Audio Extractor

A video file holds separate streams: one carrying the picture, one carrying the sound, sometimes subtitles alongside them. Extracting the audio means taking the sound stream out and writing it into a file of its own. The picture is discarded, not converted, which is why the result is a fraction of the size.

The useful thing to know is that this often does not have to be a conversion at all. An MP4 or MOV almost always carries AAC audio, and M4A is simply that same AAC in an audio-only container from the same family. The stream can be copied across untouched: no decoding, no re-encoding, nothing lost, and it finishes in about the time it takes to read the file off disk. The result is bit for bit the audio that was in the video.

Most converters answer this query with an MP3 and do not mention what that costs. AAC and MP3 are both lossy, so going from one to the other means decoding audio that has already had detail removed and compressing it again with a different algorithm. The second encoder cannot tell which artefacts were in the original recording and which the first encoder left behind, so it spends bits preserving both. That is generation loss, and it is audible on cymbals, applause and sibilance before it is audible anywhere else. MP3 is still the right choice when the destination will not read anything else, and it is offered here with the cost stated rather than buried.

WAV is the third option and a different trade. It is uncompressed PCM, so it adds no further loss, though it cannot recover what the source already discarded. It is what audio editors, transcription tools and speech models generally want as input. It is also about 10 MB a minute in stereo, which is worth knowing before extracting an hour-long recording.

All of it runs in the browser through ffmpeg compiled to WebAssembly. The video is read from your disk into the page, the audio stream is pulled out in the tab, and the file is handed back as a download. Nothing is uploaded, nothing is queued on a server, and no copy exists for anyone to delete afterwards. For a recording of a meeting, a lecture or an interview, that is the difference between a private operation and handing a stranger your audio.