8: Beyond Realtime Graphics
- Page ID
The first seven chapters of this textbook have covered real-time computer graphics, that is, graphics systems in which an image can be generated in a fraction of a second. The typical case is a video game, where new frames can be rendered as many as sixty times per second. Very complex and realistic-looking scenes can be rendered in real time, using techniques covered in this book and the immense processing power of modern GPUs, plus some tricks and advanced algorithms. However, real-time graphics still can’t match the realism of the very highest quality computer graphics, such as what can be found in movies. In fact, the CGI (computer generated imagery) in today’s movies is sometimes indistinguishable from reality. Getting graphics of that quality can require hours of computing time to render a single frame. (According to the Wikipedia article about the 2013 animated movie Frozen, some complex scenes in that movie used 30 hours of computing to render each frame.)
This chapter is a very brief look at some techniques that can be used for very high quality graphics. The discussion will be in general terms. I won’t be giving sample code or detailed discussions of the mathematics behind the techniques, but I hope to provide at least a basic conceptual understanding.
The first thing that you should understand, though, is that most of what you have leaned so far still applies. Scenes are still composed using geometric primitives, transformations, materials, textures, and light sources (although perhaps using more advanced material properties and lighting than we have encountered so far). The graphic designers working on CGI for a movie can see real time previews of their work that are rendered using techniques that we have covered. The final scene that you see in the movie is just rendered using different, much more computation-intensive techniques.