Reverse audio search

People type “reverse audio search” when they have a clip they cannot place. Often the clip is already backwards: a TikTok, a WhatsApp voice note, a puzzle, or a track that was exported reversed. This page is the first step of that workflow. It does not crawl a fingerprint database. It restores forward playback so a human ear, a lyric search, or a song ID app has something usable to work with.

What reverse audio search actually means

Reverse image search compares a photo to an index of images. There is no equivalent public Google tool for sound. Music identification products such as Shazam compare a fingerprint of the forward mix against commercial catalogues. They fail on reversed audio because the fingerprint no longer matches.

If your file plays backwards, reverse it here first. If your file already plays forwards and you still cannot name it, skip reversal and go straight to a song ID app, a hummed-search feature, or a lyric fragment in a search engine.

How to reverse a clip before you try to identify it

1. Upload the file, or record up to 60 seconds from the microphone.
2. Wait until the waveform is processed in the browser.
3. Play the reversed result. If the source was already backwards, this pass should sound like normal speech or music.
4. Switch between Play reversed and Use original to compare both directions.
5. Download the WAV if you want to run it through a phone app.

Use 0.75x or 0.5x playback when speech is fast. Use 1.25x when you are scanning a long music bed for a hook.

When this workflow helps

• A meme or challenge posted the audio backwards on purpose.
• Someone sent a reversed voice note as a joke.
• You found a “secret message” clip and need the original syllable order.
• A sample in a project was reversed and you need to hear the source phrase.

It does not help when the recording is too short, too noisy, or covered by a loud voiceover. Identification still needs a clear stretch of the original mix.

What to do after you have a forward clip

For music: play the restored WAV next to your phone and run a song ID app, or paste a distinct lyric into a search engine. For speech: listen once at 0.75x and write down unique names or phrases. For sound effects: reverse-searching a whoosh or riser rarely returns a product name; you are listening for a library you already own.

If the clip came from YouTube, extract the audio on your computer first. This site does not pull streams from YouTube. The YouTube audio reverser page explains that workflow.

Limits of this tool

Processing stays on your device. There is no account, no watermark, and no server-side matcher. Recording stops at 60 seconds. Downloads are WAV, which is large but lossless enough for a second app to fingerprint. Very large files can stall on older phones; keep uploads under about 100 MB.

If you already know the track and only want it backwards for an edit, you do not need this search workflow. Use the homepage reverser and export.

Common mistakes

Uploading a screen recording of a reversed TikTok that also contains UI sounds. The ID app then matches a notification chirp or a creator’s voice instead of the song. Crop to the music bed before you reverse, or at least start playback after the voiceover.

Expecting a 1.5 second sting to match. Fingerprint apps need a few seconds of relatively clean signal. If you only have a stab, reverse it for your own ears and search a lyric you can make out, rather than forcing a match.

Running identification on the still-reversed file. That is the usual reason people say “Shazam cannot find this.” Reverse first, confirm it sounds like music in the normal direction, then identify.

Using a Bluetooth speaker pointed at the phone mic in a noisy room. Play the restored WAV on the same device with headphones, or AirDrop/share the file into the ID app if it accepts uploads.

Privacy during a search workflow

The reverse step never leaves the tab. The identification step is whatever you choose next. A phone-based song ID app typically sends a fingerprint to that vendor. A lyric typed into a search engine is a text query. Do not upload private voice notes to a public “audio search” website if the content is sensitive; restore direction here and listen yourself.

Browser notes

Chrome and Edge decode the widest set of files. Safari on iOS can be picky with some OGGs; convert those to M4A or WAV in Files, then upload. Firefox works for WAV and MP3. If the page sits on “Processing audio…” for more than a minute, the file is too large for the device RAM. Split it or downsample before retrying.

Closing the tab drops the buffer. There is no cloud history of clips you reversed. Download before you navigate away if you still need the WAV.

Related search pages

Google-specific confusion is covered on Google reverse audio search. Songs as a class are on reverse music search. Speech is on reverse voice search. A side-by-side of methods is on reverse audio search engine.