Skip to main content
Experimental

Lab

Works in progress. Half-finished prototypes. Messy experiments pushing the boundaries of what's possible with audio and AI in the browser.

Not everything here works perfectly—that's the point.

01

Vibe Rater

Prototype

Real-time audio mood classification using machine learning. Analyzes microphone input to detect emotional characteristics in music and sound.

TensorFlow.jsWeb AudioML
02

Audio Embeddings

Concept

Exploring audio similarity search using neural audio embeddings. Upload a sound, find similar sounds.

MLEmbeddingsSearch
03

Generative Sequences

Planned

AI-assisted pattern generation for the ACID-303. Train on classic acid patterns, generate new ones.

GenerativeSequencerML

About the Lab

Research & Development

This section explores emerging intersections of audio processing and machine learning. Each project investigates a specific technical challenge—from real-time audio classification to neural embedding search—with the goal of understanding what's possible in the browser.

Why It Matters

The next generation of creative tools will combine traditional DSP with AI capabilities. Understanding both domains—and how they integrate—is essential for building production-quality music software. These experiments inform that understanding.