Transpose any audio file up or down by semitones without changing its duration — from subtle tuning to a full octave, all processed privately in your browser.
Drag and drop an audio file here, or click to browse
MP3, WAV, M4A, OGG, FLAC, AAC, Opus, or WebM — up to 100MB
Upload an audio file to shift its pitch
This free tool shifts any audio file's pitch up or down by up to 12 semitones (a full octave) while keeping the duration exactly the same. It uses FFmpeg's standard asetrate/aresample/atempo recipe: the sample rate is reinterpreted to shift pitch, then tempo is compensated back so playback length never changes. Everything runs locally in your browser via WebAssembly — nothing is uploaded.
Is this audio pitch changer really free?
Yes. There's no signup, no watermark, and no limit on how many files you process — everything runs in your browser.
Does my audio get uploaded to a server?
No. Pitch shifting happens entirely locally using FFmpeg compiled to WebAssembly. Your file never leaves your device.
What audio formats can I upload?
MP3, WAV, M4A, OGG, FLAC, AAC, Opus, and WebM are supported, up to 100MB per file.
How many semitones should I shift a voice?
It depends on the effect you want. A shift of 1-3 semitones is a subtle correction, often used to nudge a recording into tune. Around 3-5 semitones noticeably changes a voice's character while staying natural-sounding. Beyond about 7-8 semitones the result starts sounding processed, and a full octave (12 semitones) is a dramatic, obviously artificial transformation — useful for creative effects rather than natural-sounding results.
Why doesn't the audio speed up or slow down when I shift the pitch?
This tool uses a three-stage FFmpeg filter chain: it first reinterprets the sample rate to shift pitch (which would also change speed), then compensates with a tempo filter that restores the original duration without touching pitch again. The result is a pitch shift with no change in playback length.
Why does a big pitch shift sound robotic?
Sample-rate-based pitch shifting (the technique this tool uses) doesn't independently model the human voice or an instrument's timbre — it mathematically stretches or compresses the waveform's frequency content. At small shifts this is inaudible, but beyond roughly 7-8 semitones the formants (the resonant frequencies that give a voice its natural character) shift along with the pitch, which is what produces the telltale robotic or "helium" quality. This is an honest limitation of the technique, not a bug — professional formant-preserving pitch shifters use more complex algorithms to avoid it.
Your audio is processed entirely in your browser using WebAssembly. It is never uploaded, stored, or seen by our servers.