Reverse Audio

Turn a recording back to front, with the last sample first and the length unchanged.

Runs on your deviceAudioWorkingNo account needed

What this tool does

Reverse Audio decodes your file to raw samples, walks every channel from the last sample to the first, and writes the result out again. That is the whole operation, and on the samples themselves it is exact: nothing is scaled, resampled, faded or trimmed, so the length, the peak level and every value in between survive the trip, and a second pass hands back the numbers you started with. The work is done by the tab you already have open, reading the file from disk and handing the reversed copy straight back.

What the ear notices is not the order of the samples but the shape of every envelope in the file. A note that was struck and left to decay now swells out of nothing and stops dead; the fade-out at the end of a track becomes a fade-in at the start; a room's reverb tail arrives before the sound that caused it. Nothing is added to soften any of that, because no fade is applied at either edge — so a take that was cut mid-waveform still begins with the same click, only now at the other end.

Output is WAV or MP3, chosen with the same three-way setting the other audio tools use. Match the input where possible keeps an MP3 as an MP3 and sends everything else — FLAC, M4A, OGG, WAV — to WAV, because no browser exposes an AAC, FLAC or OGG encoder to write with. The bitrate list runs from 64 to 320 kbps and starts at 192; the WAV list offers 16, 24 or 32-bit float and starts at 16. Both lists stay on screen while the default output setting is in force, since it can still end either way, and picking WAV or MP3 outright hides the one that no longer applies. Neither list touches the reversal, which happens before any encoder sees the samples.

Three limits are worth knowing first. The reversal covers the whole file and nothing narrower, so isolating a passage means cutting it out with Trim Audio and reversing that. ID3 tags and cover art do not survive, because the output is written from samples rather than copied from the original. And the reversed copy is built beside the decoded original rather than in place, so at the peak of the job both are held as 32-bit floats — eight bytes per sample per channel between them, which is why a long stereo master, rather than a voice memo, is the job that will make a modest machine work.

How to use it

  1. Add your audio. MP3, WAV, M4A, AAC, FLAC, OGG and Opus are offered in the picker and read as far as your browser's own decoder reaches; WAV is parsed directly without the audio engine. Queue several files and each is reversed in turn.
  2. Choose the output format. Match the input where possible, or force WAV or MP3. While the default is in force both the bitrate and the bit-depth list are shown, because either could apply; naming a format leaves only the list that matters. Neither has any bearing on the reversal itself.
  3. Run it, then listen before you keep it. The summary reports the duration, the format and the settings used. A reversed take usually wants trimming at what has become its start, since the old ending is now the first thing anyone hears.
  4. Download the file, or the whole batch. Each result is named after its source with -reversed on the end, and a queue of several can be taken away together as a single archive.

Processing & privacy

Runs on your device. Your file is processed directly in this browser.

Common use cases

Working out what a backwards clip says

Speech recorded or played in reverse becomes intelligible again once the samples are put back in order, which is the usual reason for reaching for this at all.

Building a reverse cymbal or a riser

A crash, a piano chord or a reverb tail played backwards swells rather than decays, which is the standard way to lead into a downbeat.

Putting an effect in front of the sound

Reverse a file, apply a delay or a reverb elsewhere, then reverse it back, and the tail that would trail behind a note now arrives ahead of it instead.

Checking the end of a long recording

Reversed, the last few seconds are the first you hear, so a truncated word or an abrupt cut at the end of an interview shows up immediately instead of after an hour of playback.

Questions

Is my audio uploaded?
No. Decoding, reversing and encoding all run inside this browser tab, and no part of the file is transmitted anywhere. One caveat if you are working offline: the MP3 encoder is a separate script fetched from this site the first time an MP3 is written, so a tab that lost its connection before ever writing one can still produce WAV, but not MP3.
Does reversing cost any quality?
The reversal does not. Every sample is kept and only its position changes, so the peak level and duration come back untouched. Writing MP3 afterwards is a separate matter and costs one generation of lossy encoding — choose WAV when the file is going on into an editor.
Can I reverse only part of a file?
Not here. Reverse Audio turns the whole recording around. Cut the section you want with Trim Audio first, reverse that, and put it back in place with Merge Audio if it has to sit inside the original.
If I reverse it twice, do I get the original back?
The samples, yes — the second pass restores them exactly, and that is one of the checks the tool runs against itself. The file, only if you write WAV both times, since two trips through the MP3 encoder leave audio that has been re-encoded twice. The tags are gone either way.
What format does the result come back as?
WAV or MP3, and nothing else. Match the input where possible keeps MP3 as MP3 and sends everything else to WAV, so a FLAC or M4A input returns a considerably larger WAV file. There is no third option to offer, because browsers ship no encoder for the other formats.
Do the title, artist and cover art survive?
No. The output is built from the decoded samples alone, so it arrives with no ID3 tags and no artwork, and a player will fall back to the filename. Read what the original holds with Audio Metadata first if you mean to type them back in.
What happens to a multichannel or odd-rate file?
WAV output keeps every channel and the original sample rate as they were. MP3 cannot: anything beyond two channels is folded to stereo, and a sample rate the format does not define is moved to the nearest one it does. A surround mix reversed to MP3 therefore comes back as a stereo fold.
Which files can it open?
Whatever your browser can decode. MP3, WAV, M4A and AAC work more or less everywhere, FLAC and OGG on most browsers, AIFF only in Safari, and WMA on none of them. A file that will not decode is named back to you with the reason: WMA and AIFF are told which browsers can and cannot read them, OGG and Opus that support varies, and anything else gets the list of formats worth trying instead.

Related tools