Pitch Shifter
Transpose a recording up or down while keeping its original length
100% In-Browser · Private & Instant · Free Forever
Your file is decoded, processed, and re-encoded in this browser tab. Nothing is uploaded to an OfflineTTS processing API.
Interactive Workspace
Pitch Shifter
Advertisement
Adsterra privacyMove a recording up or down in pitch without stretching or squashing its duration. The whole track shifts by the number of semitones you choose, so a vocal line stays in time with the same backing track while sitting in a different key.
How it works: the tool oversamples the audio by the pitch ratio, then uses a phase-vocoder time stretch to bring the duration back. Pitch changes, length does not. The vocoder window is fixed, so quality is best on sustained material and vocals; dense percussion is the hardest case for any phase vocoder.
Transpose by semitones: 12 semitones equals one octave. Small shifts of one or two semitones are near-transparent on most material. Large shifts — five semitones and beyond — start to sound processed, especially on cymbals and sibilants.
Formant colour: the second control adds a further tone shift. Raising it along with pitch keeps a shifted vocal sounding brighter rather than muffled; lowering it adds chest weight. This is an approximation of formant preservation, not a true formant-correction algorithm.
Good uses: matching a sample to a project key, changing the register of a spoken take, creating a deeper or lighter character voice, or building a pitched variation of a loop.
Export: WAV keeps the processed signal exactly as rendered; MP3 is smaller for review. Decoding, processing, and encoding all run in this browser tab. The file you select is not uploaded, and the rendered audio is not stored on any server.
Supported Formats, File Limits & Local Privacy
Formats & Codecs
| Input formats | MP3, WAV, FLAC, OGG, Opus, M4A, AAC, plus MP4 & WebM audio tracks |
|---|---|
| Output formats | Uncompressed 16-bit PCM WAV; MP3 at 128, 192, or 320 kbps |
| Decoding path | Native browser AudioContext.decodeAudioData + offline Web Audio graph |
Size Limits & Privacy
- No server upload cap: files never leave your browser tab, so there is no remote file-size paywall.
- Memory ceiling: decoded audio is held in browser RAM as 32-bit float PCM (~10 MB per minute of 44.1 kHz stereo). Tracks under 30–45 minutes run reliably on standard laptops.
- Zero data retention: closing or refreshing the tab clears every audio buffer from memory immediately.
How This Compares to Cloud Audio Tools
Most browser-based audio utilities (such as 123apps or Clideo) upload your media to remote processing queues. Here is how OfflineTTS differs:
| Criterion | OfflineTTS (Pitch Shifter) | Typical Cloud Web Editors |
|---|---|---|
| File upload | None — decoded & rendered in your browser tab | Audio file uploaded to remote servers |
| Account / daily cap | No signup, no daily file quota, no watermarks | Often gated by daily file limits or paid tiers |
| Speed on large files | Instant start (no upload wait); bounded by device RAM | Upload + queue wait, then server-side CPU render |
| Works offline | Yes, once the page is loaded | Requires active network round trip |
How It Works
Load audio
Choose the file you want to transpose.
Set semitones
Pick the interval — negative values go down, positive values go up.
Add formant colour
Optionally shift the tone further to keep the voice character natural.
Render and export
Preview against the original, then download WAV or MP3.
Features
±12 Semitones
One semitone steps in both directions, up to a full octave each way.
Length Preserved
Phase-vocoder stretching restores the original duration after resampling.
Formant Colour
An extra tone shift so a transposed voice keeps its character.
A/B Preview
Switch between the original and processed audio without re-rendering.
Choosing an Interval That Sounds Natural
The most transparent intervals are the ones that other musical parts already support. Shifting a vocal up two semitones to reach a different key works because the harmonic relationships stay intact. Shifting it up seven semitones and calling it a harmony part does not — that is a different musical decision that needs its own take.
For spoken word, treat larger shifts as a character effect rather than a correction. Down four to six semitones gives a heavier, more deliberate read; up three to five gives a lighter, quicker one. Both change how the speaker is perceived, so check whether the result still matches the intent of the original recording before publishing.
Always render a short excerpt first. A fifteen-second sample that contains the highest and lowest notes of the material tells you more than a full-length render, and it costs seconds instead of minutes.
Processing Boundaries and Review
Every effect here is deterministic: the same file and the same control values produce the same output. That means you can reproduce a result later, but it also means the tool will not guess what you intended — check a short excerpt before rendering a long file, and keep the original.
This tool shifts the whole file by one interval. It does not analyse a melody, correct a performance to a scale, detect a key, or split a mix into parts that can be shifted independently. If you need note-level correction, that is a different class of tool and it is not what this page offers.
Phase-vocoder artefacts are easiest to hear on solo instruments and exposed vocals, and easiest to overlook in a dense mix. Judge the result on the material you actually intend to publish, not on a test tone. If a shift sounds smeared, reducing the interval or applying the effect to a narrower frequency range in a DAW will usually sound better than pushing this control further.
Related Tools
Audio Speed & Tempo Changer
Slow down or speed up audio while holding the original pitch
Audio Equalizer
Ten bands from 32 Hz to 16 kHz, with presets and a spectrum view
Nightcore Maker
Speed a track up with a matching pitch lift — the classic nightcore sound
Voice Changer
Sixteen character presets, each one still fully editable
Pitch Shifter — FAQ
Does pitch shifting change the speed?
No. The tool resamples to change pitch, then time-stretches back to the original duration. A three-minute song stays three minutes long.
How many semitones can I shift?
Up to 12 semitones in either direction, which is one full octave. Shifts of one to three semitones stay close to transparent; larger shifts become progressively more audible as processing.
Why does shifted audio sound metallic on drums?
Phase vocoders assume that content inside each analysis window is roughly periodic. Transients such as drum hits break that assumption, so the phase estimate smears them. This is a known limitation of the technique, not a bug in this implementation.
What is the formant control for?
Formants are the resonant peaks that give a voice its size and character. Shifting pitch alone makes a voice sound thinner or heavier than expected; the formant control compensates by shifting those resonances in the same direction.
Is my audio uploaded?
No. The shift is computed in this browser tab.
Technical Standards & Upstream References
More Local Audio Tools
Every audio utility here runs in the browser tab. Browse the full set of editing, conversion, and measurement tools.