Voice reverser

A voice reverser is the same audio engine as the homepage, aimed at speech. Upload a memo, a call export, or a clip of yourself, or record up to a minute with the microphone. The page plays the buffer from the last sample to the first. That is all. There is no cartoon character, no gender slider, and no language translation.

Why reverse a voice on purpose

Producers use reversed vowels as risers into a chorus. Podcasters hide a swear by reversing a syllable instead of bleeping. Teachers use backwards playback as a listening exercise. Puzzle makers record a sentence, reverse it, and ask students to decode it. None of those jobs need a DAW if you only need a quick WAV.

If you actually want a different identity (robot, chipmunk, lower pitch), that is a voice changer, not a reverser. The voice changer page explains the difference and still gives you the reverse engine for the reverse part of a chain.

Record or upload

Recording: click Record, allow the mic, speak, click Stop before 60 seconds. Chrome, Firefox, Safari, and Edge all prompt for permission. If permission is denied, the button stays in an error state until you fix it in site settings.

Upload: Voice Memos on iPhone usually export M4A. WhatsApp voice notes are often OGG or MP3. Android recorders vary. The decoder in the browser reads the common ones. If a file fails, convert it to WAV on the phone and try again.

Listening tips for reversed speech

Reversed English is hard to parse at 1x. Drop to 0.75x and listen for sibilants; they mark the original ends of words. Names and numbers are easier than full sentences. If the take has room echo, the reverb will run backwards and smear consonants. Record dry if you plan to reverse.

For a “say this backwards” challenge, write the sentence, record it normally, reverse here, and send the WAV. Do not try to pronounce reversed phonemes live unless that is the joke.

After you download

The file is WAV. iMessage, WhatsApp, and most editors accept it. If an app rejects WAV, convert with the phone’s files app or a desktop encoder. To put reversed speech on a story or Reel, import the WAV as a separate audio layer so you can trim to the picture.

Longer workflows: reverse a Voice Memo, record then reverse, or restore a clip with unreverse audio.

Backmasking versus this tool

Backmasking in vinyl-era records usually meant a phrase was recorded backwards on purpose so it would sound like speech when the disc was spun the wrong way. Playing a modern file backwards with this page will not reveal a hidden sentence that was never recorded. You will hear reversed phonemes, which brains love to turn into words. Treat “I hear a phrase” claims as ear-pareidolia unless you also have the original take.

If you are making a backmasked line, write the sentence, record it dry, reverse, and mix the reversed WAV under the track. Keep the dry take. Two reverses restore word order if you need to edit a syllable later.

Levels and clipping

Speech is peaky. Reversing does not add makeup gain, but the short edge fade can make the first millisecond quieter. If the take was already clipped, reversed clipping sounds like a click at the new start. Re-record instead of normalizing a clipped memo.

The meter in your editor after download is the same peak level as the decoded buffer, minus the fade. There is no limiter on export.

Classroom and call-in use

For a class: each student records the same sentence, reverses, and swaps files. Decoding at 0.75x is the exercise. Do not collect recordings on a shared drive if the school policy forbids it; the browser copy disappears when the tab closes unless they download.

For radio or podcasts: reverse a caller only if you have rights to air the call. The technical step is the same as any other voice file.

What this page will not do

It will not isolate a voice from music. It will not remove noise. It will not transcribe. It will not change gender or accent. Those are other products. Chain them yourself: denoise in an editor, reverse here, done.

See also reverse voice search if the job is “what was said,” and reverse speech generator if you meant “make reversed speech from a take,” which is still this engine plus your microphone. Keep the dry take next to the reversed WAV so you can recut a word without recording again.

Accents, languages, and singing

Reversed speech in a language you do not speak is even harder to parse, which is useful for a texture and useless for a transcript. If the job is translation, restore direction first, then use a human or a speech-to-text product you trust. This page will not detect language.

Sung vowels reverse into the same eerie pad producers already know. Consonants (t, k, p) become small reverse-swells. Harmonies stay in the original key. For a sung line as a puzzle, reverse singing is the better walkthrough; this page stays on spoken voice.