How to transcribe an M4A file

The short answer, in four steps

If you only want the file turned into text right now, this is the whole job. Every device is a version of the same four steps, and none of them involve converting the file. The device-by-device detail is further down, and the exceptions are in the sections after that.

1. 1Find the .m4a where it already lives: Voice Memos or the Files app on iPhone, Files by Google or your recorder app on Android, a folder or a cloud drive on a computer.
2. 2Send it straight into a transcription app or site. On a phone that is the Share button; in a browser it is the upload box. Do not convert it to MP3 on the way in.
3. 3Wait for the upload to finish before anything else. On a long recording the upload is the slow half, not the transcription.
4. 4Read the transcript when it lands, fix any misspelled names, then generate the summary or the action items from it.

What an M4A file actually is

M4A is audio stored inside an MP4 container. The sound itself is almost always AAC, the same codec streaming services use. The .m4a extension is just a signal that the file holds audio and no picture. That is the whole story. It is not proprietary, it is not locked to Apple hardware, and it is not a lower class of file than MP3.

It became the default because Apple made it the default. Voice Memos on iPhone saves .m4a. So do most third-party iPhone recorders, most meeting recorder apps, and anything exported out of GarageBand or the Music app. Plenty of Android recorders save it too, usually next to .mp3 and .3gp in their settings. A podcast episode exported from a recording session lands as .m4a as well, and transcribing a podcast episode into show notes takes the same route as everything below.

For transcription none of that matters much. A transcription service reads the audio track, so an M4A goes in the same way an MP3 or a WAV does. What does matter is a small set of edge cases hiding under the same extension, and those are what the rest of this page is about.

Do you have to convert an M4A to MP3 first?

No, in almost every case. Converting takes audio that is already compressed and compresses it a second time. You lose a little quality and gain nothing. If the tool accepts M4A, hand it the M4A.

There are a few real exceptions, and they are worth knowing before you waste an hour on a file that was never going to work. When you do have to convert, the Mac route below uses the Music app, and the Windows equivalent is in the first note underneath it.

1. 1On a Mac, open the Music app.
2. 2Go to Music > Settings > Files > Import Settings, set Import Using to MP3 Encoder or WAV Encoder, then click OK.
3. 3Drag the M4A into the Music window so it appears in your library.
4. 4Select the track, then choose File > Convert > Create MP3 Version (or Create WAV Version).
5. 5Right-click the new track and choose Show in Finder to get at the file itself.

• On Windows, skip steps 1 to 5 and use VLC. Open Media > Convert / Save, add the file, pick an audio profile such as Audio - MP3, name the output file, then click Start. Same result, no Apple software involved.
• Keep the original. Convert a copy, never the only version you have.
• Convert once. M4A to MP3 to something else stacks up losses. Prefer WAV as the target if the tool takes it.
• Do not drop the bitrate to save space. Transcription cares about clarity, and a thin 32 kbps MP3 of a noisy room reads worse than the file you started with.
• Play the converted file before uploading it. Ten seconds of listening saves a failed upload and a puzzled hour.

How to transcribe an M4A file on iPhone

Two routes, and neither one needs a conversion. Share the file into a transcription app from wherever it already lives, or open the app and import it. The Share sheet is faster when the recording is in Voice Memos or arrived in a chat.

If the M4A sits in the Files app or iCloud Drive instead, use the second route: open the app, tap + or New Note, choose Import Audio, then pick the file in the Files browser. That reaches anything the Files app can see, including connected cloud storage. One thing to check before you do it: if the file shows a small cloud icon next to its name, it is a placeholder rather than a real download. Tap it once and wait for the icon to disappear, or the import will fail with no useful error.

The full Share-sheet walkthrough, screen by screen, is in how to transcribe audio on iPhone, and if the file is a Voice Memo specifically, how to transcribe a voice memo on iPhone covers that path on its own. The condensed version is below.

Recent iPhone software can also show a transcript inside Voice Memos itself, for recordings made there. That is handy for a quick read of a short memo. It only covers recordings that live in Voice Memos, it will not touch a file someone sent you, and it gives you plain text with no speaker labels, no summary and no action items.

The steps use Noter AI, because it takes M4A directly and accepts files shared in from other apps. Any app with an iOS Share extension follows the same shape.

1. 1Find the .m4a. In Voice Memos, tap the recording, then the three-dot button. In the Files app, press and hold the file. If a cloud icon is showing, tap the file first and let it download.
2. 2Choose Share, then tap Noter AI in the Share sheet. If it is not there, tap More and switch it on in the list.
3. 3The file imports and transcription starts on its own, with a progress indicator while it runs.
4. 4Read the finished transcript, which carries a timestamp on every segment and speaker labels where there is more than one voice. Tap any line to hear that exact moment in the audio.

How to transcribe an M4A file on Android

The shape is the same and only the buttons differ. On Android the share menu is the standard route, and the M4A usually sits in Files by Google, in your recorder app, or in the chat it arrived in. The wider Android picture is in the best AI meeting note taker for Android.

1. 1Open Files by Google, your recorder app, or the conversation holding the file.
2. 2Long-press the M4A file, or open its menu, and choose Share.
3. 3Pick Noter AI from the list of apps. Install it from Google Play first if it is not there.
4. 4Transcription starts once the file finishes uploading.
5. 5Open the finished note for the transcript, the summary and the action items.
6. 6Translate the output from the same note if the person reading it needs another language.

How to transcribe an M4A file on a computer

Two honest options on a desktop: use a browser tool that reads M4A as it is, or move the file to your phone and transcribe it there. There is no Noter AI desktop app, so if Noter AI is the app you want, take the second route in the next section.

The best known desktop route is the Transcribe feature in the web version of Microsoft Word, if you have a Microsoft 365 subscription and use Edge or Chrome, which produces a timestamped transcript with basic speaker separation. It accepts an .m4a upload directly and labels voices Speaker 1, Speaker 2 and so on, which you can rename afterwards. Two limits to know before you start: the audio file has to be under 300 MB, and the account gets 300 minutes of uploaded audio a month.

It runs in the browser on Windows, Mac, or anything else that runs Edge or Chrome, so the steps are identical on both platforms.

1. 1Go to office.com in Edge or Chrome and sign in with the account that carries your Microsoft 365 subscription. Safari and Firefox do not show the feature.
2. 2Open Word, then start a Blank document.
3. 3On the Home tab, click the small arrow next to Dictate, then choose Transcribe. The Transcribe panel opens on the right.
4. 4Click Upload audio, pick your .m4a file, and let it upload.
5. 5Leave the tab open while it processes. A long recording takes a while, and closing the tab cancels the job.
6. 6When the transcript appears, click Add to document, then All speakers, to drop the timestamped text with its speaker labels into the document. From there you can edit it, rename the speakers, and save or export it like any Word file.

• No Microsoft 365 subscription? Transcribe simply will not appear in the Dictate menu. Free Word for the web does not include it, so use a browser transcription service that accepts M4A uploads, or the phone route below.
• Google Docs voice typing is not an alternative. It listens to your microphone in real time and cannot read a file at all, so playing the M4A out of your speakers into it is the only way, and the quality of that is exactly as bad as it sounds.
• Watch the file size. A one-hour lossless .m4a can pass 300 MB on its own. Convert it to MP3 first, or split the recording, if the upload bounces.

Or move the M4A to your phone and transcribe it there

This is the route to take when the app you actually want lives on your phone, or when the desktop tools you have refuse the file. Getting the M4A across takes about a minute, and any cloud drive both devices can see will do it.

Cable transfer works too, and is faster for a very large file: iPhone through Finder on a Mac or the Apple Devices app on Windows, Android as a plain USB drive.

1. 1On a Mac, drag the M4A into iCloud Drive in Finder. On Windows, drop it into Google Drive, OneDrive or Dropbox.
2. 2Wait for the upload to finish completely. A half-uploaded file will not import.
3. 3On iPhone, open the Files app and find the file. On Android, open your cloud app and download the file to the device first.
4. 4Tap and hold the file, choose Share, then pick your transcription app.
5. 5Stay on Wi-Fi if the recording is long. A full hour of audio is a real upload.
6. 6The finished note syncs across your devices, so you can read it on the iPad even though you imported it on the phone.

Long recordings and very large M4A files

Length is rarely the problem people expect. A one-hour voice memo at the normal compressed setting is usually tens of megabytes, not gigabytes. The same hour recorded with Lossless switched on can be several times bigger, and that is where uploads start to hurt.

What usually breaks on a long file is the upload, not the transcription. On a weak mobile signal a large file stalls partway and the app looks stuck when it is really just waiting. Move to Wi-Fi, make sure the file is genuinely downloaded to the device rather than sitting in the cloud, and leave the app in the foreground until the progress bar completes.

Transcription time itself scales roughly with length. A short memo comes back in a couple of minutes and a long meeting takes proportionally longer. There is a fuller breakdown in how long AI takes to transcribe a meeting.

If the recording is much longer than the part you care about, trim it before you upload.

1. 1Open Voice Memos and tap the recording.
2. 2Tap the three-dot button and choose Edit Recording.
3. 3Tap the crop icon, then drag the yellow handles to the section you want.
4. 4Tap Trim, then Save as a new recording so the original survives.
5. 5Share the trimmed file into your transcription app as normal.
6. 6For a two-hour file that keeps failing, split it into two halves and transcribe them separately, then read the two summaries side by side.

When the M4A came from a video

Be careful here, because two different things end up called M4A. A true M4A holds an audio track and nothing else. But .m4a, .mp4, .m4v and .mov are all the same container family, so a file can carry a video track while wearing an audio extension, either because someone renamed it or because an export tool was sloppy.

This matters because Noter AI takes audio only. Its Share extension accepts audio files and refuses everything else, so a video file will not appear as a valid target no matter what the extension says. There is no video import of any kind, no MP4 or MOV support, and no way to paste a link to a video hosted online.

The fix is to pull the audio track out on a computer first, then transcribe the extracted file like any other M4A. On a Mac that is QuickTime Player, step by step below. On Windows it is VLC instead: Media > Convert / Save, add the video, choose an audio-only profile, name the output file, click Start, then pick up at step 3.

1. 1On a Mac, open the video in QuickTime Player.
2. 2Choose File > Export As > Audio Only. QuickTime writes an .m4a containing the audio track alone.
3. 3Play the exported file and confirm you can hear speech before going further.
4. 4Move the exported audio to your phone through iCloud Drive, Google Drive or a cable.
5. 5Share it into your transcription app exactly like any other M4A.

If the recording has more than one language in it

This is where most transcription tools quietly fail, and imported files are the worst case. Nobody was thinking about the software when they pressed record, so the recording is whatever the room actually sounded like.

Many apps make you choose a language before transcription starts. When the file holds two, the second one comes back transliterated phonetically or dropped altogether. An Arabic conversation with English product names in it turns into Arabic letters spelling English words. A Hindi and English mix breaks the same way. So does a Spanish call where the numbers and the software names are said in English.

Noter AI tags language at the word level instead of locking the whole file to one language. It transcribes 60+ languages with detection always on, so there is nothing to choose before importing, and a sentence carrying two languages arrives with both halves intact. Finished notes can be translated into 18 languages: Arabic, Chinese (Simplified), Chinese (Traditional), Dutch, English, French, German, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Thai, Turkish and Vietnamese. There is a longer comparison of how the main apps handle this in the best app for transcribing multilingual meetings.

Troubleshooting: what goes wrong, and the fix

Most import problems come down to one of a few causes. Work down this list before blaming the app.

One general rule: if the file plays properly in VLC on a computer, the audio is intact and the problem is somewhere in the handoff, not in the recording.

• The file is greyed out in the picker, or the app is missing from the Share sheet. You are probably trying to share a video, a PDF or a link. Audio-only apps offer themselves for audio and stay hidden for everything else. Confirm the file plays as sound on its own. If the app is missing from an otherwise valid share, tap More in the Share sheet and switch it on.
• You shared it from Google Drive or Dropbox and nothing happened. You sent a link, not audio. A transcription app cannot follow a URL and fetch a file. Open the Drive app, tap the three-dot menu on the file, choose Save to Files or Open in, then share the real file. The same goes for a web page holding a video or a podcast, there is no URL import.
• The file has a small cloud icon next to it in the Files app. That is a placeholder, not a download. Tap it once and wait for the icon to disappear before sharing it anywhere.
• Unsupported codec, or the upload is rejected outright. Usually the file is an .m4b audiobook, an old DRM-protected .m4p bought from iTunes years ago, or Apple Lossless rather than plain AAC. Some services reject Apple Lossless; Noter AI does not, but a lossless file is several times larger, so with those the upload is the real risk and converting to MP3 or WAV is about size, not compatibility. An .m4b usually converts fine. A DRM-protected file cannot be converted or transcribed at all, so you need an unprotected copy.
• The file will not open or play anywhere. A recorder that crashed mid-session can leave a file with no usable index. Try VLC, which is more forgiving than QuickTime and the Music app. If VLC cannot play it either, the audio is gone and no transcription tool will rescue it.
• The file is over an hour and the upload keeps failing. Switch to Wi-Fi, make sure the file is fully downloaded locally instead of being a cloud placeholder, and leave the app open until the upload completes. If it still fails, trim or split the recording and do it in two passes.
• No audio track, or a transcript that comes back empty. Either the file holds video with no sound, or the microphone was blocked while recording. Play it with the volume up to check. If the first minute is silent, the recorder was probably denied microphone permission at the start.
• The transcript is in the wrong language, or half of it is nonsense. The tool locked itself to one language. Re-run the same file in something that detects language automatically and handles switching mid-sentence. You do not need to re-record anything, the original audio is fine.
• Names and technical terms are spelled wrong. Normal in any transcript, and worth fixing before you send it anywhere. Correct them in the transcript, then re-run the summary so the fix carries into the output. Editing an AI transcript and re-running the summary walks through it.
• The transcript reads fine but nobody is labelled. Speaker separation needs voices it can tell apart, and a phone that was face-down in a bag will not give it that. Can AI tell who said what in a meeting explains what helps and what does not.

Turning the transcript into something you can send

A raw transcript is material, not a result. What people usually want from an M4A is the short version, the decisions, and the list of who agreed to do what.

That is the part Noter AI is built for. The same imported file can be reshaped into minutes of meeting, key bullet points, an email, a to-do list, a formal report, or a custom instruction you describe in a sentence. Action items are pulled out on their own. You can ask questions about one note, or across every note you have, and search every note's title and AI notes at once.

The transcript stays editable, so a misspelled name gets fixed once and the corrected summary follows. Export to PDF, Word or plain text, or copy formatted text straight to the clipboard. Notes sync across your devices, and everything is encrypted in transit and at rest, with deletion available whenever you want it.

The plan is $9.99 a month or $49.99 a year, one flat price per person. For file transcription the number that actually bites is the minute allowance, not the monthly rate, because M4A files arrive in whatever length they arrive in. A single ninety-minute board recording or a two-hour panel can swallow a free tier in one import. There is no minute cap here and no per-seat multiplier if the files come from several people. Free trial on iPhone and iPad or Android.