Normal vs bump map

broken image

† Displacement mapping has less of an impact on render time and memory usage if you are using micropolygon rendering with no raytracing (for example, no raytraced shadows). Can be baked from a high-resolution model or generated from a normal pass. Requires specialized texture painting tools. And you can add some surface curvature on top of existing curvature, using Bump node. You cannot use normal maps directly, because you should convert them from local (tangent space) to world space using Normal map node. Can be baked by rendering a height pass in 3D applications. Normal map, the blueish textures, that represent curvature of surface, should be plugged in material via Normal map node. Larger impact with raytracing/PBR because renderer must generate displaced geometry.Įasy to manipulate in 2D apps, but can be hard to get desired result painting in 2D. Slower in raytracing/PBR because it adds polygons †. Small slowdown, requires multiple texture samples. Potentially large memory usage for lots of small polygons †. Uses more memory for 16/32 bits to avoid banding. This idea and the specific technique to compute these shading normals in a way that gives a plausible appearance of the actual displaced surface were developed by Blinn ( 1978 ). Memory used by the texture, ideally greyscale. For relatively small displacement functions, the visual effect of bump mapping can be quite convincing.

broken image