Convexy

Extract the audio from a MOV as M4A

Your iPhone shot a MOV, but what you actually wanted was the sound — the interview, the recital, the lecture, the thing someone said. Convexy discards the picture and writes the audio out on its own, on the device.

What you get, and what you lose

A MOV is a QuickTime container holding a video stream and an audio stream side by side. Extracting to M4A means the video stream is discarded and the audio is exported on its own, into the container that audio was already in.

Your iPhone records audio as AAC, and M4A is AAC's natural home — so this is about as direct as an audio extraction gets. Convexy runs it through Apple's AAC exporter rather than copying the bits verbatim, so treat it as one more encode rather than a free remux. From an iPhone recording, at iPhone bitrates, that is not something you will hear.

What you lose is the picture. That is the point, and it is where the entire size saving comes from: in a video, the images are nearly all of the bytes. A 2 GB recording of a school concert is mostly pixels of a stage. The audio alone is a small fraction of it.

The recordings people actually do this to

It is worth naming them, because they have something in common:

What they have in common is that they are personal, they are often confidential, and several of them are the kind of thing you cannot recreate. They are also, unavoidably, exactly the files a cloud converter would ask you to upload.

There is no server. Convexy does the extraction on your iPhone or iPad using its own media hardware. Nothing is uploaded, no account exists, and it works in Airplane Mode — which is the simplest way to prove to yourself that a recording of a private conversation is not being sent anywhere.

Choosing the output format

Convexy can extract the audio from a MOV as M4A, MP3, WAV or FLAC. They are not equivalent, and the honest ranking is:

The M4A encode itself

Convexy exports M4A using Apple's own AAC exporter at its standard high-quality setting. There is no bitrate dial for M4A output, because that exporter does not offer one.

If you want control — a small file for a two-hour lecture, say — extract to MP3 at 128 kbps in mono instead. A single person speaking into a phone is not stereo in any meaningful sense, and mono at a modest bitrate will produce a file you can actually email.

The original video is untouched

Convexy writes a new audio file. The MOV stays exactly where it was, unchanged and undeleted. If you decide later that you want the video after all, it is still there.

You can also rename the output before saving, which is worth doing — an audio file called IMG_4471 is not going to mean anything to you in a year.

How to do it

  1. Bring the MOV into Convexy

    Choose it from Photos, tap Browse files to find it in Files, or share the video into Convexy from any app.

  2. Choose M4A

    Convexy offers only what the file can genuinely become — M4A appears alongside MP4, M4V, animated GIF and still frames.

  3. Convert

    The video track is discarded and the audio is exported on your device. Nothing is uploaded at any point.

  4. Rename the output

    Give it a name you will recognise later. IMG_4471.m4a will mean nothing to you in six months.

  5. Save or share it

    Save the M4A to Files, or send it straight on to a transcription service, an editor, or whoever needs it.

Common questions

Does extracting the audio change the original video?

No. Convexy writes a new audio file and leaves the MOV exactly as it was — not modified, not moved, not deleted. If you want the video later, it is still there.

Does the extraction lose quality?

Very little, but be precise: the audio in your MOV is already lossy AAC, and Convexy exports it through Apple's AAC exporter rather than copying the bits verbatim. Treat it as one more encode rather than a free remux. From an iPhone recording it is not something you will be able to hear.

Should I extract to WAV or FLAC for better quality?

No — this is the most common mistake in audio conversion. The audio in the video is lossy AAC. Wrapping it in WAV or FLAC gives you a file many times larger containing precisely the same audio, holes and all. Lossless formats preserve what they are given; they do not reconstruct what was thrown away. Extract to WAV or FLAC only when a tool refuses to accept anything else.

How do I get a smaller file for a long recording?

Extract to MP3 at 128 kbps and set channels to Mono. A single voice recorded on a phone is not meaningfully stereo, so mono halves the file for nothing you will miss. A two-hour lecture that way lands around 60 MB, which is emailable; the source MOV might have been several gigabytes.

Is my recording uploaded anywhere?

No. Convexy has no backend at all. Everything runs on your iPhone or iPad, and the conversion works with the network switched off — worth checking yourself if the recording is an interview, a consultation or anything else you would rather not hand to a server.