I’m making a video which will be in an exhibition in 3 days and need urgent help on the following:
I am painting thousands of pixels according to a generative algorithm. I was using the CPU and on every frame I was calling ofDrawRectangle(x,y,1,1).
At 15FPS I was getting a maximum of 150k pixels. This was not enough. So, after some investigation I decided to move to the GPU using instance drawing and a TBO (texture buffer object). Surprisingly everything seemed to b working quite well.
But then I found out that I was not getting a 1 to 1 correlation between what I draw and the screen pixels (like I was getting in the CPU version).
It feels like there are two things happening:
- Everything is smaller
- Everything is blurred (anti-alias?)
I just want every vortex to be drawn as 1 screen pixel.
Here are the results (the upload did anti-alias on both but the original CPU version is crisp):
Can someone please suggest how this can be solved?
If it helps, the draw() code for both versions is here:
CPU version: https://github.com/nununo/ofMoldApp/blob/cpu-one2one/src/hypha/Hypha.cpp
GPU version: https://github.com/nununo/ofMoldApp/blob/gpu-one2one/src/app/Painter.cpp
I was hoping some simple GL configuration commands could solve this easily.