MatrixWorkz opened this issue on May 05, 2008 · 61 posts
stewer posted Thu, 08 May 2008 at 5:19 PM
There is one subtle problem that you get with gamma correction in shaders vs applying inverse gamma correction to the textures before: texture filtering will still happen in non-linear space, resulting in mip-maps that have the wrong intensities at higher levels. (for a detailed explanation, see http://www.4p8.com/eric.brasseur/gamma.html.)
For an extreme example try this image as a texture, once with PPro's gamma correction and once with gamma correction in the shader.
gamma_dalai_lama_gray.jpg