BPM

Local offline DJ track browser built with Tauri + React + Python.

Features (MVP progress)

• Choose a music folder and scan tracks recursively
• SQLite persistence with rescan deduplication
• Analyze selected tracks or all tracks via Python CLI
• BPM, raw features, and normalized Energy (1–10) in the table
• Batch analysis queue with progress events
• Two-deck preview with waveform, play/pause, seek, crossfader
• Optional simple BPM sync via playbackRate

Requirements

• Node.js 20+
• Rust (via rustup)
• Python 3.11+
• macOS build tools for native deps (Xcode CLT)

Setup

cd ~/WebstormProjects/BPM
npm install
npm run python:setup

Development

npm run tauri:dev

Python analyzer

./python/venv/bin/python python/analyze_track.py "/path/to/track.mp3"

Example output:

{
  "duration": 384.2,
  "bpm": 124.0,
  "rms": 0.18,
  "onset_rate": 3.4,
  "spectral_centroid": 2650,
  "energy_raw": 0.72
}

Architecture

• Frontend: React + TypeScript (src/)
• Backend: Rust Tauri commands (src-tauri/src/)
• Database: SQLite in app data dir (bpm.db)
• Analysis: python/analyze_track.py (librosa)
• Decks: wavesurfer.js + Web Audio volume / playbackRate