Convert MP4 to WAV

An MP4 is a container holding a video track and an audio track side by side. This operation reads the audio, decodes it and writes it out as WAV. The video is not converted, not re-encoded and not degraded — it is simply not part of the output, and the original file is untouched.

That framing explains why it is fast even for a long recording. There is no video encoding, which is the expensive part of any video operation; decoding an hour of AAC and writing samples is comparatively trivial work.

The soundtrack inside an MP4 is almost always AAC, and AAC is lossy — the encoder discarded parts of the signal when the video was made. Decoding it into WAV reproduces exactly what survived and nothing more.

So the output is a large, uncompressed file containing audio that has already been compressed once. That is not a fault, and it is worth stating because people frequently expect an improvement. The gain is that the file is now raw samples, which is what editing tools want; the sound is identical to what the MP4 held.

Editing. Every cut, fade, filter and level change in a WAV starts from exactly what was there before, and saving does not re-encode anything. A lossy file has to be decoded, altered and re-encoded on every save, and after a few rounds of ordinary work the damage becomes audible.

Compatibility is the other half. Audio editors, transcription tools, forced-alignment software, DAWs and broadcast systems all take WAV without question, and several take nothing else. If the next step in your workflow named the format, this is why.

Stereo at 48 kHz and 16 bits is about 11 megabytes a minute. A ten-minute clip is 110 megabytes; an hour-long interview is around 650; a three-hour recording is close to two gigabytes.

Two consequences worth planning for. WAV cannot describe a file larger than about four gigabytes, which at those rates is roughly six hours — long enough for almost everything and not for a full conference day. And these files compress extremely well, so zipping one before sending it typically halves the size at no cost.

A video can carry several audio tracks: an original language and a dub, a commentary, a descriptive track for accessibility, or separate microphone feeds from a multi-camera shoot. This conversion takes the first one.

For material from a phone, a screen recorder or a normal export that is the only track and the question does not arise. For a film, a broadcast recording or a professionally delivered file, check what is in it before assuming the right audio came out — the wrong language is an easy thing to discover an hour into an edit.

If the audio is going to be listened to rather than worked on — a podcast episode to publish, a lecture to keep, a soundtrack for a phone — MP3 or AAC is the right choice. A tenth of the size, and at a decent bitrate the difference is inaudible on any ordinary listening equipment.

The rule that follows is the same one that applies to recording: work in the uncompressed format, deliver in the compressed one. Extract WAV, do the editing, and export MP3 once at the end from the finished result rather than compressing at the start.

Speech recognition tools generally take WAV directly, and several want a specific sample rate or mono audio. Handing them raw samples avoids a decode step and avoids any doubt about what the tool did with a compressed file it had to open first.

For a long recording, extracting the audio also makes everything downstream faster: a 650-megabyte WAV moves through a transcription pipeline considerably more quickly than a multi-gigabyte video that has to be demuxed at every stage.

The extraction runs inside the browser tab on your own processor. The video is never uploaded, so there is no server copy, no queue behind other people and no daily count to exhaust — the free tier’s ceiling is 100 MB a file, which is the one thing a long recording runs into.

That matters for exactly the material this conversion is used on: recorded interviews, client meetings, lectures, medical and legal recordings. The most reliable way to keep a recording private is not to send it anywhere.