Customer Reviews

Realistic Ray Tracing

5 star:		(5)
4 star:		(2)
3 star:		(4)
2 star:		(2)
1 star:		(1)

 

 

The book "Realistic Ray Tracing" contains a description of all important ray tracing techniques and a guideline to the implementation of a ray tracing program. The book covers the basics like ray-object intersection, lighting, viewing and materials, but the major part of the text deals with advanced techniques monte carlo integration, antialiasing, soft shadows or path-tracing.

The book contains only some 150 pages and each technique is thus described in 3 to 10 pages. The language used is clear and the book is very readable. It is very easy to read the whole book or just to pick a specific chapter and get an idea of one topic. The main focus of this book is the implementation of a ray tracer. All techniques are described in a way that enables the reader to easily code them. All the math needed is provided and procedural pseudo code fragments are given in some chapters. Despite being quite a thin book, the selection of topics is very good and most of the important ray tracing techniques are covered.

There are some problems with this book though. First of all, this book does not describe the ray tracing algorithm very well. Readers completely unfamiliar with this method might have some difficulties understanding the overall picture. This is also true for the implementation part. Although a lot of techniques and basics are explained, the author does not cover the implementation of a ray tracing framework.

Some of the chapters are simply too brief. The mathematical background is covered but not explained. In the first chapter, the author introduces 4-dimensional homogeneous coordinate systems without explaining them. This could have been done in 1-2 pages and would have helped to better understand a lot of the transformations used throughout the book. And the chapter about triangle meshes only deals with different approaches to store a triangle mesh. No word about the triangulation process itself.

Overall, this is a lovely book that covers a lot of ray tracing techniques, but is is no introduction to this method.

I was initially excited about this book, as I have a somewhat silly notion that books produced by authors bold enough to condense a large subject into a compact little package are generally very tight, elegant works that are a joy to read. Wirth's Compiler Construction is an excellent example of this.
As I worked through this book, I found this to be anything but the case. Maybe this is my fault; after all I did say it was a somewhat silly notion.

I have found that the assumed knowledge varies widely as you work your way throughout the book, and not just in an easy-to-difficult progression from front to back. I have worked through approximately 3/4ths of this book, and have found it so riddled with errors that I often wonder if the author didn't just submit his first draft as camera-ready to the publisher. Every time I look at a pseudocode algorithm I check the errata page, and almost every time I find that there are errors. One algorithm was so incredibly wrong that rather than try and correct the code in my book with a pencil, I had to print the correct code, cut it out, and tape it over the existing one! After all of this you start to wonder how much you can trust what is being said, which is unfortunate.

The book does have some redeeming value, and if you keep in mind the large number of errors, you can actually learn quite a bit from it. It just annoys me to spend money on a "rough draft" book that could have benefited so much by a little more "proofing" by the author. I would also take with a grain of salt reviews here that are obviously written by people who read the back cover and the introduction with great zeal and formed their summary based on that. If you have not worked your way through a substantial portion of a book, you have no business writing a review of it.

I did like how the book was divided into a basic ray tracer, bells and whistles, and an advanced section. I got some nice results with just the first part. I also took some useful bits and pieces from the second part, and found the discussion about monte carlo methods and antialiasing interesting. I also did some soft shadow work, but supplemented it with a discussion from the Watt/Watt book (Advanced Animation and Rendering Techniques). This would most likely be a decent book for a dabbler new to the field (but having a decent mathematical background), and those taking a ray tracing class.

I enjoyed this small, elegant book on how to write a modern ray tracer. As someone who has written his share of rendering programs, it's always a pleasure for me to see someone capture the essence of the idea as well as Shirley has done here.

Ray tracing is particularly well-suited to being broken down conceptually into bite-sized pieces, and this book uses that concept for its organization. Each of the 18 chapters covers a specific aspect of writing a ray tracing program. The writing and illustractions are clear and easy to follow.

There is no source code in this book, but there are a number of pseudo-code listings. The heart of most of the chapters is captured in the math. The essence of ray-tracing is in the details, and the mathematics capture those details precisely.

Though Shirley has boiled down the math to its minimum, the casual reader with no mathematical background will have trouble understanding the advanced portions of this book. If the sight of an integral sign spooks you, then it's probably a good idea to start with a more basic text. The math isn't gratuitous - this is the real, practical nuts-and-bolts stuff that you need to write a mature, modern renderer.

It's also great fun. Ray tracers are the simplest rendering programs to write, and there's nothing like seeing your own code producing beautiful images. There are plenty of references in the bibliography if you want to go further.

If you're ready to roll up your sleeves and implement a state-of-the-art ray-tracer, and are comfortable with basic calculus, this book will serve you as an ideal roadmap and reference.

The book started off amazingly well. I was very excited to find out what next. The book did a good job of introducing the technique and giving hints on how to implement them. After the first several chapters, however, the implementation gets left behind and the reader is given a series of chapters on disjointed topics. The author who was weaving such a beautiful web toward the beginning stopped doing everything that make the book great. There was no longer any explanation about how one topic led to the next and implementation was no longer touched upon.
This book is an adequate introduction to the concepts but I found it sorely lacking when it came to help on implementation.

I had heard about how many errors this book originally had and assumed that these would be fixed for the Second Edition printing -- well, they weren't. There's a typo on nearly every other page, plus some serious formula errors.

Also, the book isn't very balanced. Some parts go into too much detail, and others not nearly enough.

I returned this book the day after I received it.

Riddled with errors. Towards the end of the book, the material becomes very thin. While Shirley does cover some advanced techniques, I found his explantions to be lacking. If you're going to buy this book, wait for the second edition. Glassner is a much better text on ray tracing.

I bought this book during a period it had actually good reviews. From all CG books I have bought this is the worst by far, as others explain in other reviews, the book is full of mistakes, and lacks of explanations. It's not the kind of book that will give you the insight on how algorithms work, but just present you the algorithms or formulas directly (that usually are just plain wrong). I tried on first edition to mail the author about some errors and got no replay at all (and no fix).

There is really no reason at all why someone would need this book; if you are new to ray tracing, you wont find any explanations on how it really works. If you know about raytracing, in theory it could work as a reference manual, but since its so full of mistakes its of no use even for that.

I suggest "Computer Graphics: Principles and Practice" or "Advanced Animation and Rendering Techniques" if you want to learn the basics of raytracing and actually understand it. And "Realistic Image Synthesis Using Photon Mapping" if you want to implement some global illumination through raytracing, or some more advanced effects like caustics.

I have to second the person who said that the people who recommended this book never actually sat down and tried to implement the material inside of it. The book is riddled with mistakes, and the errata online keeps growing by the minute.

Shirley (and Morley) are brilliant and pioneers in the field, but they write a book like any hardcore professor you've taken in college would have written it --- dry, coarse and assuming of much knowldege. A reviewer here mentioned "disjointed" and that actually is an understatement.

I am by far no means a graphics person, but being a college graduate with at least two classes in graphics under my belt, I expected to just shoot through this book. This was definitely not the case.

Beware of slim and short books, and this is one of them. I found the classic "An Introduction To Ray Tracing", Andrew S. Glassner (another genuis) to be much better. Also, I've heard great things about Matt Pharr/Greg Humphreys "Physically Based Rendering -- from Theory to Implementation". That one's currently in the mail shipping to my house.

No offense to Mr. Shirley -- however please let edition 3 of this book (assuming there will be one) cater to a wider audience, such as myself.

I wonder if the first edition was better...

Compare to the first edition, major improvement include libraries of C++ code but unfortunately, lack of explanations.

No offense to Mr.Shirley, but this book is riddled with errors, some of the materials provided at the end of each chapters were DIFFERENT from the code on the server. Not sure which one should be follow.

If you intend to write you own path-tracer from scratch, I will definitely not recommend this book. Reading this book requires knowledge on much other theories, such as calculus, probability and other... which were very briefly mentioned in the text.

If you intend to understand global illumination and radiometry, I would recommend "Advanced Global Illumination" by Philip Dutre, Philippe Bekaert and Kavita Bala. This is a much better book which explains things clearly.

I took a ray tracing course from the first author several years ago and though I loved the course I am not a fan of the first edition. The new edition is a very very good book however. The biggest improvement is that it has lots of C++ code, along with explanation of why it is written the way it is. Higly recommended.