Video & Audio
Voice Changer (Chipmunk, Deep, Robot, Echo)
Record your voice or upload audio and apply fun effects — chipmunk, deep voice, robot, alien, telephone, echo, cathedral reverb — then download.
How to use
- 1Record with your mic or add an audio file.
- 2Pick an effect, preview it, and download MP3 or WAV.
Frequently asked questions
Is my recording uploaded?
No — recording and effects run in your browser.
Can I use it for videos?
Yes — download the processed audio and add it in your video editor.
Why does the robot effect sound metallic?
It uses ring modulation, the classic sci-fi robot trick, which multiplies your voice with a tone.
Why multiplying your voice by a 50Hz tone makes it sound robotic, specifically
Ring modulation works by multiplying your voice's waveform sample-by-sample with a pure sine wave (50Hz for the Robot effect) — and multiplying two signals in the time domain mathematically produces the sum and difference of their frequencies rather than either original frequency surviving on its own. That's precisely why the result sounds metallic and inharmonic rather than just 'low-pitched': ring modulation doesn't shift pitch cleanly the way the Chipmunk and Deep effects do (which use genuine pitch-shifting), it fractures the voice's natural harmonic relationships into a new, dissonant set of frequencies — the exact mechanism classic sci-fi robot voices (and plenty of Doctor Who-era sound design) have relied on for decades.
Why the Cathedral effect uses random noise instead of a recording of an actual cathedral
Rather than convolving your voice with a real impulse response recorded in a real stone cathedral, this effect synthesizes one: it generates 3.5 seconds of random noise in each channel, shaped by a decay curve ((1 − position/length)³) so the noise starts loud and tapers smoothly to silence, then uses that synthetic buffer as the impulse response for a convolution reverb node. A convolution reverb works by convolving dry audio with any impulse response buffer, real or synthetic — a true recorded cathedral impulse response would capture that specific room's exact reflections and resonant frequencies, while this synthetic one approximates the general character of a large reverberant space (long, dense, smoothly decaying) without modeling any particular room.
Why Telephone and Radio sound similarly 'thin' but aren't quite the same effect
Both effects chain a highpass filter, a lowpass filter, and a waveshaper (using a tanh curve to add soft clipping distortion) — but with different corner frequencies: Telephone narrows the band to 300-3400Hz (the actual bandwidth classic analog phone lines were engineered to carry) with gentler distortion, while Radio uses a tighter 500-2800Hz band with distortion driven twice as hard. The telephone band isn't an arbitrary choice — 300-3400Hz is the real standard telephony voice-frequency band, which is exactly why cutting audio to that specific range is what makes something sound like it's coming through a phone line rather than just 'muffled.'
Chipmunk is 7 semitones up, and Deep voice is 5 semitones down
Chipmunk and Deep voice move the pitch by a fixed musical interval and leave the length of the clip alone. Chipmunk is 7 semitones up, about one and a half times the original pitch. Deep voice is 5 semitones down, about three quarters of the original pitch. A 4-second take is still 4 seconds afterward. Alien uses a smaller lift, 3 semitones, about 19% higher, and only then adds its metallic tone. The buttons open on Chipmunk. Recording a new take, or dropping in a new file, replaces what you were editing, and the selected effect is applied to that new audio.
Echo repeats after 0.28 seconds, and cathedral stays mostly dry
Echo repeats the voice 0.28 seconds later and feeds 45% of each repeat back in, so the echoes get quieter instead of stacking at full volume. Two extra seconds are kept on the end so the last repeats are not cut off when the recording ends. Alien, after the 3-semitone lift, keeps 40% of your original voice and mixes in 60% of a 520 Hz metallic tone. Cathedral is quieter than the name suggests: only about 35% of what you hear is the reverb, the rest is still the dry voice, and 4 seconds are allowed for that 3.5-second wash to finish. Robot adds a 12-millisecond slap after its metallic tone, with half of the slap fed back. Switching effects shows Applying effect… while the new one is built.
Related tools
- Audio Speed Changer (Keep Pitch)
Speed up or slow down audio from 0.5× to 2× — keep the original pitch (time-stretch) or change it like a tape — for podcasts, lessons and practice.
- Audiogram / Waveform Video Maker
Turn audio into a shareable video with an animated waveform, title and cover image — perfect for promoting podcast clips on social media.
- Change Video Speed (Slow Motion / Fast Forward)
Speed up or slow down a video from 0.25× to 4× — with audio kept in sync and pitch preserved — for time-lapses and slow motion.
- Mute Video (Remove Audio)
Remove the sound from a video instantly — no re-encoding, no quality loss — and download the silent MP4.
- Pitch Changer / Key Shifter
Change the pitch or key of a song by semitones without changing its speed — transpose for singing, practice or tuning — then download.
- Reverse Audio
Play any audio backwards — reverse songs, voice messages and sound effects — and download the reversed file as MP3 or WAV.