17th April 2004, 3:58 PM
Well what they're doing now is creating incredibly detailed characters with 6-10 million polygons in a standard 3D modeling application, then using that to render highly detailed lightmap information. Then for the engine the character models are scaled down to around 10 thousand polygons but have the detailed lightmaps applied which gives the illusion of more polygons. This technique is nothing new, but a lighting system this insanely powerful is something unprecedented.
This process really makes sense for realtime 3D graphics. There are so many lighting calculations that need to be done with high poly models that this method really cuts down on those calculations while at the same time preserving a lot of the detail.
This process really makes sense for realtime 3D graphics. There are so many lighting calculations that need to be done with high poly models that this method really cuts down on those calculations while at the same time preserving a lot of the detail.