Voice Changer

Apply a chipmunk, deep, robot, echo or telephone effect to a recording in your browser. Pitch shifts keep the original speed. Nothing is uploaded.

🌐 Español

Drop your file here (.mp3, .wav, .m4a, .ogg, .flac)

🔒 Private by design: your files are processed locally in your browser and never uploaded to any server.

Written and maintained by Last updated

Five effects, each doing something specific

A voice changer with a slider labelled from robot to alien is fun for a minute and useless for anything else. These five presets each reproduce a recognisable, describable effect, and knowing what each one actually does makes it obvious which to pick.

Two of them shift pitch, one up and one down. One modulates the signal to make it metallic. One adds a repeat. One reproduces the frequency limits of a telephone line. There is no free parameter on any of them, because a preset that needs tuning is not a preset.

Applying an effect

  1. Drop an MP3, WAV, M4A, OGG or FLAC onto the box. One file at a time.
  2. Choose a Voice effect.
  3. Run it and download the result, an MP3 named after your file with the effect appended.

The two pitch presets read the source’s real sample rate first, since their maths depends on it. The other three skip that step and start immediately.

Pitch without speed, and why it matters

Change a record player’s speed and the pitch goes up with it. That is the whole relationship, and it is why the simplest pitch shift also shortens the recording and makes the speech faster.

For a comedy chipmunk voice that might be acceptable. For anything where the words matter it is not, because a voice sped up by a third is much harder to follow.

The two pitch presets here therefore do the shift in two stages. The playback rate is changed, which moves pitch and speed together, and then a tempo correction is applied that restores the original speed while leaving the pitch where it was moved to. The result is a voice at a different pitch speaking at exactly the original pace, occupying exactly the original duration.

If you want the pitch shift with a value you choose rather than one of these two presets, Audio Pitch Changer exposes it directly.

The robot is modulation, not a vocoder

There are two very different things people mean by a robot voice.

A vocoder imposes the frequency envelope of a voice onto a synthesised carrier, which is the classic musical robot and requires real analysis and resynthesis.

This is the other kind, and it is much simpler. The pitch is wobbled 18 times a second and the amplitude 45 times a second. Both rates sit well above the range where modulation is perceived as a wobble, so rather than hearing the modulation you hear its effect on the timbre, which is a hard metallic buzz.

Calling it a modulation robot rather than a vocoder is the accurate description, and it is worth knowing if you were expecting the other thing.

The telephone effect reproduces a real limit

Analogue phone systems carried a band from roughly 300 to 3400 hertz. That was not a stylistic decision, it was what fitted in the available bandwidth, and everything outside it was simply gone.

Removing the bottom takes away the chest resonance that makes a voice sound present and close. Removing the top takes away the sibilance and air that make it sound detailed. What is left is the middle, which carries almost all the intelligibility and none of the presence, which is exactly why phone calls were perfectly understandable and never sounded like the person was in the room.

The mild compression after the filtering flattens the dynamics the same way a phone line did, so a shouted word and a quiet one arrive at similar levels.

To make the recording in the first place, Voice Recorder captures your microphone directly to a file. To clean up a noisy source before applying an effect, Remove Audio Noise is worth a pass first, since these effects will happily process hiss along with the voice.

To trim the result down to the part you want, Cut MP3 handles that. For a spatial effect rather than a character one, 8D Audio moves the sound around the stereo field. The rest are on the audio hub.

See it in action

Screenshot of the Voice Changer tool with sysfenix-sample.wav (517 KB) loaded, Voice effect set to Chipmunk, higher pitch
Voice Changer mid-process: sysfenix-sample.wav (517 KB) loaded, Voice effect set to Chipmunk, higher pitch.
Screenshot of the Voice Changer result screen showing sysfenix-sample-chipmunk.mp3 ready to download (72 KB, 86% smaller)
The finished result: sysfenix-sample-chipmunk.mp3 ready to download (72 KB, 86% smaller). The download link is a local blob URL — the file never leaves your device.

Frequently asked questions

Do the pitch presets also speed the recording up?

No, and that is the point of how they are built. The naive way to raise pitch is to play the audio faster, which raises the pitch and shortens the recording together. Here the speed change is applied and then undone by a compensating tempo adjustment, so the pitch stays shifted while the duration and the pace of speech are exactly what they were.

How much do the two pitch presets shift by?

Chipmunk multiplies the pitch by 1.35, which is around five semitones up, high enough to be clearly cartoonish while leaving the words intelligible. Deep multiplies by 0.8, roughly four semitones down, which is the film trailer or villain register. Neither is adjustable, since the point of a preset is to be a single decision.

What is the robot effect actually doing?

Two fast modulations in sequence. First the pitch wobbles at 18 times a second, then the volume wobbles at 45 times a second. Both rates are far above the few hertz where modulation is heard as a musical wobble, so instead of warbling they fuse into a metallic buzzing timbre. It is a modulation robot rather than a studio vocoder, which is a different and much more complex effect this does not attempt.

What kind of echo is it?

A single slap back repeat at 250 milliseconds with a soft decay, not a reverb tail. That produces one distinct repetition of each sound rather than the wash of reflections a room gives you. Slap back is the classic effect on early rock vocals and it is far more intelligible than reverb, which is why it suits a voice.

Why does the telephone effect sound so specific?

Because it reproduces a real constraint rather than approximating a vibe. Analogue telephone lines carried roughly 300 to 3400 hertz and nothing outside it, so the effect removes everything below 300 and above 3400, then applies a mild compression to flatten the dynamics the way a phone line did. Losing both the low warmth and the high detail is exactly what made phone calls sound like phone calls.

What format is the processed file?

Always an MP3 at 192 kbps, named after your file with the effect name appended. Every one of these effects rebuilds the audio samples, so a re-encode is unavoidable, and rather than guess a format the tool standardises on MP3 at a rate that is transparent for speech. Any embedded cover art in the source is dropped.

Can I combine two effects?

Not in one pass, since the presets are exclusive. You can run the output back through the tool with a different preset, and it works, but be aware that each pass is another lossy encode and the artefacts accumulate. Two passes is usually fine, four is not.

Related tools