Tip: Click or slide across keys, play with QWERTY keys A–;, press 1–6 for IPA vowels, or plug in a USB MIDI controller.
100% local Web Audio synthesis — no audio is ever uploaded.
Interactive human vocal tract synthesizer — drag the tongue, play MIDI or keyboard, try presets, type text to speak, and export WAV.
Tip: Click or slide across keys, play with QWERTY keys A–;, press 1–6 for IPA vowels, or plug in a USB MIDI controller.
100% local Web Audio synthesis — no audio is ever uploaded.
Click or touch anywhere inside the mouth cavity to move the tongue body and sculpt vowel formants in real time. Tap the lips, alveolar ridge, palate, or velum to add consonants and nasal resonance.
Slide along the pitch control pad to glide pitch horizontally and adjust tenseness vertically. Enable Choir Mode to harmonize your vocal tract into lush multi-voice intervals and chords.
Plug in a USB MIDI controller (Chrome, Edge, or Opera) or play notes directly on your QWERTY keyboard. Open Text-to-Speech to type any English phrase and watch the vocal tract animate each phoneme.
Hit Record to capture your performance directly in the browser with zero server uploads, then download an uncompressed 16-bit PCM WAV file ready for Ableton, FL Studio, Logic, or video edits.
| IPA | Type | Example | Tongue Position | Constriction | Velum |
|---|---|---|---|---|---|
| [i] | Vowel | fleece / "ee" | High front (forward & raised) | Open lips, narrow palatal channel | Closed (oral) |
| [u] | Vowel | goose / "oo" | High back (pulled toward soft palate) | Tightly rounded lips | Closed (oral) |
| [ɑ] | Vowel | palm / "ah" | Low back (pharyngeal narrowing) | Wide open mouth & lips | Closed (oral) |
| [æ] | Vowel | trap / "a" | Low front (flat & forward) | Open jaw, unrounded lips | Closed (oral) |
| [ə] | Vowel | comma / "uh" | Mid central (relaxed neutral tube) | No strong narrowing anywhere | Closed (oral) |
| [s] | Fricative | sip / "sss" | Tongue blade raised anteriorly | Very tight alveolar gap + voiceless hiss | Closed (oral) |
| [ʃ] | Fricative | ship / "shh" | Tongue raised slightly behind alveolar ridge | Post-alveolar gap + slight lip rounding | Closed (oral) |
| [m] | Nasal | hum / "mmm" | Neutral oral cavity | Complete lip closure (bilabial) | Open (nasal resonance) |
| [n] | Nasal | noon / "nnn" | Tongue tip raised to alveolar ridge | Complete alveolar closure | Open (nasal resonance) |
| [t / d] | Plosive | top / dog | Quick tap and release at alveolar ridge | Momentary full oral closure (voiceless [t], voiced [d]) | Closed (oral) |
Pink Trombone is an interactive physical-modeling vocal tract synthesizer based on Neil Thapen’s classic 2017 open-source project (MIT License). Rather than triggering pre-recorded audio clips, it computes a live acoustic simulation of the human vocal apparatus.
At the throat, a glottal pulse model (inspired by Rosenberg’s glottal waveform) generates pitched vocal-fold vibration mixed with aspiration noise. That signal travels through a 44-segment Kelly-Lochbaum digital waveguide representing the pharynx and mouth, coupled at the velum to a branching nasal tract. Moving the tongue, lips, or palate changes the cross-sectional diameters of the tube, shifting formant frequencies (F1, F2, F3) just like a real human voice.
| Capability | pinktrombone.net | 2017 Original |
|---|---|---|
| Audio Engine | Dedicated Web Audio AudioWorklet thread (low-glitch) | Deprecated ScriptProcessorNode on main thread |
| Layout & Mobile | Responsive multi-touch canvas + dark & light themes | Fixed 600×600 canvas without viewport meta |
| Musical Playability | Web MIDI controller, QWERTY keyboard & Choir Mode | Single-voice mouse/touch pitch pad only |
| Phonetics & Speech | One-click IPA presets + animated English Text-to-Speech | Manual dragging only |
| Export & Sharing | Client-side WAV recording + shareable preset links | No built-in recording or preset saving |
Master Pink Trombone controls, drag the vocal tract to shape every IPA vowel and consonant, control nasal resonance, and make Pink Trombone talk with Text-to-Speech.
Connect a USB MIDI controller or play QWERTY keys on Pink Trombone, perform vocal songs with Choir Mode harmonies, and export WAV recordings into your DAW.
Compare Pink Trombone Online in your browser with the 2017 original, Xfer Records free VST3/AU plugin, Max/MSP ports, Blob Opera, Incredibox, and vocal synth alternatives.
Explore the acoustic physics inside Pink Trombone: source-filter theory, Rosenberg glottal pulses, the 44-segment Kelly-Lochbaum waveguide, formants, and nasal coupling.
Pink Trombone is a real-time articulatory speech synthesizer originally created by Neil Thapen in 2017. Instead of playing back recorded vocal samples, it simulates the physics of a human voice — pairing a Rosenberg glottal pulse generator with a 44-segment Kelly-Lochbaum vocal tract waveguide and nasal cavity. On pinktrombone.net, the audio engine runs inside a modern Web Audio AudioWorklet directly in your browser.
Yes. Pink Trombone on pinktrombone.net is 100% free, requires no account or downloads, and runs over standard HTTPS in modern desktop and mobile browsers (Chrome, Safari, Firefox, and Edge). It is built for classrooms, linguistics demos, and music students who need a clean, reliable browser synthesizer with multi-touch support.
Plug in any class-compliant USB or Bluetooth MIDI keyboard and open Pink Trombone in a Web MIDI browser (Chrome, Edge, or Opera) to play notes, pitch bend, and modulation automatically. No MIDI gear handy? Switch on Keyboard mode and use your QWERTY keys (Z–M for the lower octave and Q–P for the upper octave) while shaping vowels with your mouse or trackpad.
Open the Speak tab, type an English word or sentence, adjust the speaking rate if you like, and press Speak. The built-in phoneme sequencer maps text to articulatory targets and animates the tongue, lips, velum, and voicing in real time so you can both hear and see how each syllable is formed.
Yes. Press Record to capture the live synthesizer output locally in browser memory, press Stop when finished, and click Download WAV to save a lossless PCM WAV file. All audio synthesis and WAV encoding happen strictly on your device — nothing is ever uploaded to a server.
Choir Mode turns the single vocal tract into a multi-voice harmonic ensemble by adding tuned harmony voices (such as octave, fifth, major triad, or minor triad) alongside your live articulation. Moving the tongue or triggering presets morphs the entire choir together in real time.
Mobile browsers block web audio until you interact with the page. Tap the Start Sound button or touch the vocal tract canvas once to unlock the AudioContext. On iPhone and iPad, also check that the physical Ring/Silent switch on the side of your device is not muted and that media volume is turned up.
The original Pink Trombone vocal tract model and interface were created by Neil Thapen in 2017 and released under the MIT License at dood.al/pinktrombone/. Pink Trombone Online (pinktrombone.net) builds on that open-source physical model with an AudioWorklet engine, responsive multi-touch layout, MIDI and keyboard support, choir harmony, IPA presets, text-to-speech, and WAV recording.