How to use Audio Speed Changer
- 1Drop in an audio file
- 2Pick a speed preset, or type any factor
- 3Decide whether to keep the original pitch
- 4Play the preview to check it sounds right
- 5Export and download
What Audio Speed Changer does
Pitch and tempo are bound together in the physical world. Play a record faster and everything rises in pitch, because you are compressing the same waveform into less time and every frequency in it goes up proportionally. That is one operation, not two, and for most of recording history it was the only one available.
Separating them requires rebuilding the sound rather than replaying it faster: the audio is cut into short overlapping grains which are then repositioned in time and blended back together. The frequencies inside each grain are untouched, so pitch stays put while the timing changes. That is what makes a lecture at 1.5× still sound like the person speaking.
When to use it
- Listening to a lecture or podcast faster
- 1.5× is the usual sweet spot — substantially quicker with speech still entirely natural. Exporting at that speed means it plays that way in any player.
- Slowing down to transcribe
- 0.75× or 0.5× makes fast or unclear speech far easier to catch, and with pitch preserved the voice stays recognisable rather than becoming a drawl.
- Learning a piece of music
- Slowing a passage without changing its pitch keeps the notes correct while giving you time to follow them — the reason this feature exists in practice tools.
- Fitting a fixed duration
- A recording slightly too long for a slot can be nudged to 1.05× or 1.1×, which is imperceptible and avoids cutting content.
Audio Speed Changer settings explained
- Speed
- Any factor. Above 1 is faster and shorter, below 1 is slower and longer. 2 halves the duration, 0.5 doubles it. Presets cover the values people reach for most.
- Keep pitch
- When on, the tempo changes and the pitch does not — voices stay recognisable at any speed. When off, the audio is resampled like a record played fast, and pitch rises or falls with the tempo.
- Presets
- Common factors in one click, so the usual listening speeds do not need typing. Any value can still be entered directly.
- Preview before export
- The retimed audio plays in the tab before you commit to it, which matters because the artefacts of time-stretching are much easier to hear than to predict.
Keep pitch on or off
| Keep pitch on | Keep pitch off | |
|---|---|---|
| What changes | Tempo only | Tempo and pitch together |
| A voice at 2× | Same voice, faster | Chipmunk |
| Processing | Grains repositioned and blended | Simple resampling |
| Artefacts | Possible at extremes | None — but the pitch has moved |
Limits and what to expect
- Beyond about 2× or below 0.5×, time-stretching artefacts become audible even with pitch preserved.
- Music is less forgiving than speech: sustained notes and reverb tails show stretching first.
- Extreme factors are applied as a chain of smaller steps, which compounds any artefacts.
- 150 MB per file, and long recordings decode slowly because samples are held uncompressed.
- Exporting to a lossy format re-encodes, so the result is a second generation unless you choose WAV or FLAC.
Troubleshooting
- The voice sounds unnaturally high or low
- Keep pitch is off, so the audio is being resampled like a record played at the wrong speed. Turn it on to change only the tempo — that is almost always what you want for speech.
- It sounds watery or metallic at high speed
- Those are time-stretching artefacts, and they grow past about 2×. Preserving pitch means rebuilding the sound from overlapping grains, and the blending becomes audible when the grains are pushed far apart. Try a lower factor.
- Music sounds worse than speech at the same setting
- Speech is full of natural gaps that hide the seams; sustained notes and reverb tails have none. Music generally tolerates about half the speed change speech does before artefacts show.
- The exported file is bigger than the original
- You slowed the audio down, so there is more of it, or you exported to a lossless format. Both are expected — check the duration and the chosen format before assuming something went wrong.
Your files and your privacy
Decoding, retiming and export all happen inside your browser using your device's own audio engine. Nothing is uploaded, including during preview playback — what you hear is played from the in-memory buffer being edited, not streamed from a server.
What to use next
FAQ
How do I speed up audio without the chipmunk effect?
Turn on keep pitch. The audio is then rebuilt from short overlapping grains repositioned in time, so the tempo changes while the frequencies inside stay put. Without it, the audio is simply resampled and the pitch rises with the speed.
What speed should I use for a podcast or lecture?
1.5× is where most people settle — noticeably quicker while speech remains entirely natural. 2× is followable for a clear speaker. Past that, comprehension falls off faster than the time saved is worth.
Can I slow audio down to transcribe it?
Yes. 0.75× or 0.5× with pitch preserved makes fast speech much easier to catch while keeping the voice recognisable. Below 0.5× the stretching artefacts start to interfere.
Why does it sound strange at very high speeds?
Preserving pitch means cutting the audio into grains and repositioning them. Push them far enough apart and the blending between them becomes audible as a watery or metallic quality. Speech tolerates this better than music, which has fewer natural gaps to hide the seams.
Does changing speed reduce quality?
The retiming itself is done at full precision. What costs quality is the export format — WAV or FLAC preserve the result exactly, MP3 and AAC re-encode it. At extreme speed factors the stretching artefacts are a separate matter from compression.
Is my audio uploaded?
No. Everything runs in your browser on your own device, with nothing transmitted at any point.