Gesture Synth: Play Synthesizer with Hand Gestures
Experience the world's most intuitive browser-based gesture synth. Hold up your hands to your webcam to shape chords, trigger inversions, sweep cutoff filters, and create evolving soundscapes in real time — with no equipment required.
How Hand Gesture Tracking Works
Tunevate uses your device’s webcam and client-side computer vision to track 21 hand landmarks in 3D space with sub-millisecond precision. Here is how your hands shape chords, melodies, and sound parameters:
Fingers Select Harmonic Scale Degree
Raise 1 to 5 fingers on your left hand to select the root note according to your chosen musical key (1 = Tonic / I, 2 = Supertonic / ii, 3 = Mediant / iii, 4 = Subdominant / IV, 5 = Dominant / V).
Voicing, Inversions & Octaves
Raise fingers on your right hand to dictate the voicing structure: 1 finger plays root position, 2 fingers plays 1st inversion, 3 fingers adds a 7th, and 4 fingers adds extended color.
Continuous Filter Sweeps & Dynamic Volume
Tilt your right wrist up and down to sweep the lowpass filter cutoff frequency. Move your hand higher or lower vertically to swell and attenuate master volume.
Built for Real-Time Music Creation
Engineered with Tone.js, Web Audio API, and MediaPipe AI for professional synthesis, responsive acoustic instruments, and seamless loop recording.
Subtractive & FM Synth Engines
Switch between multi-oscillator Warm Pads, Bright Saws with chorus, and punchy Square Lead voices powered by Tone.js Web Audio.
MediaPipe 3D Landmark Tracking
Tracks 21 hand joints per hand at 60 FPS directly inside the browser canvas using WebAssembly and GPU acceleration.
Zero Video Recording or Uploads
Your webcam feed is processed strictly in your local device memory. No video or audio is ever sent over the network.
Full 12-Key Harmony Engine
Select any tonic from C to B. The music theory engine automatically maps gestures to musically valid diatonics so you never hit a wrong note.
Integrated Multitrack Looper
Record your gesture synth phrases directly to quantised loop tracks, then overdub drums, pianos, and basslines.
Instant Zero-Install Access
No VSTs, no DAW, no drivers, and no account creation. Open the page, enable camera permissions, and start performing.
Audio & Machine Learning Architecture
| Synthesis Architecture | Subtractive Dual/Triple Sawtooth + Polyphonic FM |
| Vision Latency | < 16ms (60 FPS MediaPipe HandLandmarker) |
| Audio Engine | Web Audio API / Tone.js v15 |
| Filter Section | State-variable 12dB/octave lowpass modulated by wrist angle |
| Polyphony | Full polyphonic chord voicing up to 8 voices |
| Platform Support | Chrome, Edge, Firefox, Safari on Windows, macOS, Linux, ChromeOS |
Frequently Asked Questions
Everything you need to know about playing this instrument online with hand gestures.
What is a gesture synth?
A gesture synth is an electronic musical instrument that uses motion or computer vision sensors to translate physical hand and body movements into musical notes and audio synthesis parameters, without physical keys, wires, or wearable gloves.
How does Tunevate turn hand gestures into synth music?
Tunevate captures video from your webcam and uses Google's MediaPipe AI to track 21 coordinate landmarks on each hand. The left hand detects finger counts and wrist tilt to pick the musical root and quality, while the right hand selects chord inversions, filter sweeps, and volume.
Do I need any special hardware or sensors to use this gesture synth?
No special hardware is required. Any standard built-in laptop webcam, USB webcam, or phone camera connected to a modern browser will work instantly.
Is my webcam video private when playing the gesture synth?
Yes, 100% private. All video landmark detection runs locally on your computer via client-side WebAssembly. No video frames, images, or audio clips are ever transmitted to any external server.