Hey Lyxa, I've been tinkering with a new DSP library for real‑time audio synthesis. Ever thought about how a neural net could learn a synth's voice and play it back with more emotion? Maybe we could experiment together?