Audio Speed Changer - Slow Down or Speed Up Music

Shift pitch by up to ±12 semitones and change speed from 0.25× to 4× — independently — with the studio-grade Signalsmith Stretch engine, then export as WAV or MP3, entirely on your device.

Load a file, set the pitch in semitones (0 keeps the original pitch), set the speed (1× keeps the original speed and duration), and process. Pitch and speed never interfere with each other: a pitch shift preserves the duration exactly, and a speed change preserves the pitch. Nothing is uploaded.

Audio Speed Changer - Slow Down or Speed Up Music
Shift pitch by up to ±12 semitones and change speed from 0.25× to 4× — independently — with the studio-grade Signalsmith Stretch engine, then export as WAV or MP3, entirely on your device.
0 st

0 st keeps the original pitch. Each semitone is one piano key: +12 st is one octave up, −12 st one octave down. Shifting pitch never changes the duration.

Speed changes tempo and duration only; the pitch stays wherever the slider above puts it. The default 1× leaves the timing untouched.

Pitch: 0 st · Speed: 1.00× · Duration: 100%

New duration: 100% of the original

At 1× a 10:00 file becomes 10:00.

Output duration = input duration ÷ speed, and pitch factor = 2 raised to the power of semitones ÷ 12. The two are independent: +3 st at 1× keeps the length at 100%, and 2× at 0 st keeps the key while halving the length.

Local-only processing: the file is decoded, stretched, and re-encoded on your device and never sent to a server.

About this pitch and speed changer

This tool changes the pitch and the speed of an audio file as two completely independent controls. The primary control is the pitch shift, measured in semitones — the steps of a piano keyboard. Move it to +3 st and every note comes out three keys higher while the file keeps exactly its original length; move it to −12 st and everything drops a full octave, still at the original tempo. The secondary control is playback speed, from a quarter of the original speed to four times it, and it works the other way around: it changes tempo and duration while the pitch stays wherever you set it. Keeping the original speed is simply leaving the speed at its default 1×, which the readout confirms by reporting the duration as 100% of the original. Under the hood the processing runs on Signalsmith Stretch, an open-source, studio-grade time-stretch and pitch-shift engine compiled to WebAssembly and running entirely inside your browser. It analyzes the audio spectrally in overlapping blocks and rebuilds it at the new pitch and tempo, which keeps transients tight and voices natural far beyond what simple granular methods manage — vocals shifted a few semitones stay intelligible, and music slowed to half speed stays clean enough to practice against. If the WASM engine cannot load — an old browser, a strict content-security policy — the tool automatically falls back to a built-in WSOLA granular stretcher and tells you so in the result: the engine readout shows either Signalsmith (studio quality) or WSOLA (fallback), so you always know what produced the file you are downloading. The classic uses split neatly along the two controls. Pitch only: transpose a backing track into a singable key, fix a recording that is a semitone flat, deepen or raise a voice without the chipmunk speed-up. Speed only: slow a solo to 0.5× to learn it in the original key, listen to a lecture at 1.5× without the narrator turning cartoonish, gently slow fast speech to 0.75× for transcription. And because the controls are independent you can combine them freely — up two semitones and 10 percent faster in one pass, with no intermediate render. The math is shown live: output duration equals input duration divided by speed, and the pitch factor is 2 raised to the power of semitones divided by 12. Files longer than ten minutes are refused so the render stays responsive; trim first, then process. Everything — decoding, stretching, and encoding to WAV or 192 kbps MP3 — happens locally, so nothing is uploaded and nothing waits in a queue.

Common pitch and speed recipes

Set pitch and speed by task — the readout shows the resulting duration before you process, and the engine line after processing shows whether the studio-grade WASM engine or the fallback produced the file.

GoalSettingsWhy
Transpose a backing track into your vocal range+3 st, speed 1×, WAVThe pitch moves three semitones up while the duration stays at 100%, so the track still lines up with any click or lyric sheet.
Deepen a voice-over without slowing it down−4 st, speed 1×Duration-preserving pitch shift avoids the sluggish drawl of tape-style slowing; the pacing is untouched.
Practice a fast guitar solo in the original key0 st, 0.5×, WAVHalf speed keeps the solo in the same key so you can play along; WAV avoids stacking lossy artifacts if you re-export.
Get through a 60-minute lecture faster0 st, 1.5×, MP3 — 192 kbpsThe hour becomes 40 minutes and the voice stays natural — pitch is untouched, so the narrator sounds brisk, not cartoonish.
Nightcore-style edit+4 st and 1.25× togetherBoth controls in one pass: the pitch lift and the tempo lift are set separately, so you can tune each to taste instead of accepting a coupled resample.
Down-tune a song from E standard to D standard−2 st, speed 1×Two semitones down matches a whole-step down-tuning, and the runtime stays identical for play-along practice.

How to change pitch and speed in your browser

  1. Load an audio file up to 10 minutes long — anything your browser can decode: MP3, M4A/AAC, OGG/Opus, FLAC, WAV, or WebM. Its duration, sample rate, and channel count appear once decoded.
  2. Set the pitch shift with the semitone slider or the −1 st and +1 st buttons, from −12 to +12. Leave it at 0 st to keep the original pitch; shifting never changes the length.
  3. Pick a playback speed from 0.25× to 4×, or leave the default 1× — keep original speed. The readout immediately shows the new duration as a percentage of the original, plus this file's exact output length.
  4. Choose WAV or MP3 — 192 kbps and press Process audio. Processing runs locally on the Signalsmith Stretch WASM engine, falling back to the built-in WSOLA stretcher if the engine cannot load; the result panel reports which one ran.
  5. Preview the result in the player and download the file.

Pitch and speed changer FAQ

Is my audio uploaded to a server?
No. The file is decoded with the browser's built-in codecs, processed by the Signalsmith Stretch WebAssembly engine running in the page (or the built-in JavaScript fallback), and re-encoded locally. Nothing is transmitted, and the tool keeps working offline once loaded.
Are pitch and speed really independent?
Yes. Shifting pitch by any number of semitones keeps the duration at exactly 100% of the original, and changing speed keeps every note at its original pitch. You can also combine them — the engine applies both in a single pass. To keep the original speed while changing pitch, just leave the speed at its default 1×.
What engine does the processing, and what is the WSOLA fallback?
The primary engine is Signalsmith Stretch, an open-source studio-grade time-stretch and pitch-shift library compiled to WebAssembly and driven through an offline audio graph. If it cannot load — very old browsers, or WASM blocked by policy — the tool falls back to a built-in WSOLA granular stretcher written in plain JavaScript. The result panel always states which engine produced the output.
How does the speed setting affect the file's length?
Output duration is input duration divided by speed. At 2× a 10:00 file becomes 5:00 (50% of the original); at 0.5× it becomes 20:00 (200%). The pitch setting has no effect on length. The readout computes this live, and shows the exact output length once a file is loaded.
Why is there a 10-minute limit?
Offline stretching renders the whole file in one pass, and both memory and processing time grow with length — especially at slow speeds, where the output is longer than the input. Beyond about ten minutes the wait and memory use get unreasonable in a browser tab, so longer files are refused; trim the file first or process it in sections.
How large a pitch shift sounds natural?
A few semitones in either direction is essentially transparent on the Signalsmith engine. Past about ±7 st any duration-preserving shifter starts to color the timbre — voices take on a slightly synthetic quality because the formants move with the pitch. That is inherent to the effect, not a defect; for creative extremes it is exactly the sound people want.
Which export format should I pick?
WAV if the result feeds into an editor or will be processed again — it adds no further loss. MP3 at 192 kbps for listening and sharing — it is far smaller and transparent for most material. Remember each MP3 export is one lossy generation.