I’ve wondered this for a while and a quick Google hasn’t yielded great results.
When films, particularly full CGI films, are made, do they truly make it frame by frame? Or do they desgin the full 3d world and put cameras into the software to “film” the scene? And how extensive to they get?
For example, in the car chase ending for Toy Story, would they have designed and modeled the entire neighborhood and used “cameras” in the scene to film it? Would they fully 3d model each house and “play” the scene to film it? Or would they design and create small pieces and put them together like traditional cartoon?
I always imagined it was similar to the StarCraft map editor. I could build the map, add in units and then assign them AI cues to move through routes or take actions based on programmed cues. Very curious how it works on big budget films.
In: 331
There are a few different technologies that go into CGI. Usually, what you see on screen is a mix of them.
So you might have a program that specifically simulates destruction physics, or a program that specifically does water/cloth. You might also do simulation-based animations – or maybe you’ll do motion capture, and then use that data to drive an animation.
Sometimes the simulators allow you to apply textures on top of them. Sometimes they just export raw data, and you’ll have to start applying shaders, textures – basically everything that makes it pretty. Usually, when working on them, they’ll look pretty bad because making photorealistic images is VERY resource-intensive, and it’s faster to do work with simplified visuals and only apply the fancy pictures at the very end.
Then once everything’s done, you’ll throw it into a renderer that basically does all that. You’ll take the simple shapes, and start applying textures so that it looks real.
Latest Answers