Introduction to “CG Magic” by Noriko Kurachi
The first computer graphics textbook I owned was “Computer Graphics: Principles and Practice” by James Foley, Andries van Dam, Steve Feiner, and John Hughes ? I had bought the book in 1990 for a computer graphics class that I was taking at the University of Michigan. I was interested to read the equations for perspective projection, and the most efficient way to draw a circle, but what I truly found amazing was how some of the CG images looked almost real ? almost like real photographs. Having a bit of experience in photography, I was intrigued that somehow math and programming could create images that seemed to have been produced by real places, real cameras, and real light ? and I was fascinated by the visual possibilities that such technologies could offer.
From reading the Foley and van Dam book I learned how polygons could form surfaces, how splines could form contours, how reflection models could mimic metal and plastic, and how light sources could create shading, highlights, and shadows. And I learned the beginnings of some important techniques that were then only a few years old ? advanced lighting simulation techniques such as radiosity, ray tracing, and global illumination. It seemed like these tools must hold the answers to how to create any form of photoreal imagery.
What I learned in the next few years is that these tools did contain answers, but not all of them. As I endeavored to use computer graphics to create “photoreal” renderings of things like my car, my high school, and my college campus, I realized that only an enormous amount of effort could create the 3D models, texture maps, and reflectance property specifications necessary to create the photoreal renderings I hoped to see. The solution that I and others discovered was to flip the traditional CG creation process around ? to begin with real-world photographs, and from these photographs to derive whatever geometry, reflectance, and lighting is necessary to create the CG models and renderings. By starting with the rich information available in photographs, these techniques ? known as image-based modeling, rendering, and lighting ? made it possible to produce photoreal computer graphics quickly and efficiently.
Interestingly, the photoreal renderings made possible by image-based techniques helped challenge the capabilities of the traditional computer graphics pipeline. Image-based techniques allowed us to move virtual cameras through real-world scenes, but they did not allow for control over the scene’s structure and lighting the way that traditional CG usually produced. What we discovered is that in order to regain this control, we needed to use traditional techniques such as global illumination and reflectometry to understand and unravel the lighting and reflectance seen in the photographs to adapt image-based models according to a new creative vision. These sorts of ideas allowed me to place towering monoliths into St. Peter’s Basilica and to light people’s faces by environments they had never been in. We also have seen the realism of image-based computer graphics push forward a new wave of traditional computer graphics methods ? efficient techniques for lighting CG objects with lighting environments, and for simulating light scattering through translucent surfaces.
The book you hold today tells the story of this new era of computer graphics. Working closely with researchers who helped lead this revolution, Noriko Kurachi describes these key innovations and brings them together as a coherent body of knowledge. Please read this book, practice the techniques, and figure out if they will allow you to create the visions you have in your mind. If they do, then make your visions a reality ? put them on the web, submit them to SIGGRAPH, and let the world see what can imagine. If they don’t, do it anyway ? and create the techniques you need that don’t yet exist. Your work may be what is written about in the next book, some ten or fifteen years from now.
