GL_RGB32F_ARB and GL_LINEAR

Started by
3 comments, last by _the_phantom_ 18 years, 10 months ago
Hi! Is there a way to have floating point textures and bilinear filtering work simultaneously? As soon as I change the 'internalformat' parameter of glTexImage2D from GL_RGB to GL_RGB32F_ARB, bilinear filtering is disabled. My first idea was to do the filtering myself in the shader that uses this texture, but that's not trivial as I'm using a textureCube lookup (basically a HDRI-texture-cube). Any ideas? ;) Regards, Leonhard
Advertisement
on the majority of current and older hardware, no, they simply cant do it. I do belive the GF6 series however can bilinear filter floating point textures (as to if it can do 64 and 128bit is another matter, i'm certain it can do 64bit however)
Interesting to see that GF6 cards can handle this.
Thanks a lot, _the_phantom_!
GF6 is the only card that can perform filtering other than NEAREST on float textures of type F16.
No card can do such thing on a F32 target, for now that is ;)
yeah, i thought that was the state of play.. now if only NV and ATI would tell us about the next generation of their GPUs so we can see if owt is gonna change [grin]

This topic is closed to new replies.

Advertisement