How to change your voice
- Press Record and allow microphone access, say something, then press Stop. Or press Open audio file (you can also drop a file on the box) to use a recording you already have — a voice memo, a WhatsApp voice message saved to your device, a clip from a video.
- Choose an Effect. The tool renders the new version within a second or two and puts it at the top of the list with a player; the original stays below it so you can compare.
- Adjust it: Extra pitch shift moves the voice up or down in semitones on top of the effect, Speed makes it faster or slower without changing the pitch. Each change renders a fresh version.
- Press Download WAV to save the version you like. The file is named after the original and the effect, for example voice-2026-10-15_14-03-22-chipmunk.wav.
The settings are remembered for your next visit, but the recording is not: it lives in this tab only, so download what you want to keep.
The effects
| Effect | What it does |
|---|---|
| Chipmunk / helium | Raises the pitch by 8 semitones at the same speed. Because the whole spectrum moves up, the voice sounds small and cartoonish. |
| Deep voice | Lowers the pitch by 5 semitones — a darker, heavier voice that still sounds fairly human. |
| Monster | Down 9 semitones, high frequencies softened and a little distortion added for a growl. |
| Robot | A channel vocoder: the loudness of your voice in 16 frequency bands shapes a buzz at a fixed 100 Hz, so the words stay clear but the melody of speech disappears — the classic monotone machine voice. |
| Alien | Up 3 semitones, a fast 6 Hz wobble and ring modulation at 400 Hz, which adds metallic, non-musical overtones. |
| Echo (cave) | Repeats every quarter of a second, each repeat at 45% of the one before, until it fades below −60 dB — which adds about 2¼ seconds of tail to the clip. |
| Telephone / old radio | Keeps only 300–3,400 Hz, the band of a classic landline call, and overdrives it slightly like a small speaker. |
| Whisper | The same vocoder as the robot, but driven by noise instead of a buzz: your words come out breathy and toneless, like a loud whisper. |
| Reverse | Plays the recording backwards. Combine it with any pitch or speed. |
Pitch, semitones and speed
Pitch is measured here in semitones, the steps between neighboring piano keys. Each semitone multiplies every frequency by the twelfth root of two (about 1.0595), so 12 semitones — an octave — double it: frequency ratio = 2^(semitones ÷ 12). Typical speaking voices sit around 85–155 Hz for adult men and 165–255 Hz for adult women (R. J. Baken, Clinical Measurement of Speech and Voice). The table shows where a voice at 120 Hz and one at 210 Hz end up:
| Semitones | Ratio | 120 Hz voice | 210 Hz voice |
|---|---|---|---|
| +12 | 2 | 240 Hz | 420 Hz |
| +8 | 1.587 | 190 Hz | 333 Hz |
| +5 | 1.335 | 160 Hz | 280 Hz |
| +3 | 1.189 | 143 Hz | 250 Hz |
| +1 | 1.059 | 127 Hz | 222 Hz |
| −3 | 0.841 | 101 Hz | 177 Hz |
| −5 | 0.749 | 90 Hz | 157 Hz |
| −9 | 0.595 | 71 Hz | 125 Hz |
| −12 | 0.5 | 60 Hz | 105 Hz |
Worked example: the chipmunk effect is +8 semitones, a ratio of 2^(8 ÷ 12) ≈ 1.587. Add an extra +4 and the total is +12: a 120 Hz voice now speaks at 240 Hz, still at its original speed. Set Speed to 1.25 as well and a 60-second clip lasts 60 ÷ 1.25 = 48 seconds.
How the pitch changes without changing the speed
Playing a tape faster raises the pitch and shortens the recording at the same time. To change only the pitch, the tool first stretches the sound in time, then resamples it. The stretch uses WSOLA (waveform-similarity overlap-add): it cuts the sound into overlapping 23-millisecond slices, lays them out at the new spacing, and nudges each slice by up to 6 ms to where its waveform lines up best with the slice before, so the joins don’t click or warble. Resampling then shortens the stretched sound back to its original length, which multiplies every frequency by the chosen ratio. Speed works the same way with a different stretch.
Be aware of the limits: this method keeps a single voice clear for shifts of a few semitones, but large shifts sound increasingly artificial, and music with several instruments can smear. It also moves the formants — the resonances of your throat and mouth — together with the pitch, which is exactly why +8 sounds like a chipmunk and not like a different adult. Turning a man’s voice convincingly into a woman’s (or the reverse) needs formant control or an AI voice model, which this tool doesn’t pretend to offer.
Live voice changing, Discord and phone calls
This is an offline voice changer: it records first and transforms afterwards. It does not change your voice live in Discord, games or phone calls — that requires an app installed on your device that creates a virtual microphone. What it does well is the fast path from a clip to a funny or disguised version you can send as a file: record, choose, download, share.
File format
The result is a 16-bit mono WAV file at 44,100 samples per second — uncompressed, about 5.3 MB per minute, and readable by every phone, computer and audio editor. To send it in a chat app as a smaller file, convert it to MP3 or M4A in an audio app, or with FFmpeg: ffmpeg -i voice-chipmunk.wav -q:a 4 voice-chipmunk.mp3. Files up to 5 minutes are processed; longer ones are cut after 5 minutes.
Privacy
Your recording, the file you open and the result never leave your device: decoding, effects and WAV encoding all run in this browser tab, and nothing is sent to a server. Close the tab and they are gone. Need a plain recording instead? Use the online voice recorder, and if the microphone stays silent, check it with the mic test.
Frequently asked questions
Is this voice changer free and private?
Yes. There is no sign-up and no upload: the recording or file is decoded, changed and saved as WAV inside your browser, and it disappears when you close the tab.
Can I use it live on Discord, in games or on calls?
No. It changes recordings, not a live microphone. Live voice changing needs an installed app that adds a virtual microphone to your system. Here you record (or open a file), apply the effect and download the result.
Which audio files can I open?
Whatever your browser can decode: MP3, WAV, M4A/AAC and OGG almost everywhere, WebM/Opus and FLAC in most current browsers, and the soundtrack of many video files. Stereo is mixed to mono, and only the first 5 minutes are used.
Why does a big pitch change sound robotic or watery?
Changing the pitch while keeping the speed means stretching the sound and gluing short slices back together. For a few semitones the joins are hard to hear; beyond about an octave they become audible. For the cleanest result, keep the shift small or let speed and pitch change together by setting the speed close to the frequency ratio shown on the card.
Can it make me sound like a woman, a man or a celebrity?
Not convincingly. Moving the pitch up or down also moves the resonances of your voice, so large shifts sound cartoonish rather than like another person. Sounding like a specific person needs an AI voice model, and imitating a real person without consent can be illegal — this tool is meant for fun effects and for disguising your own voice.
Why WAV and not MP3?
WAV can be written exactly in the browser with no extra software, and it keeps full quality. It is larger — about 5.3 MB per minute — so convert it to MP3 or M4A with an audio app if you need a smaller file.
Last updated: October 15, 2026