I’ve worked with the to_texture method in my patch for some time and it
always worked well without taxing my cpu too much. I started cleaning up
some things all over the place and now to_texture has become a big problem.
I’ve been debugging for many hours now and it’s really driving me insane.
A texture of 640×480 is already increasing my cpu from 25 to 100 and
dropping my framerate to 6fps. If I just use to_texture, without actually
drawing the texture on an object, nothing happens. Also processing it with
slabs doesn’t really change my cpu. Its when I send "texture_xxx" to a grid
or videoplane and bang it when my computer goes nuts.
Is there anyone that has experience with a situation like this or can think
of a possible reason why this could happen? I feel like an ass because it
has always worked. I know this sounds vague, and I’d post the patch if I
could but I can’t. Just hoping someone has run into this problem before…
I finally found the little bugger after 2 days of hair pulling. By accident
I passed the jit.window dimensions to the jit.gl.videoplane. This screws up
your cpu real bad.
The standard dimension for the videoplane is 20×20, but since I just use it
as a flat plane setting @dim 1 1 gave me a small performance boost. About 2%
on my computer for just one plane, so I you use multiple it might be worth
changing this setting.
I’ve always wondered, is there an important difference between [
jit.gl.videoplane @interp 0] and texturing a [jit.gl.gridshape @shape