I asked one of the coders here about the performance cost of that relief mapping method. He said it's horribly costly since it's ray-casting per-pixel to achieve the depth. Also you can't change the mesh in any way, so there's no deformation allowed (unless you animate the texture, which is then a memory hog). Oh well.…
I always thought RGB normal maps were more accurate than a greyscale heightmap (bump map) because each pixel actually stores a vector for the high-detail surface's angle, rather than just a height-difference. I don't know much about the tech though... I just imagined RGB normal maps to be a more accurate depiction of the…
Basically: Yeah, they're about the same thing. A normal map stores the slope/tilt of each pixel in the texture. A bump map stores the simulated height of each pixel in the map. The renderer has to then convert these heights into slopes when it renders. It does this by comparing the current pixel with all it's neighbors to…
Yeah, but greyscale for displacement will actually displace the geometry of a really dense mesh, so its not in the same category really. I think a good basic point to be made for the advantage of normal maps is that even if you could theoretically paint a greyscale image of, say, a nude figure to be converted to a normal…
Hi eric, I don't think what you have said is true. The length of a normal is irrelavant, normalizeing a normal map is probably only nessasary if you attempt to hand edit one. Normal data is extracted from a hightmap to be calcuated on the video card so a normal map cannot be less accurate than a hightmap. Changing the…
I think ( dont quote me on this ) that im right in saying that you can apply the bump map directly to the HIGH poly model and then render out the normal map and get a more detailed map than that of just the high poly on its own and screw the photoshop plugin!? cut out the middle man so to speak.... on a related topic iv…