Hey Shara, I’ve been trying to use a little bit of code to help plan murals that react to city noise—thought you’d find the idea of mixing algorithmic patterns with street art intriguing?

I’m leaning toward JavaScript with the Web Audio API for the capture part—easy to hook into a browser, no extra installs. The data flow is basically: microphone → audio stream → analyser node → getFrequencyData → feed that array into a tiny shader or canvas loop. For mapping, I keep it simple: low frequencies become warm reds and oranges, mid frequencies turn into greens, high frequencies go to cool blues or glitchy white sparks. I scale the amplitude to brush size or opacity, and I add a little beat‑synchronisation to trigger splashes when the volume spikes. To stay snappy I use requestAnimationFrame for the paint loop and only pull the FFT data every 50 ms, then debounce any heavy calculations. If I need to push it further I can move the frequency crunching to a Web Worker so the UI thread isn’t blocked. That’s the skeleton—let’s tweak it until it feels like the wall is breathing with the city noise.

Thanks for the pointers—will crank the fftSize to 2048, build a small hue table, and add a quick moving‑average for the splash trigger. I’ll keep an eye on sample rates and push the analyser into a worker if the frame drops hit a wall. If something feels off, I’ll holler.