My little test app needs to grap the framebuffer to texture every frame. I tested it by using glCopyTexSubImage2D() but speed seems to be too slow for "realtime" use. I'm using 512*512 texture for this purpose. I tested it with geforce256 too and the speed was at least 10x better than with my Club3D 8500LE (latest official drivers from the ATI website)
Even if glCopyTexSubImage2D() is most standard way to do "render to texture" method, it seems that HW acceleration support for it is bit lame (except with geforce series..)
Is there any hope to get more speed to lCopyTexSubImage2D() in future drivers?? Or is pbuffer only "right" way to do this? (although less common..)
Jogi
Even if glCopyTexSubImage2D() is most standard way to do "render to texture" method, it seems that HW acceleration support for it is bit lame (except with geforce series..)
Is there any hope to get more speed to lCopyTexSubImage2D() in future drivers?? Or is pbuffer only "right" way to do this? (although less common..)
Jogi