basicwiz opened this issue on Apr 07, 2009 · 88 posts
MikeJ posted Thu, 09 April 2009 at 5:37 PM
Quote -
And ... as for whether using 4000 is stupid and bad practice, it just means you could have had 96 more pixels of detail, but you don't. The renderer still expands the image to a power of 2 before doing mip-mapping.
Graphics cards render textures in real time only in powers of two, and if a texture is, say, 4K, the GPU has to internally resize it to 4096. I don't know if it does that every time the camera is moved, or only once as the texture is loaded, though. If the GPU is doing that ALL the time, there will probabaly be a considerable display performace hit, especially if you have a scene with a whole lot of off-size textures.
It probably doesn't make a big difference in Poser, but I've been told it's an absolute necessity for game textures and animations, so this is why people make textures in powers of two.
The former limit isn't the only reason people are doing it (probably most people don't even know about tthet) - I think they do it because they see everyone else doing it that way, started by DAZ. Also, there are plenty of 3000x3000 and 3500x3500 and oher sized textures out there, too, so that's not explained away by any former 4K limit.
And when GPU's are being used for actual rendering, along with the CPU, it will probably become more of a necessity to have "correct" image sizes. Well, not like Poser is ever going to get GPU rendering anyway....
At any rate, if nothing else, it would just lseem more professional to use industry standard texture sizing rules, no?
But aside from that, if the renderer is expanding an image size, isn't it getting stretched slightly? Why not just make it that size to begin with? I can't see any good reason why not.