It's using a camera to detect particles only based on rendering data (no GPU Event involve at this stage).
The computeShader "Reduction" extracts the needed information to minimize the transfer from GPU to CPU.
Reference:
- Inspired by this thread https://forum.unity.com/threads/getparticlesysteminfo-latency.1243120/#post-7919116
- Taking inspiration from this article https://pixeleuphoria.com/blog/index.php/2021/01/23/synthesizing-procedural-audio/#htoc-creating-audio for synthesizing sound
Next Step:
- Avoid missing collision using a child system and GPU event to render collision data in an offscreen camera
See also: