The torch has been passed to Opus, and I'd say we're in business! Still 200+ fps.
As it turns out the reason my previous attempt had so much trouble with the sphere primitives is that they aren't supported by my python bindings. Fable just implemented sphere intersection off the raytracing cores.
Dribbles
You can do this with Julia fractals as well with a constant instead of start_pos.
The glowing spots where particles bunch up clearly have to do with the fractal structure, but I haven't looked into the math at all. Maybe I'll see if Claude can figure out what's going on.
These are reminding me of my favorite particle-test-pattern. It's really simple but creates a cool structure:
- Initialize particles in a grid
then loop:
- Calculate the particle's image under a mandelbrot iteration and apply it as a force:
vel+= dt * (complex_square(pos)+start_pos);
pos+= dt* vel;
Video
Video
Video
Oops! All Paperclips
Oops! All Paperclips
Oops! All Paperclips
Oops! All Paperclips
Cool. Still 700 fps. I'm curious to see how long it takes to rebuild the acceleration structure, would be sweet to get this in realtime with Fluoddicles
