About
A browser-based acoustic analysis tool that extracts voice quality metrics from audio files — no server, no uploads, no privacy concerns.
👥 Who is this for?
- Speech-language pathologists – quick voice assessments
- Voice researchers – batch analysis without Python setup
- Students & educators – learn acoustic features hands-on
- Clinicians – privacy-first patient voice screening
⚙️ How it works
- Runs entirely in your browser using WebAssembly (praat-wasm)
- Audio never leaves your machine — 100% private
- Supports WAV, MP3, M4A, OGG
- Calculates ~40 acoustic metrics — F0, jitter, shimmer, HNR, CPP, DSI, AVQI, CSID, and more
✨ Key features
- Batch processing – analyze multiple files at once
- CSV export – download results for further analysis
- Toggle advanced metrics – disable AVQI/CSID for speed
- No file size limits (browser-dependent)
💡 Why build this?
- Parselmouth (Python) is powerful but requires coding knowledge
- This tool makes the same metrics accessible to anyone with a browser
- Built for rapid prototyping, teaching, and clinical screening
📊 Metrics supported
The tool extracts over 40 acoustic metrics, identical to Parselmouth (Praat in Python):
F0 mean / min / max / std
Pitch range (semitones)
Jitter (local, RAP, PPQ5, DDP)
Shimmer (local, APQ3, APQ5, APQ11, DDA)
HNR / NHR
Voicing breaks (DUV, NVB, DVB)
CPP / CPPS
LH ratio
LTAS slope & tilt
DSI
AVQI
CSID
⚠️ Limitations & caveats
- Results are not a clinical diagnosis — consult a qualified professional for medical decisions
- Accurate results require a quiet recording environment and consistent mic distance (~15 cm from mouth)
- This tool is for research and educational purposes only
🛠️ Tech stack
- Praat (via praat-wasm)
- Vanilla JavaScript
- No external dependencies or tracking
- Audio processing: 100% client-side