You can change the pitch or the speed of an audio file online by loading it into a browser tab, moving the speed and pitch sliders independently, and exporting a new copy. The processing runs locally with FFmpeg compiled to WebAssembly, so the audio is never uploaded, and the pitch can move by up to 12 semitones in either direction.
Speed and pitch are separate controls by default. The speed slider retimes the track from 0.5x to 2x without altering the key, so a voice at 1.5x still sounds like the same person. The pitch slider moves the key alone, from minus 12 to plus 12 semitones, and leaves the running time untouched. A linked mode joins the two when you want the tape deck effect. Every change is written to a new file, and the original stays on your device.
Audio sits in the middle of a lot of everyday work: practising an instrument along to a track, fitting a voice note into a shorter slot, tuning a karaoke backing to a comfortable key. Speed and pitch come up in all of them, and they are not the same request. This guide explains what each control actually does, how the linked mode differs, and where the approach stops working.
Audio Pitch and Speed Changer
Retime a track from 0.5x to 2x and shift its key by up to 12 semitones, independently or linked. The FFmpeg filter chain runs in your browser.
Speed and pitch are two separate things
Speed is how long the audio takes to play. Pitch is how high or low it sounds. On a tape or a vinyl record you cannot move one without the other, because running the medium faster shortens the music and raises its key at once. Digital processing splits the two, so a track can play faster at its original key.
Most online changers still bind the two together, because plain resampling is the easiest way to alter either one. That is where the chipmunk effect comes from: speeding the file up resamples it and lifts the pitch along with the tempo. This tool keeps the controls independent by default and uses a tempo filter that stretches time instead of resampling, so voices keep their natural tone while the track gets faster or slower.
How to change the pitch or speed of an audio file
The whole job runs in the browser tab, with no install and no account. Open the Audio Pitch and Speed Changer and follow five steps.
- Add your audio. Drop an MP3, WAV, M4A, AAC, OGG, FLAC or Opus file onto the upload box. The tool reads the length and sample rate on your device; if the browser cannot decode the format, it points you to the Audio Converter instead.
- Set the speed. Drag the speed slider between 0.5x and 2x. It moves in 0.05x steps, so 1x is the original speed, anything below slows the track down and anything above speeds it up. Leave it at 1x if you only want to change the key.
- Set the pitch. Drag the pitch slider from minus 12 to plus 12 semitones in whole steps. Zero leaves the key untouched, 12 raises it a full octave, and minus 12 drops it an octave. Skip this slider if you only want to change the tempo.
- Choose the output format. MP3 is the default and is small enough to share. WAV is uncompressed, for a file that goes back into an editor. Keep original format still exports MP3, because any change to speed or pitch has to be re-encoded.
- Change the audio and download. Click Change audio, check the preview in the result card, then press Download. The download carries a suffix such as -1.5x+3st, so versions of the same track never collide.
One detail worth knowing: with the speed at 1x and the pitch at 0 semitones there is nothing to encode, so the button stays disabled. The first change downloads the FFmpeg engine, about 30 MB, into the browser, and every run afterwards works offline.
What a semitone is and how much to shift
A semitone is the smallest standard step in Western music, the distance between two adjacent keys on a piano, black or white. An octave holds 12 of them, which is why the pitch slider stops at 12 in each direction. The full slider is therefore exactly one octave up or down, and every number in between is a musical interval you can name.
| Shift | Interval | What it does |
|---|---|---|
| +1 semitone | Minor second up | The smallest audible move, useful for nudging a karaoke track into range |
| +2 semitones | Major second up | A whole tone up, a common transposition for a singer who needs headroom |
| +7 semitones | Perfect fifth up | A bright, open lift, often used for a harmony line |
| +12 semitones | One octave up | The same melody in a higher register |
| -12 semitones | One octave down | The same melody an octave lower, a deeper voice |
The slider moves in whole semitones, so a finer shift below one step is not possible here; a quarter tone would need a different tool. Small shifts generally sound clean, while a full octave up or down is where the processing starts to become audible.
Independent mode and linked mode
Independent mode is the default and gives each slider its own job: speed changes only the tempo, pitch changes only the key. You can slow a track to 0.75x for practice and still lift the key by two semitones, and the two settings do not interfere with each other.
Linked mode makes the key follow the tempo, imitating a tape deck or a record player. The relationship is fixed by one formula: the shift in semitones is 12 times the base two logarithm of the speed. At 0.5x that works out to minus 12 semitones, exactly one octave down, and at 2x it is plus 12 semitones, one octave up. A speed of 1.5x lands near plus 7 semitones. In this mode the pitch slider is disabled and only shows the current value, because the speed alone decides it.
Linked mode is the right choice for a slowed and reverb edit or a nightcore version, where the pitch movement is the point. If you want the tempo change without the pitch movement, leave it off; there is no way to link only part of the pitch in this tool.
What happens to the length
The two controls affect the running time differently, and that difference is easy to predict. A pitch shift resamples the audio to move the key and then resamples it back to the original rate, with a tempo stage correcting the timing afterwards, so the two operations cancel out and the length is unchanged. The speed slider is the only one that alters the duration.
The rule is simple: the output length equals the original length divided by the speed. Half speed doubles the playing time, roughly double speed halves it, and 1.5x cuts it to two thirds. When both controls are set at once, the length still follows the speed alone. The tool shows the estimated length next to the controls and updates it live.
That number is an estimate calculated from the source length, not a promise. The final duration comes from the encoded file once it is ready.
What this tool cannot do
The limits are worth reading before you commit, because in these cases a different route is genuinely better.
- The sliders stop at 0.5x to 2x and minus 12 to plus 12 semitones. Beyond those ranges the algorithms introduce artefacts, so a stronger result means running the file twice and accepting a second re-encode.
- Large pitch shifts lose quality. Moving a whole octave is a stretch for any resampling method, and the result sounds more processed than a one or two semitone nudge. Moderate shifts stay clean.
- The browser has to be able to decode the file. The tool reads the sample rate and length with the Web Audio API, so a format it cannot open, such as some FLAC or AMR variants, is refused with a note to convert it to MP3 or WAV first.
- Long recordings depend on device memory. No duration limit is written into the tool, but the engine runs in browser memory, so a very long file can fail on a low-memory device.
- Nothing is written losslessly. A speed or pitch change cannot be saved into the original container, so the output is always re-encoded, and Keep original format exports MP3 for that reason.
- One file at a time. There is no batch queue, so a set of clips is processed one after another rather than in one pass.
- The first run downloads about 30 MB. That is the FFmpeg engine, cached afterwards, and on a slow connection the first click sits at the loading stage for a while.