Maybe it’s a newbie question, but can’t find any infos, not clear how jit.gl.pix is placed in the dataflow…
It seems for me, all the examples of jit.gl.pix works with data from the CPU (camera data, sent every frame to the GPU) not texture data stored on the GPU. Can someone clear it for me what happens if i render an openGL conext on GPU including all objects, lights and shaders written by me (JXS). I know vertex and pixel shaders will work on GPU with the texture and render data in the memory of GPU, but what happens if i want to use jit.gl.pix in this pipeline?
- Should i grab the whole scene into a texture and process the texture-matrix with jit.gl.pix? Will this move the data back to the CPU, or can i use jit.gl.pix on data resided in the GPU memory?
- If i want to post process a scene with jit.gl.pix then i should capture it with jit.gl.node, send it to jit.gl.pix , then jit.gl.videoplane to display it? Will this whole process run on the GPU or will any texture data come back to the computer memory?
- I’d be interested in the way how i can move the less data beetween CPU and GPU…
- Is jit.gl.pix somehow equivalent to the fragment shader of a JXS file or replace it, or just an additional post processing shader in that case if i capture the output of an vp/fp?
thanx in advance if someone could enlighten me or at leat point to somewhere i can learn about it,
nevermind, Tutorial 42′s end shed some light on the RAM/VRAM thing…
C74 RSS Feed | © Copyright Cycling '74