Feature request: WLED Audio Sync

[netmindz]

Would be nice to also have support for WLED audio sync so that you can use the built-in audio effects too

[rlfnb]

Thanks for the suggestion — adding support for WLED Audio Sync is definitely possible and technically quite simple.

Right now, rusty_lights already performs its own audio analysis and sends fully rendered RGB frames to WLED via DDP or E1.31. So the audio itself never reaches WLED — it just displays precomputed pixels.

WLED's built-in audio-reactive effects are based on relatively lightweight FFT data (around 16 bands, 8-bit precision) processed on the ESP32. That’s great for embedded use, but it’s considerably simpler than rusty_lights’ DSP pipeline (257 FFT bins, Mel scaling, beat detection, etc.).

Because of that, sending FFT data to WLED for it to re-render effects would overlap with what rusty_lights already does, offering limited benefit in most setups.

That said, adding basic Audio Sync output wouldn’t be hard and could make rusty_lights feel more “native” within the WLED ecosystem — especially if there’s broader interest from the WLED side. It’s not entirely clear yet how much demand there is (my small doc PR to WLED-Docs has been quiet too), but if more people show interest, it’s easy to add.

I’ll keep this open for feedback — if others would like to use rusty_lights together with WLED’s internal effects, we can make that happen quickly.

[netmindz]

I'm releasing the 16beta this weekend that has quite a few more Audio Reactive effects. Sync is not just the 16 fft values. Having more fft bands is pretty useless for anything other than a spectral analizer as otherwise they are too limited in frequency coverage, so you need to combine them to map to affect attributes, franky 16 is too many really, you are best grouping into bass mid and treble

[rlfnb]

Good point about the bass/mid/treble grouping - that's exactly the approach we use internally, but with 24 Mel-scaled bands and dedicated beat detection on top. We're currently working on effects that go further than what's possible on an ESP32, like a full-resolution spectrogram and multi-band onset detection (independent kick/snare/hi-hat triggers).
That said, we've started building WLED integration from the other side: auto-detecting LED count and config directly from the WLED JSON API, so setup becomes zero-config when using DDP. We're not ruling out AudioSync support - the protocol is straightforward - but our priority is making the effects themselves strong enough that you wouldn't want WLED to re-render them. Happy to revisit if there's more demand.