Reverse audio recorder

A reverse audio recorder is two steps in one tab: capture (or open) a sound, then play it from the last sample to the first. This page is the general version of that job. It is not limited to speech. If you pointed a phone at a kettle, a synth, or a hallway, this is the page. If you only care about spoken memos and chat notes, the tighter guide is reverse voice recorder.

There is no timeline with regions, no punch-in, and no metronome. Think of a field recorder that has a reverse playback switch, implemented in Web Audio.

Record in the tab

Press Record. The browser asks for microphone permission for this site. Grant it or the button cannot capture. Permission is per origin; blocking it in Chrome site settings will keep failing until you reset that toggle. More on mics, monitoring, and permission UX: reverse audio microphone.

The take stops at 60 seconds. That cap is intentional. A reverse recorder in a browser tab is a sketch tool, not a two-hour interview deck. For a longer capture, use Voice Memos, a Zoom H-series, a DAW, or the phone’s stock recorder, then upload the file here.

While recording, watch input level in whatever UI the page shows. Clipping at the analog stage cannot be un-reversed into clean audio. If you are capturing a loud amp, back the phone up or lower the preamp.

Stop, then play. Default playback is reversed. The original end of the take is now the attack. A door slam becomes a rising suck; a sung note’s decay becomes a swell.

Toggle Normal to hear the take as you performed it. That A/B is the reason to keep the buffer in the tab instead of immediately downloading. Speed controls are preview only: they do not change the WAV duration.

Upload when the take already exists

Supported: MP3, WAV, FLAC, OGG, M4A, AAC. A Zoom WAV, an iPhone M4A, a WhatsApp OGG, a desktop FLAC bounce—all go through the same decoder. If the browser cannot decode a proprietary voice-file wrapper, bounce to WAV first.

Download is always WAV. Sample rate and channel count follow the decoded buffer. Local processing: the recorder buffer is not posted to a server.

This is still one file at a time. A session of twenty foley takes means twenty uploads or twenty mic takes. Reverse audio batch states that limit without a fake queue.

General recorder versus voice-specific pages

Job	Use
Speech, memos, chat	Voice recorder reverse, Voice Memo
Any mic source, SFX, instruments	This page
Mic hardware, permission, headphones	Microphone
Installable app store listing	There is no required app; audio reverse app explains the website-as-app pattern

The DSP is identical. The difference is which mistakes the copy warns you about. Here: watch for guitar-amp bleed, handling noise, and stereo files that reverse per channel (left stays left; this is not an L/R swap). Flip audio clarifies time flip versus stereo flip.

After you have a reversed WAV

Drop it on a video timeline, attach it in chat, or import it into a DAW as a one-shot. If you need the forward version again, Normal mode still works until you close the tab. Later, upload the reversed WAV and reverse once more; edge fades mean the round trip is not bit-perfect (unreverse).

Do not expect the download to include your preview speed, a limiter, or a fade you did not hear as a designed fade. The engine applies a short edge fade to reduce clicks from hard trims. That is not an artistic fade-out.

Phone path: reverse audio on mobile. Desktop path for large WAVs: reverse WAV online.

What this recorder will not do

• Multi-take comping
• MIDI
• Pitch correction
• Polarity invert (time reverse only)
• Background recording with the tab closed
• Cloud backup of takes

If you need those, use a DAW and, if you still want reverse, either reverse here per bounce or use the DAW’s reverse region command. This page stays a single-buffer reverse recorder on purpose.

Instruments, rooms, and naming

A guitar DI reversed is a swell into a pick attack. A room mic reversed puts the room before the source, which is the pre-verb trick if you bounce with reverb already on—usually you do not want that on a dry practice take.

Name files with _fwd and _rev as soon as they hit disk. The recorder UI will not do it. A week later you will not remember which WAV is which.

Clap / stick / metronome: reverse is rarely useful except as a gag. Record those in a DAW with a click; this 60-second cap is the wrong tool for a take that must hit grid.

When to bounce from another recorder

If you already recorded in Voice Memos, do not re-record on this page through speakers. Upload the memo. Re-recording is a generation of air and clock.

If the take is video, extract audio; this recorder will not strip a camera file reliably. See the video page.

The voice-specific recorder article spends more time on speech exports from chat apps. This page is the same button when the source is a kettle, a synth, or a hallway. Do not look for a different DSP behind the two URLs.