℘r = lim ⎨
zo→−∞ zo→−∞ ⎪ − sgn(z )n r 2 + (z − z ) 2 r′2 + z ′2
o a o a ( a a )
⎪
⎩
(10.35)
n 2 [ r r + (z − z )z ]
2 ⎫
1 − 2⎡ 2o a a′ a 2⎤ o 2 a′ 2 ⎪
n ⎣ra + (zo − za ) ⎦(ra′ + za′ ) ⎪
− za′ ⎬
ra′2 + za′2 ⎪
⎪
⎭

and,
⎧
⎪
⎪ no[rara′ + (za − zo)za′]za′
lim ℘z = lim ⎨
zo→−∞ zo→−∞ ⎪ − sgn(z )n r 2 + (z − z ) 2 r′2 + z ′2
o a o a ( a a )
⎪
⎩
(10.36)
n 2 [ r r + (z − z )z ]
2 ⎫
1 − 2⎡ 2o a a′ a 2⎤ o 2 a′ 2 ⎪
n ⎣ra + (zo − za ) ⎦(ra′ + za′ ) ⎪
+ ra′ ⎬
ra′2 + za′2 ⎪
⎪
⎭

10-20
 Stigmatic Optics

computing the aforementioned limit over equations (10.35) and (10.36),

⎛ ⎞
′⎜ ′ 2
za nora − n ra′ + za′ 2
( n 2 − n o2 )za′ 2 + n 2ra′ 2
⎟
⎜ n 2(ra′ 2 + za′ 2 ) ⎟ (10.37)
⎝ ⎠
lim ℘r = 2 2
zo→−∞ n(ra′ + za′ )
and,

ra′
(n2 − n o2)za′2 + n2ra′2
(
n 2 ra′ 2 + za′ 2 ) noza′2 (10.38)
lim ℘z = + .
zo→−∞ ra′2 + za′2 n(ra′2 + za′2 )

Equations (8.14)–(8.19), (10.29)–(10.38) of this section are the ones implemented to

design a singlet stigmatic collector lens with the paradigm proposed in this section.

10.10 Illustrative comparison

Once again, we compare the three paradigms presented in the previous three
sections. This time the comparison is for collector lenses, that is, zo → −∞.
In figures 10.8–10.11 are lenses described by the methods of the previous three
sections. The constant grey line is for the refractive surfaces of section 10.7. The
dashed purple line is for the refraction surfaces of section 10.8. The dotted blue line
represents the lines of the equations of section 10.9. Again, in all examples the black
line, dotted purple line and dotted blue line are overwritten.
We repeat that this is not a coincidence; we have tested hundreds of examples.
This result only confirms that Nature has made stigmatic systems unique for each
particular configuration. There is an infinite number of stigmatic lenses, but for a
given first surface, provided indices of refraction, the distance between a given object
and image, and given a central thickness, there is only one solution.
Given a distance zi and zo → −∞, there is only a single reflective surface that
causes the rays emerging from the object at zo → −∞ to converge on the image at zi.
That curve is a parabola given by

ra2
za = . (10.39)
4zi
z
20

15

10

r
-40 -20 20 40

Figure 10.8. Input values: no = ni = 1, n = 1.7 , zo = −∞, zin = 150 mm , ze = 70 mm, τ = 20 mm , zi = ze − τ .

10-21
 Stigmatic Optics

z
20

15

10

r
-30 -20 -10 10 20 30

Figure 10.9. Input values: no = ni = 1, n = 1.5, zo = −∞, zin = 140 mm , ze = 70 mm , τ = 20 mm , zi = ze − τ .

z
20

15

10

r
-30 -20 -10 10 20 30

Figure 10.10. Input values: no = ni = 1, n = 1.5, zo = −∞, zin = 100 mm , ze = 70 mm , τ = 20 mm ,

zi = ze − τ .

15

10

r
-30 -20 -10 10 20 30

Figure 10.11. Input values: no = ni = 1, n = 1.5, zo = −∞, zin = 100 mm , ze = 70 mm , τ = 15 mm ,

zi = ze − τ .

There are no others, Nature does not allow that! This is what we discovered in
chapter 5, and in chapter 7, and we reaffirm it.
The implications of this comparison are not trivial and show a profound
characteristic nature of light under SVEA and stigmatism as an implication of it.
In the following section, we discuss these implications in more detail.

10-22
 Stigmatic Optics

10.11 Implications
The implications of the uniqueness of stigmatic surfaces are not left in the mirror or
single-lens stigmatic systems.
In our previous treatise, Analytical Lens Design, in chapters 11 and 12, we show
the generalization of the general equation of stigmatic lenses in chapter 9.
The generalization consists in giving an arbitrary series of the refractive surfaces
as it should be one last refractive surface of the system for the system to become
stigmatic on the axis. Not only is the shape of the refractive surfaces known, but also
their central thicknesses, and their indices of refraction.
The procedure for finding said last refractive surface is very similar to that
presented in chapter 9, which, as in this chapter, predicts that the last refractive
surface is unique for a given configuration. In other words, given the shapes of the
refractive surfaces, their central thickness and the index of refraction, there is only
one additional refractive surface such that the system is stigmatic.
Resolving the system for the last refractive surface does not limit the model in
general, since for example, with the comparisons prescribed, we can say that given
the input parameters and the first surface of the lens, there is the only one-second
surface that makes up for the first with the ability to make the lens stigmatic.
So now we talk about a pair of surfaces that a stigmatic lens makes. For example,
if our first surface is a sphere, with the general equation of stigmatic lenses, we can
design an aspheric surface such that, given the already mentioned first surface, the
system is stigmatic.
But the geometric optics is reversible, that is to say, given the aforementioned
stigmatic surface, there is a spherical surface that makes the system stigmatic on the
axis.
The same thing happens if we have more surfaces. Given a series of refractive
surfaces, there is only one that can group with the previous refractive surfaces such
that the complete system is stigmatic on the axis. So once you see that the system is
stigmatic on the axis, there is no way to distinguish which surface does the work, all
of them work together to make the system stigmatic on the axis.
The standard for optical design is the use of commercial optical design software to
design all kinds of lenses and many systems with a lot of lenses. Commercial optical
design software uses a series of algorithms that optimize an optical system to help
the user to reduce aberrations of the optical system in question. In general, what is
expected is that the system is more or less stigmatic on the optical axis. The filter to
accept a design, among other considerations, is that at least the spot diagram is
inside the air disk for the object that is on-axis that produced an on-axis image.
To achieve this, the commercial optical design software uses purely numerical
algorithms that approximate the system to the fulfilment of the requirement above.
On the other hand, it is essential to note that in the optical design industry, the
use of mostly spherical surfaces is encouraged, due to the simplicity of their
production.
The exciting thing here is that, if we start from a system made by pure spheres and
such that it has already been optimized to be more or less stigmatic on the axis, we

10-23
 Stigmatic Optics

can describe the outermost surface as we described in chapters 11 and 12 of

Analytical Lens Design and that refractive surface will look similar to a sphere.
The equations presented in chapters 11 and 12 of Analytical lens design will
approximate the more or less last refractive spherical surface of the system above,
depending on how stigmatic the system is.
The same is true for a lens made in commercial software or designed with the
methodology in chapter 9. Given the input values, the second surface of the lens
designed in commercial software will be so similar to the second surface of the lens in
chapter 9, like how stigmatic it is. All of this because of the nature of stigmatism
confirmed with the comparisons made in this chapter.
It may be that mathematical expressions of the commercial software look very
different from the expression of chapter 9. But the curves will look very similar if the
expression of the commercial software is close to being stigmatic. All this is taking
the section of the curves that are exposed by the axis object and that converge in the
axis image.
Therefore, we can understand the existence and uniqueness of stigmatism.
When a problem of initial value mathematically models a physical situation, the
existence and uniqueness of the solution is of utmost importance, since it is surely
expected to have a solution, because physically something must happen. On the
other hand, the solution is supposed to be unique, because if we repeat the
experiment under identical conditions, we can expect the same results, as long as
the model is deterministic. Therefore, when considering an initial value problem of
the stigmatism, it is natural to ask the following.
Existence: will there be a solution to the problem? From the results of chapters 6–
10 we know that stigmatic lenses exist. Uniqueness: if there is a solution, will it be
unique? From equation (9.39), the general equation of stigmatic lenses we know that
stigmatism is unique. The results of this chapter just confirm it. Determination: if
there is a solution, how do we determine it? Geometrical optics is deterministic.

10.12 End notes

In this chapter, we compared a series of stigmatic lenses generated by Cartesian
ovals that are described by different approaches. But it does not imply that the
model is the one chosen, given the same input values for all the standards compared,
the results are precisely the same.
This implies that stigmatism, given the input values, supports a single solution.

Further reading
Estrada J C V, Calle Á H B and Hernández D M 2013 Explicit representations of all refractive
optical interfaces without spherical aberration J. Opt. Soc. Am. A 30 1814–24
González Acuña R G and Gutiérrez-Vega J C 2019 General formula of the refractive telescope
design free spherical aberration Novel Optical Systems, Methods, and Applications XXII vol
11105 ed C F Hahlweg and J R Mulley (Bellingham, WA: SPIE) pp 162–6

10-24
 Stigmatic Optics

González Acuña R G and Gutiérrez-Vega J C 2019 General formula to design freeform

collimator lens free of spherical aberration and astigmatism Novel Optical Systems,
Methods, and Applications XXII vol 11105 (Bellingham, WA: SPIE) p 111050A
González-Acuña R G and Chaparro-Romo H A 2018 General formula for bi-aspheric singlet lens
design free of spherical aberration Appl. Opt. 57 9341–5
González-Acuña R G and Guitiérrez-Vega J C 2018 Generalization of the axicon shape: the
gaxicon J. Opt. Soc. Am. A 35 1915–8
González-Acuña R G and Gutiérrez-Vega J C 2019 Analytic formulation of a refractive-reflective
telescope free of spherical aberration Opt. Eng. 58 085105
González-Acuña R G, Avendaño-Alejo M and Gutiérrez-Vega J C 2019a Singlet lens for
generating aberration-free patterns on deformed surfaces J. Opt. Soc. Am. A 36 925–9
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2019b General formula to
design freeform singlet free of spherical aberration and astigmatism Appl. Opt. 58 1010–5
González-Acuña R G, Chaparro-Romo H A and Gutíerrez-Vega J C 2019c Single lens telescope
(arXiv:1903.11129)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020a Analytic aplanatic
singlet lens: setting and design for three-point objects and images in the meridional plane Opt.
Eng. 59 055104
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020b Analytical Lens
Design (Bristol: IOP Publishing)
González-Acuña R G, Chaparro-Romo H A and Gutiérrez-Vega J C 2020c Analytic solution of
the eikonal for a stigmatic singlet lens Phys. Scr. 95 085201
González-Acuña R G and Gutiérrez-Vega J C 2019a General formula to eliminate spherical
aberration produced by an arbitrary number of lenses Opt. Eng. 58 985106
González-Acuña R G and Gutiérrez-Vega J C 2019b Analytic formulation of a refractive-
reflective telescope free of spherical aberration Opt. Eng. 58 085105
González-Acuña R G and Gutiérrez-Vega J C 2019c General formula for aspheric collimator lens
design free of spherical aberration Current Developments in Lens Design and Optical
Engineering XX vol 11104 ed R B Johnson, V N Mahajan and S Thibault (Bellingham,
WA: SPIE) pp 181–4
González-Acuña R G and Gutiérrez-Vega J C 2020 Analytic design of a spherochromatic singlet
J. Opt. Soc. Am. A 37 149–53
Silva-Lora A and Torres R 2020 Explicit Cartesian oval as a superconic surface for stigmatic
imaging optical systems with real or virtual source or image Proc. R. Soc. A 476
Silva-Lora A and Torres R 2020 Superconical aplanatic ovoid singlet lenses J. Opt. Soc. Am. A 37
1155–65
Valencia-Estrada J C and Malacara-Doblado D 2014 Parastigmatic corneal surfaces Appl. Opt.
53 3438–47
Valencia-Estrada J C, Flores-Hernández R B and Malacara-Hernández D 2015 Singlet lenses free
of all orders of spherical aberration Proc. R. Soc. A 471 20140608

10-25
 IOP Publishing

Stigmatic Optics
Rafael G González-Acuña and Héctor A Chaparro-Romo

Chapter 11
Algorithms for stigmatic design

11.1 Programs for chapter 6

11.1.1 Case: real finite object—real finite image

In[1]:= ClearAll["Global‘*"]
(** Definition of set variables **)

n+1
C2 = ;
n−1

n+1
C4 = ;
n−1

(n+1)(n2 +6n+1)
C6 = ;
(n−1)3

n+1
C8 = ;
n−1

(n+1)(7n4 +124n3 +122n2 +124n+7)

C10 = ;
(n−1)5

(n+1)(3n4 −44n3 −46n2 −44n+3)

C12 = ;
(n−1)5

ra2 ra4 2ra6 5ra8 2ra10

za[ra_] := C2 + C4 + C6 + C8 + C10 + C12
2f 8 f3 32f5 128f7 512f9
14ra12
;
2048f11

doi:10.1088/978-0-7503-3463-1ch11 11-1 ª IOP Publishing Ltd 2020

 Stigmatic Optics

no(za’ [ra] (ra+(−zo+za[ra]) za’ [ra]))

 z[ra_] := 
n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2

 2

1 √ no2 (ra+(−zo+za[ra]) za’ [ra])
+  1− 2
1+(za’ [ra])
2 n2 (ra2 +(zo−za[ra])2 ) (1+(za’ [ra]) )

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n
 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(ra)/(za[ra] −zi)];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 16;
density = 4;
zo = −40;
zi = 40;
f = −zo

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],

11-2
 Stigmatic Optics

Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]}
}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]}
}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo+ n zi,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{tt,0,−zo+ n zi},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-3
 Stigmatic Optics

11.1.2 Case: real infinity object—real finite image

In[2]:=
(** Definition of set variables **)


n((n−1)f−Sign[f] (n−1)((n−1)f^2−(n+1)ra2 ))
za[ra_]:= ;
(n2 −1)
  

za [ra]
2 1 1 1
 z[ra_]:= + 1+ −1+ ;
n+n za [ra] 1+za [ra]
2
n2 2
1+za [ra]
2

   
 
za [ra] 
2  1 1

 r[ra_]:= 1−n 1+za [ra] 1+ 2 −1+ ;
n (1+za [ra] )
2
1+za [ra]
2
n

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[Limit[(ra)/(−zo+za[ra]),zo→
→−Infinity]]

θ 2[ra_] := ArcTan[(ra)/(za[ra]−zi )];

c1[ra_] := (−Zo+za[ra]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 16;
density = 4;
(* zo −> −Infinity *)
Zo = −55;
zi = 40;
f=zi;

11-4
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{ra,−rmax,rmax},
PlotRange→→All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+n zi,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{tt,0,−Zo+ n zi},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-5
 Stigmatic Optics

11.1.3 Case: real infinity object—virtual finite image

In[3]:=
(** Definition of set variables **)


n((n−1)f−Sign[f] (n−1)((n−1)f^2−(n+1)ra2 ))
za[ra_]:= ;
(n2 −1)

  
za [ra]
2
1  1 1
 z[ra_]:= + 1+ 2 −1+ ;
n+n za [ra]
2
1+za [ra]
2
n
1+za [ra]
2

   
 
za [ra]  1 1
1+za [ra] 1+ 2 −1+
 2
 r[ra_]:= 1−n ;
n (1+za [ra] )
2
1+za [ra]
2
n

(** Homotopy for Eikonal **)

→−Infinity]]
θ 1[ra_] := ArcTan[Limit[(ra)/(−zo+za[ra]),zo→

θ 2[ra_] := ArcTan[(ra)/(za[ra]−zi )];

c1[ra_] := (−Zo+za[ra]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 16;
density = 4;
(* zo −> −Infinity *)
Zo = −55;
zi = −40;
f = zi;

11-6
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],
Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo −n zi,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{tt,0,−Zo − n zi},
PlotRange→→All,PlotStyle →{Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-7
 Stigmatic Optics

11.1.4 Case: real finite object—virtual finite image

In[4]:= ClearAll["Global‘*"]
(** Definition of set variables **)


n((n−1)f−Sign[f] (n−1)((n−1)f^2−(n+1)ra2 ))
za[ra_]:= ;
(n2 −1)

za [ra]
2
1 1 1
 z[ra_]:= + 1+ (−1+ );
 2 2
1+za [ra]
2
n+n za [ra] n
1+za [ra]
2


za [ra] 1 1
1+za [ra]
2
 r[ra_]:= (1−n 1+ (−1+ ));
 2
n2 1+za [ra]
2
n (1+za [ra] )

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 16;
density = 4;
zo = −55;
zi = −40;
f =zi;

11-8
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},
{ ra,−rmax,rmax},
PlotRange→→All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo − n zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{tt,0,−zo+ n zi+0.25},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-9
 Stigmatic Optics

11.1.5 Case: real finite object—real infinite image

In[5]:= ClearAll["Global‘*"]
(** Definition of set variables **)

(n−1)f−Sign[f] (n−1)((n−1)f2 +(n+1)ra2 )
za[ra_]:=
(n2 −1)

za [ra] (ra+(−zo+za[ra]) za [ra])

 z[ra_]:= 
−Sign[zo]n ra 2 +(zo−za[ra])2 (1+(za [ra]) )
2



1  (ra+(−zo+za[ra]) za [ra])
2
+ 1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

ra+(−zo+za[ra]) za [ra]

 r[ra_]:= 
−Sign[zo]n ra 2 +(zo−za[ra])2 (1+(za [ra]) )
2




za [ra]  (ra+(−zo+za[ra]) za [ra])
2
− 1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 16;
density = 4;
zo = −40;
(* zi −> Infinity*)
Zi = 20;
f = zo;

11-10
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},

PlotRange→→All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo + n Zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,−zo + n Zi + 0.25},
PlotRange→→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-11
 Stigmatic Optics

11.1.6 Case: virtual finite object—real infinite image

In[6]:= ClearAll["Global‘*"]
(** Definition of set variables **)


(n−1)f−Sign[f] (n−1)((n−1)f2 +(n+1)ra2 )
za[ra_]:=
(n2 −1)

za [ra] (ra+(−zo+za[ra]) za [ra])

 z[ra_]:= 
−Sign[zo]n ra 2 +(zo−za[ra])2 (1+(za [ra]) )
2



1  (ra+(−zo+za[ra]) za [ra])
2
+ 1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

ra+(−zo+za[ra]) za [ra]

 r[ra_]:= 
−Sign[zo]n ra 2 +(zo−za[ra])2 (1+(za [ra]) )
2




za [ra]  (ra+(−zo+za[ra]) za [ra])
2
− 1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

c1[ra_] := (zo + n Zi) + Sec[θθ 1[ra]](−zo + za[ra]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 16;
density = 4;
zo = 40;
(* zi −> Infinity*)
Zi = 20;
f = zo;

11-12
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(n zo+n τ ))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(n zo+n τ )),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0, zo + n Zi ,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(n zo+n τ ))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(n zo+n τ )),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{tt,0, zo + n Zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-13
 Stigmatic Optics

11.1.7 Case: virtual finite object—virtual finite image

In[7]:= ClearAll["Global‘*"]
(** Definition of set variables **)
α n+1
C2 = − ;
α (n−1)

α3 +2α
α 3 n2 +(α α2 −2α
α −1)n+1
C4 = − ;
α 3 (n−1)2

α5 +3α
α 5 n3 +(2α α4 −3α
α3 +α
α2 +3α
α +1)n2 +(α
α5 +3α
α4 +α
α3 −3α
α2 +3α
α+2)n−1
C6 = − ;
α 5 (n−1)3

1
C8 = − α7 n4 +(3α
(α α7 +4α
α6 −4α
α5 +2α
α4 +2α
α3 −4α
α2 −4α
α +1)n3
α 7 (n−1)4

α 7 +8α
+(3α α 6 −8α
α 4 +8α
α 3 −8α
α −3)n2 +(α
α 7 +4α
α 6 +4α
α 5 −2α
α 4 −2α
α 3 +4α
α 2 −4α
α
−3)n−1);
ra2 ra4 2ra6 5ra8
za[ra_] := C2 + C4 3
+ C6 5
+ C8 ;
2f 8f 32f 128f7

za [ra] (ra+(−zo+za[ra]) za [ra])

 z[ra_]:= 
−Sign[zo]n ra2 +(zo−za[ra])2 (1+(ra [ra]) )
2



1  (ra+(−zo+za[ra]) za [ra])
2
+ 1− ;
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

ra+(−zo+za[ra]) za [ra]

 r[ra_]:= 
−Sign[zo]n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2



za [ra]  (ra+(−zo+za[ra]) za [ra])
2
− 1− ;
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(ra)/(za[ra] −zi)];

c1[ra_] := (zo + n zi) + Sec[θθ 1[ra]](−zo+za[ra]);

11-14
 Stigmatic Optics

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

α = 1.5;
rmax = 16;
density = 4;
zo = α 40;
zi = −40;
f = zo

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(zo + n zi))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(zo + n zi)),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,zo+n τ − n zi+0.25,density}];

11-15
 Stigmatic Optics

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(zo + n zi))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(zo + n zi)),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
}]},

{tt,0,zo+ n − n zi+0.25},
PlotRange→→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11.2 Programs for chapter 7

11.2.1 Case 1: real finite object—real finite image

In[8]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

11-16
 Stigmatic Optics

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(ra[ρρ])/(−za[ρρ] + zi)];

c1[ρρ_] := Sec[θθ 1[ρρ]](za[ρρ] − zo);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

ρ max = 16;
density = 4;
zo = −40;
zi = 40;
zin = n zi;

(** Plot Section **)

SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ 1.1*zi,−ra[ρρ]}},{ρρ,−ρρmax,ρρmax},
PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]

11-17
 Stigmatic Optics

OvalSurfaceV:=ParametricPlot[{za[ρρ],ra[ρρ]},{ρρ,− ρ max−250,ρρmax+250},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,− zo + n zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}}]},

{tt,0,−zo + n zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-18
 Stigmatic Optics

11.2.2 Case 2: real infinity object—real finite image

In[9]:=
(** Definition of set variables **)

no2
Ka = − ;
n2
n
coa = ;
n zi−no zi
c1a = 0;

b1a = 0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no za [ρρ]
2
ra [ρρ]
 z[ρρ_] := +
n (ra [ρρ] + za [ρρ] )
2 2
ra [ρρ] + za [ρρ]
2 2
 
√ n2 ra [ρρ]2 + (n2 −no2 ) za [ρρ]2
n2 (ra [ρρ] +za [ρρ] )
2 2


za [ρρ]
 r[ρρ_] := no ra [ρρ]−n
ρ]2 +za [ρρ]2 )
n (ra [ρ
 
 2  ρ 2
] +(n2 −no2 )za [ρρ]
2
2√ n ra [ρ
ra [ρρ] +za [ρρ]
2

n2 (ra [ρρ] +za [ρρ] )

2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[Limit[(ra[ρρ])/(−zo+za[ρρ]),zo→
→−Infinity]]

θ 2[ρρ_] := ArcTan[(ra[ρρ])/( zi − za[ρρ] )];

c1[ρρ_] := −Zo+za[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

11-19
 Stigmatic Optics

ρ max = 16;
density = 4;
(* zo −> −Infinity *)
Zo = −40;
zi = 40;
zin = n zi;
(** Plot Section **)
SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ 1.1*zi,−ra[ρρ]}},{ρρ,−ρρmax,ρρmax},
PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]

OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−250,ρρmax+250},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]
WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}}],
Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo + n zi+0.25,density}];
RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}}],
Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},
{tt,0,−Zo+ n zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];
Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-20
 Stigmatic Optics

11.2.3 Case 3: real infinity object—virtual finite image

In[10]:=
(** Definition of set variables **)

no2
Ka = − ;
n2
n
coa = ;
n zi−no zi
c1a = 0;

b1a = 0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no za [ρρ]
2
ra [ρρ]
 z[ρρ_] := +
 2  2
n (ra [ρρ] + za [ρρ] ) 2
ra [ρρ] + za [ρρ]
2
 
2
√ n2 ra [ρρ] + (n2 −no2 ) za [ρρ] 2

n2 (ra [ρρ] +za [ρρ] )

2 2


za [ρρ]
 r[ρρ_] := no ra [ρρ]−n
ρ]2 +za [ρρ]2 )
n (ra [ρ
 
 2  ρ 2
] +(n2 −no2 )za [ρρ]
2
 2  2√ n ra [ρ
ra [ρρ] +za [ρρ]
n2 (ra [ρρ] +za [ρρ] )
2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[Limit[(ra[ρρ])/(−zo+za[ρρ]),zo→
→−Infinity]]

θ 2[ρρ_] := ArcTan[(ra[ρρ])/(zi−za[ρρ] )];

c1[ρρ_] := (−Zo+za[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

11-21
 Stigmatic Optics

ρ max = 16;
density = 4;
(* zo −> −Infinity *)
Zo = −40;
zi = 40;
zin = n zi;
(** Plot Section **)
SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ 1.1*zi,−9ra[ρρ]}},{ρρ,−ρρmax,ρρmax},
PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]
OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−250,ρρmax+250},
PlotStyle→→Directive[{Black,Dashed}]]
LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]
WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt+ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+n zi+0.25,density}];
RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],
Piecewise[{
{Sin[θθ 1[ρρ]]tt+ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},
{tt,0,−Zo+ n zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-22
 Stigmatic Optics

11.2.4 Case 4: real finite object—virtual finite image

In[11]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[ra[ρρ]/(zi−za[ρρ] )];

c1[ρρ_] := Sec[θθ 1[ρρ]](−zo+za[ρρ]);

11-23
 Stigmatic Optics

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;

n = 1.5;

ρ max =12;
density =3;
zo = − 40;
zi = −15;
zin = n zi;

(** Plot Section **)

SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ −1.1*zi,−5 ra[ρρ]}},{ρρ,−ρρmax,ρρmax},

PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]

OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−150,ρρmax+150},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
}],
Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
}]},

{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo− n zi+0.25,density}];

11-24
 Stigmatic Optics

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
}]},
{tt,0,−zo − n zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11.2.5 Case 5: real finite object—real infinite image

In[12]:= ClearAll["Global‘*"]
(** Definition of set variables **)

n2
Ka = − ;
no2
no
coa = − ;
n zo−no zo
c1a = 0;

b1a =0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ 2 + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

11-25
 Stigmatic Optics

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

c1[ρρ_] := Sec[θθ 1[ρρ]](−zo+za[ρρ]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

ρ max = 24;
density = 4;
zo = −40;
(* zi −> Infinity*)
Zi = 25;

(** Plot Section **)

SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ −1.1*zi, −ra[ρρ]}},{ρρ,−ρρmax,ρρmax},

PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]

OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−30,ρρmax+30},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]

11-26
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v (tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo + n Zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v (tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{tt,0,−zo + n Zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-27
 Stigmatic Optics

11.2.6 Case 6: virtual finite object—real infinite image

In[13]:= ClearAll["Global‘*"]
(** Definition of set variables **)

n2
Ka = − ;
no2
no
coa = − ;
n zo−no zo

c1a = 0;

b1a = 0;
ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ2+ 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n

 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

11-28
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

c1[ρρ_] := (zo + n Zi) + Sec[θθ 1[ρρ]](−zo + za[ρρ]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

ρ max = 16;
density = 4;
zo = 25;
(* zi −> Infinity*)
Zi = 40;

(** Plot Section **)

SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ 1.1*zo, −ra[ρρ]}},{ρρ,−ρρmax,ρρmax},

PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]

OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−30,ρρmax+30},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],

Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]

11-29
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−( zo+n Zi)) + zo,tt<=c1[ρρ]},
{v (tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + n Zi)),tt<=c1[ρρ]},
{ ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0, zo + n Zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−( zo+n Zi)) + zo,tt<=c1[ρρ]},
{v (tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + n Zi)),tt<=c1[ρρ]},
{ ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{tt,0, n zo +n Zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-30
 Stigmatic Optics

11.2.7 Case 7: virtual finite object—real finite image

In[14]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

11-31
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[ra[ρρ]/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[ra[ρρ]/(zi−za[ρρ])];

c1[ρρ_] := (zo − n zi) + Sec[θθ 1[ρρ]](−zo+za[ρρ]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

ρmax = 12;
density = 2;
zo = 40;
zi = 25;

(** Plot Section **)

SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ 1.1 * zi, −ra[ρρ]}},{ρρ,−ρρmax,ρρmax},

PlotStyle →{Red,Transparent},AspectRatio→ →Automatic,AxesLabel →{r,z}]

OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−250,ρρmax+250},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo + n zi))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

11-32
 Stigmatic Optics

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + n zi)),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,zo+zi+10,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo + n zi))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + n zi)),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{tt,0,zo+zi+10},

PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},

{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-33
 Stigmatic Optics

11.2.8 Case 8: virtual finite object—virtual finite image

In[15]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ 2 + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n

 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

11-34
 Stigmatic Optics

θ 2[ρρ_] := ArcTan[ra[ρρ]/(zi−za[ρρ] )];

c1[ρρ_] := (zo − n zi) + Sec[θθ 1[ρρ]](−zo+za[ρρ]);

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
n = 1.5;

ρmax = 16;
density = 4;
zo = 25;
zi = −40;
zin = n zi;

(** Plot Section **)

SurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]},{ −n* zi+n, −3.5ra[ρρ]}},{ρρ,−ρρmax,

ρ max},
PlotStyle → {Red,Transparent},AspectRatio→ →Automatic,AxesLabel → {r,z}]

OvalSurfaceV:=ParametricPlot[{{za[ρρ],ra[ρρ]}},{ρρ,− ρ max−250,ρρmax+250},
→Directive[{Black,Dashed}]]
PlotStyle→

LensV:=Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[ Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],List[{0,0}],#[[2,1]],#[[3, 1]],List[{0,0}],#[[4,1]]]&
[Cases[SurfaceV,_Line,∞∞]]]}]

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo + nττ ))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

11-35
 Stigmatic Optics

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + nττ )),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,4 zo − zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo + nττ ))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]}
}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + nττ )),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]}
}]},

{tt,0,4*zo − zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-36
 Stigmatic Optics

11.2.9 Case 9: real infinite object—real infinite image

In[16]:=
(** Definition of set variables **)

no2
Ka = − ;
n2
coa = 0;

c1a = 0;

b1a = 0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no za [ρρ]
2
ra [ρρ]
 z[ρρ_] := +
n (ra [ρρ] + za [ρρ] )
2 2
ra [ρρ] + za [ρρ]
2 2
 
√ n2 ra [ρρ]2 + (n2 −no2 ) za [ρρ]2
n2 (ra [ρρ] +za [ρρ] )
2 2


za [ρρ]
 r[ρρ_] := nora [ρρ]−n
ρ]2 +za [ρρ]2 )
n (ra [ρ
 
 2  ρ 2
] +(n2 −no2 )za [ρρ]
2
 2  2√ n ra [ρ
ra [ρρ] +za [ρρ]
n2 (ra [ρρ] +za [ρρ] )
2 2

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[Limit[(ra[ρρ])/(−zo+za[ρρ]),zo→
→−Infinity]];

θ 2[ρρ_] := ArcTan[Limit[(ra[ρρ])/(zi−za[ρρ]),zi→
→Infinity]];

c1[ρρ_] :=za[ρρ];
(** Setting variable’s **)

c = 1;
v = c/n;
no = 1;
n = 1.5;

ρ max = 25;
density = 5;

11-37
 Stigmatic Optics

(* zo −> −Infinity *)
Zo = −50;
(* zi −> Infinity*)
Zi = 60;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle →{Black,Red}, AspectRatio→ →Automatic, AxesLabel →{r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−500,ρρmax+500},

PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV:=Table[{
ParametricPlot[{
Piecewise[{
{−tt,tt<=c1[ρρ]},
{v (tt−c1[ρρ])−za[ρρ],
c1[ρρ]<tt}}],
Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{ra[ρρ],c1[ρρ]<tt}}]},
{ρρ,−ρρmax,ρρmax},

PlotRange→→All,
PlotStyle → {Blue, Thickness[0.002]}]},{tt,0,−Zo+( n Zi),density}];

RaysV:=Table[{
ParametricPlot[{
Piecewise[{
{−tt,tt<=c1[ρρ]},
{v (tt−c1[ρρ])−za[ρρ],c1[ρρ]<tt}}],
Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{ra[ρρ],c1[ρρ]<tt}}] },
{tt,0,−Zo+( n Zi)},PlotRange→ →All,

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-38
 Stigmatic Optics

11.3 Programs for chapter 8

11.3.1 Case 1: real finite object—real finite image

In[17]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]


f[ρρ_] := −no zo + n τ + ni zi + Sign[zo] ra[ρρ]2 +(zo−za[ρρ])2

z [ρρ_] := za[ρρ] − τ − zi

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

1
ϑ [ρρ_] := 2 f[ρρ] n + ni2 (ra[ρρ]  r[ρρ] + z [ρρ]  z[ρρ]) + Sign[n]Sign[zi]
n −1

11-39
 Stigmatic Optics


√
ni2 (f[ρρ]2 + n2 (ra[ρρ]2 + z [ρρ]2 ) − ni2 (zz [ρρ]  r[ρρ] − ra[ρρ]  z[ρρ])2

+ 2 f[ρρ] n (ra[ρρ] r[ρρ] + z [ρρ] z[ρρ]))

zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ] )];

θ 3[ρρ_] := ArcTan[rb[ρρ]/(zb[ρρ]−ττ −zi)];

c1[ρρ_] := Sec[θθ 1[ρρ]](−zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ ρ ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 16;
density = 4;
τ = 12;
zo = −40;
zi = 40;
zin = n zi;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

11-40
 Stigmatic Optics

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−500,ρρmax+500},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo+nττ +zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0,−zo+ n τ + zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-41
 Stigmatic Optics

11.3.2 Case 2: real infinity object—real finite image

In[18]:=
(** Definition of set variables **)

no2
Ka = − ;
n2
n
coa = ;
n zi−no zi
c1a = 0;

b1a = 0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

f[ρρ_] := n τ + ni zi − za[ρρ]

z [ρρ_] := za[ρρ] − τ − zi

no za [ρρ]
2
ra [ρρ]
 z[ρρ_] := + 
n (ra [ρρ] + za [ρρ] )
2 2
ra [ρρ] + za [ρρ]
2 2
 
√ n2 ra [ρρ]2 + (n2 −no2 ) za [ρρ]2
n2 (ra [ρρ] +za [ρρ] )
2 2


za [ρρ]
 r[ρρ_] := no ra [ρρ]−n
ρ]2 +za [ρρ]2 )
n (ra [ρ
 
 2  ρ 2
] +(n2 −no2 )za [ρρ]
2
 2  2√ n ra [ρ
ra [ρρ] +za [ρρ]
n2 (ra [ρρ] +za [ρρ] )
2 2


1
ϑ [ρρ_] := n f[ρρ]+ni2 (ra[ρρ]  r[ρρ]+zz [ρρ]  z[ρρ])+Sign[n]Sign[zi]
−1+n2

√
ni2 (f[ρρ]2 +n2 (ra[ρρ]2 +zz [ρρ]2 )−ni2 (zz [ρρ]  r[ρρ]−ra[ρρ]  z[ρρ])2

+ 2 n f[ρρ] (ra[ρρ]  r[ρρ]+zz [ρρ]  z[ρρ]))

11-42
 Stigmatic Optics

zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[Limit[(ra[ρρ])/(−zo+za[ρρ]),zo→
→−Infinity]]

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ])];

θ 3[ρρ_] := ArcTan[rb[ρρ]/(zb[ρρ]−ττ −zi)];

c1[ρρ_] := (−Zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ ρ ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 16;
density = 4;
τ = 12;
(* zo −> −Infinity *)
Zo = −40;
zi = 40;
zin = n zi;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−500,ρρmax+500},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

11-43
 Stigmatic Optics

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+nττ +zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0,−Zo+ n τ + zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-44
 Stigmatic Optics

11.3.3 Case 3: real infinity object—virtual finite image

In[19]:=
(** Definition of set variables **)

no2
Ka = − ;
n2
n
coa = ;
n zi−no zi
c1a = 0;

b1a = 0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

f[ρρ_] := n τ + ni zi − za[ρρ]

z [ρρ_] := za[ρρ] − τ − zi

no za [ρρ]
2
ra [ρρ]
 z[ρρ_] := +
n (ra [ρρ] + za [ρρ] )
2 2
ra [ρρ] + za [ρρ]
2 2
 
√ n2 ra [ρρ]2 + (n2 −no2 ) za [ρρ]2
n2 (ra [ρρ] +za [ρρ] )
2 2


za [ρρ]
 r[ρρ_] := no ra [ρρ]−n
ρ]2 +za [ρρ]2 )
n (ra [ρ
 
 2  ρ 2
] +(n2 −no2 )za [ρρ]
2
2√ n ra [ρ
ra [ρρ] +za [ρρ]
2

n2 (ra [ρρ] +za [ρρ] )

2 2


1
ϑ [ρρ_] := n f[ρρ]+ni2 (ra[ρρ]  r[ρρ]+zz [ρρ]  z[ρρ]) − Sign[n]Sign[zi]
−1+n2

√
ni2 (f[ρρ]2 +n2 (ra[ρρ]2 +zz [ρρ]2 )−ni2 (zz [ρρ]  r[ρρ]−ra[ρρ]  z[ρρ])2

+ 2 n f[ρρ] (ra[ρρ]  r[ρρ]+zz [ρρ]  z[ρρ]))

11-45
 Stigmatic Optics

zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[Limit[(ra[ρρ])/(−zo+za[ρρ]),zo→
→−Infinity]]

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ])];

θ 3[ρρ_] := ArcTan[rb[ρρ]/(zb[ρρ]−ττ −zi)];

c1[ρρ_] := (−Zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ ρ ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 15;
density = 5;
τ = 20;
(* zo −> −Infinity *)
Zo = −25;
zi = −20;
zin = n zi;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−500,ρρmax+500},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

11-46
 Stigmatic Optics

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt+ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+nττ −zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+Zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt+ra[ρρ],tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0,−Zo+ n τ − zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-47
 Stigmatic Optics

11.3.4 Case 4: real finite object—virtual finite image

In[20]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ 2 + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

f[ρρ_] := −no zo + n τ + ni zi + Sign[zo] ra[ρρ]2 +(zo−za[ρρ])2

z [ρρ_] := za[ρρ] − τ − zi

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 
ra’[ρρ] √ no (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])2
+  1− 2
2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] √ no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])2
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

1
ϑ [ρρ_] := 2 f[ρρ] n + ni2 (ra[ρρ]  r[ρρ] + z [ρρ]  z[ρρ]) − Sign[n]Sign[zi]
n −1

√
ni2 (f[ρρ]2 + n2 (ra[ρρ]2 + z [ρρ]2 ) − ni2 (zz [ρρ]  r[ρρ] − ra[ρρ]  z[ρρ])2

11-48
 Stigmatic Optics


+ 2 f[ρρ] n (ra[ρρ]  r[ρρ] + z [ρρ]  z[ρρ]))

zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ] )];

θ 3[ρρ_] := ArcTan[rb[ρρ]/(zb[ρρ]−ττ −zi)];

c1[ρρ_] := Sec[θθ 1[ρρ]](−zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ ρ ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 5;
density = 1;
τ = 10;
zo = −15;
zi = −5;
zin = n zi;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

11-49
 Stigmatic Optics

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−500,ρρmax+500},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo+nττ −zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0,−zo+ n τ − zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-50
 Stigmatic Optics

11.3.5 Case 5: real finite object—real infinite image

In[21]:= ClearAll["Global‘*"]
(** Definition of set variables **)

n2
Ka = − ;
no2
no
coa = − ;
n zo−no zo
coa = 0;

c1a = 0;

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
ρ2
1+b1a ρ + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

ϑ [ρρ_] := (−1+n)ττ −zo+za[ρρ] − Sign[n]Sign[Zi]Sqrt[ra[ρρ]^2 + (za[ρρ]−zo)^2] /

(n−z[ρρ]);
zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ[ρρ] r[ρρ];

11-51
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ] )];

θ 3[ρρ_] := ArcTan[Limit[rb[ρρ]/(zb[ρρ]−ττ −zi),zi→

→Infinity]];

c1[ρρ_] := Sec[θθ 1[ρρ]](−zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ρρ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 45;
density = 5;
τ = 30;
zo = −55;
(* zi −> Infinity*)
Zi = 35;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−50,ρρmax+50},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-52
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo + n τ + Zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]]tt+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},

{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]]tt,tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0,−zo + n τ + Zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-53
 Stigmatic Optics

11.3.6 Case 6: virtual finite object—real infinite image

In[22]:= ClearAll["Global‘*"]
(** Definition of set variables **)

n2
Ka = − ;
no2
no
coa = − ;
n zo−no zo
coa = 0;

c1a = 0;

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ2 + 1+(2 b1a−coa2 Ka) ρ2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

ϑ[ρρ_] := (−1+n)ττ −zo+za[ρρ] + Sign[n]Sign[zo]Sqrt[ra[ρρ]^2 + (za[ρρ]−zo)^2] /

( n−z[ρρ]);
zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

11-54
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ] )];

θ 3[ρρ_] := ArcTan[Limit[rb[ρρ]/(zb[ρρ]−ττ −zi),zi→

→Infinity]];

c1[ρρ_] := (n zo + n τ ) + Sec[θθ 1[ρρ]](−zo + za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ρρ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 16;
density = 2;
τ = 1;
zo = 16;
(* zi −> Infinity*)
Zi = 16;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−5,ρρmax+5},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-55
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(n zo+n τ ))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(n zo+n τ )),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0, zo + n τ + n Zi + 0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(n zo+n τ ))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(n zo+n τ )),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0, zo + n τ + n Zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-56
 Stigmatic Optics

11.3.7 Case 7: virtual finite object—real finite image

In[23]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ 2 + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]


f[ρρ_] := −no zo + n τ + ni zi + Sign[zo] ra[ρρ]2 +(zo−za[ρρ])2

z [ρρ_] := za[ρρ] − τ − zi

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

 r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n
 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

1
ϑ [ρρ_] := 2 f[ρρ] n + ni2 (ra[ρρ]  r[ρρ] + z [ρρ]  z[ρρ]) − Sign[n]Sign[zi]
n −1

11-57
 Stigmatic Optics


√
ni2 (f[ρρ]2 + n2 (ra[ρρ]2 + z [ρρ]2 ) − ni2 (zz [ρρ]  r[ρρ] − ra[ρρ]  z[ρρ])2

+ 2 f[ρρ] n (ra[ρρ] r[ρρ] + z [ρρ] z[ρρ]))

zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ] )];

θ 3[ρρ_] := ArcTan[rb[ρρ]/(zb[ρρ]−ττ −zi)];

c1[ρρ_] := (zo+zi) + Sec[θθ 1[ρρ]](−zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ ρ ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 15;
density = 3;
τ = 20;
zo = 40;
zi = 25;
zin = n zi;

(** Plot Section **)

11-58
 Stigmatic Optics

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−100,ρρmax+100},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],

Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
∞]]]}];
[Cases[SurfaceV,_Line,∞

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo+zi))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo+zi)),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,zo+n τ +zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo+zi))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo+zi)),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},

{tt,0,zo+ n τ +zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-59
 Stigmatic Optics

11.3.8 Case 8: virtual finite object—virtual finite image

In[24]:= ClearAll["Global‘*"]
(** Definition of set variables **)
2
(n2 zin−no2 zo)
Ka = ;
n no (n zin−no zo) (n zo−no zin)

n zo−no zin
coa = ;
zin zo(n−no)

(n−no)(n+no)2
c1a = ;
4 n no zin zo(n zin−no zo)

(n+no)(n2 zin−no2 zo)

b1a = ;
2 n no zin zo(n zin−no zo)

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ 2 + 1+(2 b1a−coa2 Ka) ρ 2

 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]


f[ρρ_] := −no zo + n τ + ni zi + Sign[zo] ra[ρρ]2 +(zo−za[ρρ])2

z [ρρ_] := za[ρρ] − τ − zi

no(za’ [ρρ] (ra[ρρ]ra’ [ρρ]+(−zo+za[ρρ]) za’ [ρρ]))

 z[ρρ_] := 
n ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2

 2

ra’ [ρρ] √ no2 (ra[ρρ]ra ’[ρρ]+(−zo+za[ρρ]) za ’[ρρ])
+  1− 2 2
2
(ra’ [ρρ]) +(za’ [ρρ])
2 n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra’ [ρρ]) +(za’ [ρρ]) )

no(ra [ρρ](ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ]))

r[ρρ_]:= 
ra[ρρ]2 +(zo−za[ρρ])2 ((ra [ρρ]) +(za [ρρ]) )
2 2
n

11-60
 Stigmatic Optics

 
za [ρρ] no2 (ra[ρρ]ra [ρρ]+(−zo+za[ρρ]) za [ρρ])
2
√
−  1−
n2 (ra[ρρ]2 +(zo−za[ρρ])2 ) ((ra [ρρ]) +(za [ρρ]) )
2 2
(ra [ρρ]) +(za [ρρ])
2 2

1
ϑ [ρρ_] := 2 f[ρρ] n + ni2 (ra[ρρ]  r[ρρ] + z [ρρ]  z[ρρ]) + Sign[n]Sign[zi]
n −1

√
ni2 (f[ρρ]2 + n2 (ra[ρρ]2 + z [ρρ]2 ) − ni2 (zz [ρρ]  r[ρρ] − ra[ρρ]  z[ρρ])2

+ 2 f[ρρ] n (ra[ρρ]  r[ρρ] + z [ρρ]  z[ρρ]))

zb[ρρ_] := za[ρρ] + ϑ [ρρ]  z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[(ra[ρρ])/(−zo+za[ρρ])];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ] )];

θ 3[ρρ_] := ArcTan[rb[ρρ]/(zb[ρρ]−ττ −zi)];

c1[ρρ_] := (zo + nττ ) + Sec[θθ 1[ρρ]](−zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ ρ ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 30;
density = 3;
τ = 30;
zo = 50;
zi = −60;
zin = n zi;

11-61
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−10,ρρmax+10},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];
WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo + nττ ))+zo,tt<=c1[ρρ]},

{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + nττ )),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,zo+n τ − n zi+0.25,density}];
RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ρρ]](tt−(zo + nττ ))+zo,tt<=c1[ρρ]},
{v Cos[θθ 2[ρρ]](tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ])Cos[θθ 3[ρρ]],tt>=c2[ρρ]}}],

Piecewise[{
{Sin[θθ 1[ρρ]](tt−(zo + nττ )),tt<=c1[ρρ]},
{v Sin[θθ 2[ρρ]](tt−c1[ρρ])+ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ]+(tt−c2[ρρ])Sin[θθ 3[ρρ]],tt>=c2[ρρ]}}]},
{tt,0,zo+ n τ − n zi+0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→ →{640,Automatic}]
→All,ImageSize→

11-62
 Stigmatic Optics

11.3.9 Case 9: real infinite object—real infinite image

In[25]:=
(** Definition of set variables **)

no2
Ka = − ;
n2
coa = 0;

c1a = 0;

b1a = 0;

ρ 2 (coa+c1a ρ 2 )
za[ρρ_] := 
1+b1a ρ 2 + 1+(2 b1a−coa2 Ka) ρ 2
 
ra[ρρ_] := Piecewise[{{ ρ 2 −za[ρρ]2 ,ρρ≥ 0},{− ρ 2 −za[ρρ]2 ,ρρ<0}}]

no za [ρρ]
2
ra [ρρ]
 z[ρρ_] := +
 2  2
n (ra [ρρ] + za [ρρ] ) 2
ra [ρρ] + za [ρρ]
2
 
√ n2 ra [ρρ]2 + (n2 −no2 ) za [ρρ]2
n2 (ra [ρρ] +za [ρρ] )
2 2


za [ρρ]
 r[ρρ_] := nora [ρρ]−n
ρ]2 +za [ρρ]2 )
n (ra [ρ
 
 2  ρ 2
] +(n2 −no2 )za [ρρ]
2
 2  2√ n ra [ρ
ρ ρ
ra [ρ ] +za [ρ ]
n2 (ra [ρρ] +za [ρρ] )
2 2

 
ϑ [ρρ_] := (−1+n)ττ −zo+za[ρρ]+Sign[n]Sign[zo]Sqrt[ra[ρρ]^2+(za[ρρ]−zo)^2] /(n

−z[ρρ]);
zb[ρρ_] := za[ρρ] + ϑ[ρρ] z[ρρ];

rb[ρρ_] := ra[ρρ] + ϑ [ρρ]  r[ρρ];

(** Homotopy for Eikonal **)

θ 1[ρρ_] := ArcTan[Limit[(ra[ρρ])/(−zo+za[ρρ]),zo→
→−Infinity]];

θ 2[ρρ_] := ArcTan[(rb[ρρ]−ra[ρρ])/(zb[ρρ]−za[ρρ])];

θ 3[ρρ_] := ArcTan[Limit[rb[ρρ]/(zb[ρρ]−ττ −zi),zi→

→Infinity]];

11-63
 Stigmatic Optics

c1[ρρ_] := Sec[θθ 1[ρρ]](−zo+za[ρρ]);

c2[ρρ_] := (Sec[θθ 2[ρρ]](zb[ρρ]−za[ρρ])/v)+c1[ρρ];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

ρ max = 25;
density = 5;
τ = 25;
(*zo −> −Infinity*)
Zo = 50;
(* zi −> Infinity*)
Zi = 60;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},{ρρ,−ρρmax−1,ρρmax+1},
PlotStyle → {Black,Red}, AspectRatio→ →Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ρρ],ra[ρρ]},{zb[ρρ],rb[ρρ]}},
{ρρ,−ρρmax−5,ρρmax+5},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-64
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{−tt,tt<=c1[ρρ]},
{v (tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ]),tt>=c2[ρρ]}}],

Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{v ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ],tt>=c2[ρρ]}}]},

{ρρ,−ρρmax,ρρmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+nττ + Zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{−tt,tt<=c1[ρρ]},
{v (tt−c1[ρρ])+za[ρρ],c1[ρρ]<tt<c2[ρρ]},
{zb[ρρ]+(tt−c2[ρρ]),tt>=c2[ρρ]}}],

Piecewise[{
{ra[ρρ],tt<=c1[ρρ]},
{v ra[ρρ],c1[ρρ]<tt<c2[ρρ]},
{rb[ρρ],tt>=c2[ρρ]}}]},

{tt,0,−Zo+ n τ + zi+0.25},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→

{ρρ,−ρρmax,ρρmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-65
 Stigmatic Optics

11.4 Programs for chapter 9

11.4.1 Case 1: real finite object—real finite image

In[26]:= ClearAll["Global‘*"]
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])


f[ra_] := −no zo + n τ + ni zi + Sign[zo] ra2 +(zo−za[ra])2

z [ra_] := za[ra] − τ − zi

no(za’ [ra] (ra+(−zo+za[ra]) za’ [ra]))

 z[ra_] := 
n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2

 2

1 √ no2 (ra+(−zo+za[ra]) za’ [ra])
+  1− 2
1+(za’ [ra])
2 n2 (ra2 +(zo−za[ra])2 ) (1+(za’ [ra]) )

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n
 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

1
ϑ [ra_] := 2 f[ra] n + ni2 (ra  r[ra] + z [ra]  z[ra]) + Sign[n]Sign[zi]
n −1


√
ni2 (f[ra]2 + n2 (ra2 + z [ra]2 ) − ni2 (zz [ra]  r[ra] − ra  z[ra])2

+ 2 f[ra] n (ra  r[ra] + z [ra]  z[ra]))

zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

11-66
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[rb[ra]/(zb[ra]−ττ −zi)];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 20;
density = 4;
τ = 29;
zo = −55;
zi = 30;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-67
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo+nττ +zi,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,−zo+ n τ + zi},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-68
 Stigmatic Optics

11.4.2 Case 2: real infinity object—real finite image

In[27]:=
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])

f[ra_] := n τ + ni zi − za[ra]

z [ra_] := za[ra] − τ − zi



1  n2 +(−1+n2 ) za [ra]2
 z[ra_] := 
n2 (1+za [ra] )
2 3/2 2

n(1+za [ra] )

  

2
za [ra]  n2 +(−1+n2 ) za [ra]2
+
2
1+za [ra] +n 
n2 (1+za [ra] )
2 3/ 2 2

n (1+za [ra] )


za [ra]
1+za [ra]
2
 r[ra_]:=
 2 3/2
n (1+za [ra] )
  
 
 n2 +(−1+n2 ) za [ra]2  n2 +(−1+n2 ) za [ra]2
−n  −n za [ra] 
2
;
 2
n2 (1+za [ra] )
2 2
n (1+za [ra] )


1
ϑ[ra_] := n f[ra]+ni2 (ra r[ra]+zz [ra] z[ra])+Sign[n]Sign[zi]
−1+n2

√
ni2 (f[ra]2 +n2 (ra2 +zz [ra]2 )−ni2 (zz [ra]  r[ra]−ra  z[ra])2

+ 2 n f[ra] (ra  r[ra]+zz [ra]  z[ra]))

zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

(** Homotopy for Eikonal **)

→−Infinity]]
θ 1[ra_] := ArcTan[Limit[(ra)/(−zo+za[ra]),zo→

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

11-69
 Stigmatic Optics

θ 3[ra_] := ArcTan[rb[ra]/(zb[ra]−ττ −zi)];

c1[ra_] := (−Zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 26;
density = 4;
τ = 29;
(* zo −> −Infinity *)
Zo = −55;
zi = 30;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

11-70
 Stigmatic Optics

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+nττ +zi,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{ra,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,−Zo+ n τ + zi},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→

{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-71
 Stigmatic Optics

11.4.3 Case 3: real infinity object—virtual finite image

In[28]:=
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])

f[ra_] := n τ + ni zi − za[ra]

z [ra_] := za[ra] − τ − zi



1  n2 +(−1+n2 ) za [ra]2
 z[ra_] := 
n2 (1+za [ra] )
2 3/2 2

n(1+za [ra] )

  

za [ra]
2  n2 +(−1+n2 ) za [ra]2
1+za [ra] +n 
 2
+
n2 (1+za [ra] )
2 3/ 2 2
n (1+za [ra] )


za [ra]
1+za [ra]
2
 r[ra_]:=
 2 3/2
n (1+za [ra] )
  
 
 n2 +(−1+n2 ) za [ra]2  n2 +(−1+n2 ) za [ra]2
−n  −n za [ra] 
2
;
 2
n2 (1+za [ra] )
2 2
n (1+za [ra] )


1
ϑ [ra_] := n f[ra]+ni2 (ra  r[ra]+zz [ra]  z[ra]) − Sign[n]Sign[zi]
−1+n2

√
ni2 (f[ra]2 +n2 (ra2 +zz [ra]2 )−ni2 (zz [ra]  r[ra]−ra  z[ra])2

+ 2 n f[ra] (ra  r[ra]+zz [ra]  z[ra]))

zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

11-72
 Stigmatic Optics

(** Homotopy for Eikonal **)

→−Infinity]]
θ 1[ra_] := ArcTan[Limit[(ra)/(−zo+za[ra]),zo→

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[rb[ra]/(zb[ra]−ττ −zi)];

c1[ra_] := (−Zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 28;
density = 4;
τ = 29;
(* zo −> −Infinity *)
Zo = −55;
zi = −55;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},

11-73
 Stigmatic Optics

{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt+ra[ra],tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+nττ −zi,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+Zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt+ra[ra],tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,−Zo+ n τ − zi},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-74
 Stigmatic Optics

11.4.4 Case 4: real finite object—virtual finite image

In[29]:= ClearAll["Global‘*"]
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])


f[ra_] := −no zo + n τ + ni zi + Sign[zo] ra2 +(zo−za[ra])2

z [ra_] := za[ra] − τ − zi

no(za’ [ra] (ra+(−zo+za[ra]) za’ [ra]))

 z[ra_] := 
n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2

 2

1 √ no2 (ra+(−zo+za[ra]) za’ [ra])
+  1− 2
1+(za’ [ra])
2 n2 (ra2 +(zo−za[ra])2 ) (1+(za’ [ra]) )

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n
 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2


1
ϑ [ra_] := 2 f[ra] n + ni2 (ra  r[ra] + z [ra]  z[ra]) + Sign[n]Sign[zi]
n −1

√
ni2 (f[ra]2 + n2 (ra2 + z [ra]2 ) − ni2 (zz [ra]  r[ra] − ra  z[ra])2

+ 2 f[ra] n (ra  r[ra] + z [ra]  z[ra]))

zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

(** Homotopy for Eikonal **)

11-75
 Stigmatic Optics

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[rb[ra]/(zb[ra]−ττ −zi)];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;

ni = no;
n = 1.5;

rmax = 28;
density = 4;
τ = 29;
zo = −55;
zi = −55;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

11-76
 Stigmatic Optics

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−zo+nττ −zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,−zo+ n τ − zi+0.25},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-77
 Stigmatic Optics

11.4.5 Case 5: real finite object—real infinite image

In[30]:= ClearAll["Global‘*"]
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])

no(za’ [ra] (ra+(−zo+za[ra]) za’ [ra]))

 z[ra_] := 
n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2

 2

1 √ no2 (ra+(−zo+za[ra]) za’ [ra])
+  1− 2
1+(za’ [ra])
2 n2 (ra2 +(zo−za[ra])2 ) (1+(za’ [ra]) )

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n
 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

ϑ [ra_] := (−1+n)ττ −zo+za[ra] − Sign[n]Sign[Zi]Sqrt[ra^2 + (za[ra]−zo)^2] /

(n−z[ra]);
zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[Limit[rb[ra]/(zb[ra]−ττ −zi),zi→

→Infinity]];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

11-78
 Stigmatic Optics

no = 1;
ni = no;
n = 1.5;

rmax = 28;
density = 4;
τ = 29;
zo = −55;
(* zi −> Infinity*)
Zi = 30;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle →{Black,Red}, AspectRatio→ →Automatic, AxesLabel →{r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle →{Blue, Thickness[0.002]}]},
{tt,0,−zo + n τ + Zi + 0.25,density}];

11-79
 Stigmatic Optics

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]]tt+zo,tt<=c1[ra]},

{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]]tt,tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,−zo + n τ + Zi + 0.25},
PlotRange→→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-80
 Stigmatic Optics

11.4.6 Case 6: virtual finite object—real infinite image

In[31]:= ClearAll["Global‘*"]
(** Definition of set variables **)

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n
 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2

ϑ [ra_] := (−1+n)ττ −zo+za[ra] + Sign[n]Sign[zo]Sqrt[ra^2 + (za[ra]−zo)^2] /

(n−z[ra]);
zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[Limit[rb[ra]/(zb[ra]−ττ −zi),zi→

→Infinity]];

c1[ra_] := (n zo + n τ ) + Sec[θθ 1[ra]](−zo + za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 28;
density = 4;
τ =291;
zo = 70;
(* zi −> Infinity*)
Zi = 30;

11-81
 Stigmatic Optics

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},
{ra,−rmax−500,rmax+500},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];
WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(n zo+n τ ))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(n zo+n τ )),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0, zo + n τ + n Zi ,density}];
RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(n zo+n τ ))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(n zo+n τ )),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0, zo + n τ + n Zi + 0.25},
PlotRange→ →All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-82
 Stigmatic Optics

11.4.7 Case 7: virtual finite object—real finite image

In[32]:= ClearAll["Global‘*"]
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])


f[ra_] := −no zo + n τ + ni zi + Sign[zo] ra2 +(zo−za[ra])2

z [ra_] := za[ra] − τ − zi

no(za’ [ra] (ra+(−zo+za[ra]) za’ [ra]))

 z[ra_] := 
n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2

 2

1 √ no2 (ra+(−zo+za[ra]) za’ [ra])
+  1− 2
1+(za’ [ra])
2 n2 (ra2 +(zo−za[ra])2 ) (1+(za’ [ra]) )

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n

 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2


1
ϑ [ra_] := 2 f[ra] n + ni2 (ra  r[ra] + z [ra]  z[ra]) − Sign[n]Sign[zi]
n −1

√
ni2 (f[ra]2 + n2 (ra2 + z [ra]2 ) − ni2 (zz [ra]  r[ra] − ra  z[ra])2

+ 2 f[ra] n (ra  r[ra] + z [ra]  z[ra]))

zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

11-83
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[rb[ra]/(zb[ra]−ττ −zi)];

c1[ra_] := (zo+zi) + Sec[θθ 1[ra]](−zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 24;
density = 4;
τ = 29;
zo = 60;
zi = 50;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},

→Automatic, AxesLabel → {r,z}]

PlotStyle → {Black,Red}, AspectRatio→

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-84
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(zo+zi))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(zo+zi)),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,zo+n τ +zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(zo+zi))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(zo+zi)),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,zo+ n τ +zi+0.25},
PlotRange→→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-85
 Stigmatic Optics

11.4.8 Case 8: virtual finite object—virtual finite image

In[33]:= ClearAll["Global‘*"]
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])


f[ra_] := −no zo + n τ + ni zi + Sign[zo] ra2 +(zo−za[ra])2

z [ra_] := za[ra] − τ − zi

no(za’ [ra] (ra+(−zo+za[ra]) za’ [ra]))

 z[ra_] := 
n ra2 +(zo−za[ra])2 (1+(za [ra]) )
2

 2

1 √ no2 (ra+(−zo+za[ra]) za’ [ra])
+  1− 2
1+(za’ [ra])
2 n2 (ra2 +(zo−za[ra])2 ) (1+(za’ [ra]) )

no(ra+(−zo+za[ra]) za [ra])

 r[ra_]:= 
ra2 +(zo−za[ra])2 (1+(za [ra]) )
2
n
 
za [ra] no2 (ra+(−zo+za[ra]) za [ra]) )
2
√
−  1−
n2 (ra2 +(zo−za[ra])2 ) (1+(za [ra]) )
2
1+(za [ra])
2


1
ϑ [ra_] := 2 f[ra] n + ni2 (ra  r[ra] + z [ra]  z[ra]) + Sign[n]Sign[zi]
n −1

√
ni2 (f[ra]2 + n2 (ra2 + z [ra]2 ) − ni2 (zz [ra]  r[ra] − ra  z[ra])2

+ 2 f[ra] n (ra  r[ra] + z [ra]  z[ra]))

zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

11-86
 Stigmatic Optics

(** Homotopy for Eikonal **)

θ 1[ra_] := ArcTan[(ra)/(−zo+za[ra])];

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra] )];

θ 3[ra_] := ArcTan[rb[ra]/(zb[ra]−ττ −zi)];

c1[ra_] := (zo + nττ ) + Sec[θθ 1[ra]](−zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 28;
density = 4;
τ = 29;
zo = 50;
zi = −25;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

OvalSurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},
{ra,−rmax−500,rmax+500},
PlotStyle → {Directive[{Black,Dashed}],Directive[{Red,Dashed}]}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-87
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(zo + nττ ))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(zo + nττ )),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},
{ra,−rmax,rmax},
PlotRange→ →All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,zo+n τ − n zi+0.25,density}];

RaysV := Table[{

ParametricPlot[{
Piecewise[{
{Cos[θθ 1[ra]](tt−(zo + nττ ))+zo,tt<=c1[ra]},
{v Cos[θθ 2[ra]](tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra])Cos[θθ 3[ra]],tt>=c2[ra]}}],

Piecewise[{
{Sin[θθ 1[ra]](tt−(zo + nττ )),tt<=c1[ra]},
{v Sin[θθ 2[ra]](tt−c1[ra])+ra,c1[ra]<tt<c2[ra]},
{rb[ra]+(tt−c2[ra])Sin[θθ 3[ra]],tt>=c2[ra]}}]},

{tt,0,zo+ n τ − n zi+0.25},
PlotRange→→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
{ra,−rmax,rmax,density}];

Show[OvalSurfaceV,SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-88
 Stigmatic Optics

11.4.9 Case 9: real infinite object—real infinite image

In[34]:=
(** Definition of set variables **)

za[ra_] := −(29−Sqrt[29^2−ra^2])



1  n2 +(−1+n2 ) za [ra]2
 z[ra_] := 
n2 (1+za [ra] )
2 3/2 2
n(1+za [ra] )

  

2
za [ra]  n2 +(−1+n2 ) za [ra]2
+  2
1+za [ra] +n 
n2 (1+za [ra] )
2 3/ 2 2
n (1+za [ra] )


za [ra]
1+za [ra]
2
 r[ra_]:=
2 3/2
n (1+za [ra] )
  
 
 n2 +(−1+n2 ) za [ra]2  n2 +(−1+n2 ) za [ra]2
−n  −n za [ra] 
2
;
n2 (1+za [ra] )
2
n2 (1+za [ra] )
2

 
ϑ [ra_] := (−1+n)ττ −zo+za[ra]+Sign[n]Sign[zo]Sqrt[ra^2+(za[ra]−zo)^2] /(n

−z[ra]);
zb[ra_] := za[ra] + ϑ [ra]  z[ra];

rb[ra_] := ra + ϑ [ra]  r[ra];

(** Homotopy for Eikonal **)

11-89
 Stigmatic Optics

→−Infinity]];
θ 1[ra_] := ArcTan[Limit[(ra)/(−zo+za[ra]),zo→

θ 2[ra_] := ArcTan[(rb[ra]−ra)/(zb[ra]−za[ra])];

θ 3[ra_] := ArcTan[Limit[rb[ra]/(zb[ra]−ττ −zi),zi→

→Infinity]];

c1[ra_] := Sec[θθ 1[ra]](−zo+za[ra]);

c2[ra_] := (Sec[θθ 2[ra]](zb[ra]−za[ra])/v)+c1[ra];

(** Setting variable’s **)

c = 1;
v = c/n;

no = 1;
ni = no;
n = 1.5;

rmax = 28;
density = 4;
τ = 29;
(*zo −> −Infinity*)
Zo = −55;
(* zi −> Infinity*)
Zi = 55;

(** Plot Section **)

SurfaceV := ParametricPlot[{{za[ra],ra},{zb[ra],rb[ra]}},{ra,−rmax−1,rmax+1},
PlotStyle → {Black,Red}, AspectRatio→→Automatic, AxesLabel → {r,z}]

LensV := Graphics[{EdgeForm[Directive[{Black,Thin}]],
Directive[Gray,Opacity[0.3]],
Polygon[Join[#[[1,1]],#[[2,1]],Reverse[#[[4,1]]],Reverse[#[[3,1]]]]&
[Cases[SurfaceV,_Line,∞∞]]]}];

11-90
 Stigmatic Optics

WavesV := Table[{
ParametricPlot[{
Piecewise[{
{−tt,tt<=c1[ra]},
{v (tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra]),tt>=c2[ra]}}],

Piecewise[{
{ra,tt<=c1[ra]},
{v ra,c1[ra]<tt<c2[ra]},
{rb[ra],tt>=c2[ra]}}]},

{ra,−rmax,rmax},
PlotRange→→All,
PlotStyle → {Blue, Thickness[0.002]}]},
{tt,0,−Zo+nττ + Zi+0.25,density}];

RaysV := Table[{
ParametricPlot[{
Piecewise[{
{−tt,tt<=c1[ra]},
{v (tt−c1[ra])+za[ra],c1[ra]<tt<c2[ra]},
{zb[ra]+(tt−c2[ra]),tt>=c2[ra]}}],

Piecewise[{
{ra,tt<=c1[ra]},
{v ra,c1[ra]<tt<c2[ra]},
{rb[ra],tt>=c2[ra]}}]},

{tt,0,−Zo+ n τ + zi+0.25},
→All,PlotStyle → {Lighter[Purple], Thickness[0.002]}]},
PlotRange→
{ra,−rmax,rmax,density}];

Show[SurfaceV,LensV,WavesV,RaysV,
PlotRange→→All,ImageSize→
→{640,Automatic}]

11-91