Audio-to-LED
guest@rkulskis.github.io:~/projects$ echo LINKS: $(ls -a)
I built a system that syncs LEDs to music in real-time which using the following steps:
- Split the audio using AI to get isolated tracks.
- Map each track's transients to rhythm data.
- Stream the rhythm data to a microcontroller, e.g. Teensy 4.1, and perform an on-the-fly algorithm to flash LEDs with each track.
I built this with a couple friends while at Boston University for fraternity purposes. If I were to roll out a V2, I'd have the AI process the audio online rather than apriori, with appropriate buffer to match the frequency of the control loop.