Powerful computer graphics hardware is now virtually ubiquitous. If only the corresponding software expertise was equally widespread. Lots of people know the basics. Surprisingly few people know more than that. If you’re in the first group, this book will move you into the second.
Tom McReynolds and David Blythe start with a conceptual overview of computer graphics -- solidifying the knowledge you already have, for the journey ahead. They also offer a detailed introduction to the OpenGL pipeline, focusing especially on four critical areas: lighting, texture mapping, rasterization, and depth buffering.
In Part II, they present several core techniques that can serve as building blocks for achieving practically anything with OpenGL. For instance, there’s a full chapter on multiple rendering passes: rendering a primitive multiple times using different modes, and combining the results. Once, this technique impacted performance too much to be widely useful. But that’s changing. And using it can powerfully improve image quality. Among the other core techniques covered here: anti-aliasing, compositing, blending, transparency, point operations, convolution, and basic transforms.
Building on this, McReynolds and Blythe then present more than 200 pages of advanced techniques: CAD and modeling applications (for enhancing scene realism and natural detail), illustration for scientific visualization, and more. The book closes with a practical primer on performance optimization. The authors’ tips address graphics processing, frame time management, application bottlenecks, and -- crucially -- measurement.
Of course, since the authors chose OpenGL, you can use all this knowledge on practically any platform -- from Windows and Macintosh to Linux and Unix, from “big iron” compute clusters down to embedded devices running OpenGL ES. Bill Camarda, from the June 2005 Read Only