(A)live check-in 2

Notes to self, because in the past I have done interesting simulations with recordings and later no idea what I did. So for today's June 6, 2025, these are simulations of combined pulse and orbit "veras" that happily took some very interesting forms. Here are the specs for my future self, although of course, every simulation has random starting parameters as far as color etc. And not all my simulations with these exact specs yielded such interesting results. Nevertheless: Species = 3, particles = 1000, speed = 1, weight = 0.5, rotate = 4, file = osc_pulse.py

I would post the raw vids, but they are long, and our voices are on them, and the audio is there for useful auditory feedback (something which I am learning to value), but annoying (too high pitched, plain sine wave). I'll later make a silent composite.

Thanks to the auditory feedback, I have made strides in finding potentially musically interesting mapping parameters for attraction and radius. I am still having trouble with the repulse parameter - I think there is something wrong in the code:

And the stupid thing is, because I didn't take better notes and comments, I don't know where this repulse idea came from, I left no credit, so I don't know if I made it up wrong or copied it from somewhere wrong. [edit: it was a Max patching problem, and when it worked, everything kept flying away no matter how small the repulse, so I ditched this parameter for now.]

Looking for roles in agency, I think I am going to have the player provide audio triggers to re-map the species rules. If the player wants a change in the visuals, they can play their trigger sound (either trained on an mlp classifier, or through machine listening real-time descriptors), and the mappings will randomly change. The player can continue triggering until they see something interesting. Just a working idea for now.

Final thought for the day, use 4 species instead of 3. Could be because sometimes the random colors are too close to each other to be distinguishable. However, I like the 3-species interaction better (at least today), so an alternative would be to monitor the simulation before fading in the visuals, and abort (I hate that word though) any simulation with an uninteresting color and movement scheme. Sometimes, the starting seed is just too snowy, and no matter how many parameters I change, nothing "interesting" happens (too much random particle snow).