Originally posted by: HexiumVII
So what are the differences really in CG Film vs games? Is it 10x the polygons and 10x the size in textures? I've been trying to look up more technical details on Pixars films but haven't been able to find many numbers. They seem to use similar raster rendering techniques like in games rather than ray tracing.
There are a lots of things different.
The polygon counts are way higher.
The texture size is higher.
Physics calculations can be much higher. Calculating 20 million particles to make up the goo in the spiderman3 movie takes a lot of cpu power.
Lighting - radiosity, ray tracing, subsurface scattering, all devour processors.
compositing - blending the animation data with green screen film or backgrounds.
bones - each character has a rig that makes up how the mesh will move. Games like UT3 may have many bones for the character. But movies, for example narnia, had over 700 bones just for the lions face.
Animation keys - each item that is doing anything in the scene has to have a key.At every frame each key has to be calculated. So if you have a scene like the rat in pixars film, you have to have keys for every single bone in the rat, thousands of them, for every background object, for every light, for every sound.
Pixar generally writes a renderer for each work they produce. They don't use off the shelf rendering software most of the time. They have a very talented team of programmers.
about the lion in narnia
Animation Rig controls:
1851 controls, 742 in just the face alone
98 facial shapes
53 body shapes
Fur and Hair:
3 versions of Aslan were built: normal, shorn, and golden
7 hair pelt types
5.2 million strands of hair
R&H technical animators used thousands of guide hairs to simulate the effect of physics on Aslan's mane. The guide hairs drove a high-density pelt rendered as moving fur.
Rendering:
13 hours a frame render
?1.6 million render hours = 200 years of computer rendering time
Lo res 21,884 polys