These days obviously texture maps / pixel shaders/ bumpmapping are the mainstay for rendering surface textures on walls, peoples and other things. From my knowledge HL2 uses loop subdivision surfaces with bumpmapping in their engine, and that''s pretty advanced. But i haven''t heard the use of visualization stuff in game engines yet? The reason why I ask is because, I have an oppertunity for my senior thesis to do research with a C.S professor of mine who is currently working on visualization theories and algorithms. Basically the idea from what he described was to take something like a stereoscopic picture of a texture on a wall, and then from the 2 images, generate a fairly accurate high resolution 3d texture from the 2 images. I''m wondering whether that can be applied in a game engine, where you can use actual 3d textures (along with pixel shaders). I bet it''s gonna be a bit more expensive than implmeenting bumpmaps, but as hardware goes up imo it''s problably a more feasible idea. Thoughts on this subject?