How to convert apple voice memos to mp3 — free, no upload
- Step 1Export the memo as M4A — In the Voice Memos app, tap the memo, then Share (or the ⋯ menu) and save/AirDrop it as a file — it comes out as
.m4a. On a Mac, drag the memo out of the app or use File → Export. - Step 2Drop the .m4a onto the converter — Drag the exported memo onto the dropzone. It loads locally — the recording is never sent anywhere, so even confidential interviews stay on your device.
- Step 3Pick a voice-friendly bitrate — For speech, choose 64 kbps (voice) for the smallest file or 128 kbps (podcast) for a touch more clarity. The dropdown is the only control; the memo's mono channel and sample rate pass through unchanged.
- Step 4Convert in the browser — FFmpeg WebAssembly decodes the AAC and re-encodes to MP3 locally. The memo's saved title is copied to an ID3 tag so you can still tell recordings apart.
- Step 5Download and use the MP3 — Save the MP3 and hand it to your transcription tool, attach it to an email, or drop it into a case/patient record — it now plays on Windows, Android, and web players.
- Step 6Trim or clean up first if needed — If a memo has a long silent run-up or you only need part of it, use the audio trimmer or the silence stripper before converting. To even out a quiet recording, run the loudness normalizer; to label the result, see edit MP3 metadata for a podcast.
Bitrate options in the picker
The Bitrate dropdown is the only control on this tool. MP3 is encoded with LAME (libmp3lame) at the constant bitrate (CBR) you choose; the default is 192 kbps. There is no sample-rate or channel control here — the source's sample rate and channel layout pass through unchanged.
| Picker value | FFmpeg flag | Typical use | Approx. size / minute (stereo) |
|---|---|---|---|
| 320 kbps · best | -b:a 320k | Music, mastered exports, the highest MP3 quality available here | ~2.4 MB |
| 256 kbps | -b:a 256k | Music and interviews where you want headroom but a smaller file than 320 | ~1.9 MB |
| 192 kbps · default | -b:a 192k | All-round default — transparent for most spoken-word and casual music | ~1.4 MB |
| 128 kbps · podcast | -b:a 128k | Podcasts and talk content where file size matters more than music fidelity | ~0.94 MB |
| 64 kbps · voice | -b:a 64k | Mono voice notes, dictation, transcription source where size is critical | ~0.47 MB |
Per-file limits by plan
Limits come straight from the audio tier table. Note the per-file DURATION cap is separate from the file-size cap — a long-but-small recording can hit the minutes limit before the megabytes limit. M4A to MP3 is on the Free tier.
| Plan | Max file size | Max duration / file | Files per run |
|---|---|---|---|
| Free | 50 MB | 30 minutes | 1 |
| Pro | 200 MB | 120 minutes | 10 |
| Pro + Media | 100 GB | Unlimited | 100 |
| Developer | 100 GB | Unlimited | 100 |
What is preserved vs. changed in the transcode
The converter runs -c:a libmp3lame -b:a <bitrate> -map_metadata 0 -map 0:a -map 0:v? -c:v copy -disposition:v:0 attached_pic -id3v2_version 3. Tags and embedded artwork are carried across; the audio itself is re-encoded.
| Property | Behaviour | Why |
|---|---|---|
| Title / artist / album / year tags | Copied to the MP3 (ID3v2.3) | -map_metadata 0 carries container metadata; -id3v2_version 3 writes the most widely-read ID3 flavour |
| Embedded cover art | Re-attached, not re-encoded | -map 0:v? picks up the attached picture and -c:v copy copies it byte-for-byte, then marks it attached_pic |
| Sample rate (e.g. 44.1 / 48 kHz) | Passed through unchanged | No -ar flag is set for this tool |
| Channels (mono / stereo) | Passed through unchanged | No -ac flag is set for this tool |
| Audio fidelity | Re-encoded (lossy → lossy) | AAC is decoded then LAME re-encodes to MP3 — a second lossy generation |
Cookbook
Real voice-memo conversions. Voice recordings are mono speech, so these lean on the low bitrates — the engine builds the FFmpeg call from your Bitrate choice.
Interview memo to MP3 for transcription
A 40-minute recorded interview straight from an iPhone. Transcription services want MP3; 64 kbps mono is plenty for speech recognition and keeps the upload to the service small.
Input: interview-jan12.m4a (mono, 40 min, 19 MB)
Picker: 64 kbps · voice
ffmpeg -i interview-jan12.m4a -c:a libmp3lame -b:a 64k ... \
interview-jan12.mp3
Output: interview-jan12.mp3 (64 kbps mono, ~18 MB)
Note: 40 min > Free's 30-min cap — needs Pro (120 min) for one
pass, or trim it first.Quick reminder memo emailed to a Windows colleague
A 2-minute spoken note. Convert to a small MP3 that opens in Windows Media Player / Outlook without the 'unsupported file' headache M4A causes on some setups.
Input: reminder.m4a (mono, 2 min) Picker: 128 kbps · podcast Output: reminder.mp3 (128 kbps, plays in any Windows player)
Clinical dictation kept entirely offline
Patient-related dictation must not touch a third-party server. Because the whole conversion runs in the browser via WASM, the audio never leaves the machine — suitable where upload-based converters are off-limits.
Input: dictation-ptA.m4a (mono, 6 min)
Picker: 64 kbps · voice
Processing: 100% local (FFmpeg WASM) — no network request with
the audio. Output dictation-ptA.mp3 goes straight into
the records system.Lecture / meeting memo at 128 kbps
Longer talks with a couple of speakers benefit from 128 kbps over 64 — clearer consonants for note-taking, still a modest file.
Input: lecture.m4a (mono, 28 min, 13 MB) Picker: 128 kbps · podcast Output: lecture.mp3 (128 kbps, ~26 MB) Within Free's 50 MB / 30 min caps.
Tidy a memo before converting
Memos often start with fumbling and dead air. Trim or strip silence first, then convert — a cleaner, shorter MP3 for the archive.
1. /audio-tools/silence-stripper → cut the silent run-up 2. m4a-to-mp3 (64 kbps) → final memo.mp3 Both steps are local; the trimmed audio feeds the converter without an upload in between.
Edge cases and what actually happens
Memo longer than 30 minutes on Free
RejectedInterviews and lectures routinely run past Free's 30-minute per-file cap even when they're well under 50 MB (speech is small). Upgrade to Pro (120 minutes) or trim the memo with the audio trimmer into sub-30-minute parts first.
Memo shared as .caf instead of .m4a
SupportedSome Apple flows export a .caf (Core Audio Format) wrapper. FFmpeg reads CAF and converts to MP3 like any other source — the engine looks at the codec, not the extension.
Voice memo recorded in stereo
By designMost memos are mono, but external-mic or some iPad recordings can be stereo. The tool passes channels through, so a stereo memo stays stereo (and uses more bits at the same bitrate). It won't fold to mono — that's a separate operation.
Memo title with emoji or accents
PreservedVoice Memos titles can include emoji or accented characters. These copy into the ID3 tag fine; if a downstream player shows them oddly, that's the player's tag rendering, not the conversion.
Output sounds the same size as the M4A
ExpectedAAC is efficient for speech, so a low-bitrate MP3 can be a similar size to the source. To go smaller, drop to 64 kbps or trim silence first; M4A to MP3 is about compatibility more than shrinkage.
Memo still recording / not finished exporting
ErrorIf you export while a memo is mid-save, the M4A can be missing its index (moov atom) and FFmpeg can't decode it. Make sure the memo has finished and re-export the complete file.
Transcription service still rejects the MP3
Check service limitsIf your transcriber rejects the result, it's usually their size/length limit, not the format. Lower the bitrate (smaller file) or split the recording — the MP3 itself is a standard CBR file every service accepts.
Multiple memos at once on Free
LimitedFree converts one memo per run. To process a batch of interviews, upgrade to Pro (10 per run) or higher, or convert them sequentially.
Frequently asked questions
How do I get my iPhone voice memo as an M4A file?
In the Voice Memos app, open the memo, tap Share (or the ⋯ menu), and save it to Files or AirDrop it — it exports as .m4a. On a Mac, drag the memo out of the app or use File → Export.
Why convert a voice memo to MP3 at all?
Transcription tools, email clients on Windows, case-management and clinical systems, and non-Apple players often expect MP3 and choke on M4A. Converting makes the recording usable everywhere.
Which bitrate is best for a voice recording?
64 kbps (voice) for the smallest file that's still clearly intelligible, or 128 kbps (podcast) if you want crisper consonants for note-taking. Voice memos are mono speech, so higher bitrates mostly waste space.
Is my recording private when I convert it?
Yes. The conversion runs locally in your browser via FFmpeg WebAssembly — the audio is never uploaded. That's why it's suitable for interviews, statements, and dictation that can't go to a third-party server.
Will the memo's name be kept?
Yes — the title Voice Memos saved is copied to an ID3 tag, so you can still tell recordings apart after conversion.
Can I convert a memo on a Windows PC or Android phone?
Yes. It runs in any modern browser, so you can convert an AirDrop'd or emailed memo without a Mac.
My interview is 40 minutes — why is it rejected?
Free caps each file at 30 minutes (separate from the 50 MB size cap). Trim it into shorter parts with the audio trimmer, or upgrade to Pro for 120-minute files.
Does converting hurt the speech quality?
Barely, for spoken word. M4A to MP3 is a lossy-to-lossy transcode, but at 64–128 kbps voice stays clear. If you want maximum fidelity, keep the original M4A and convert a copy.
Will the memo stay mono?
Yes — channels pass through unchanged, so a mono memo produces a mono MP3 and doesn't bloat to stereo.
Can I cut out the silence at the start of a memo?
Yes — run it through the silence stripper or audio trimmer first, then convert. Both run locally, so the audio isn't uploaded between steps.
Does this work for Mac Voice Memos too?
Yes. Mac Voice Memos also save as M4A; export and drop the file in exactly as you would from an iPhone.
What if my memo is a .caf file?
FFmpeg reads Core Audio Format (.caf) and converts it to MP3 too — the engine detects the codec rather than relying on the extension.
Privacy first
Every JAD Audio tool runs entirely in your browser via FFmpeg (WebAssembly) and RNNoise. Your audio files never leave your device — verified by zero outbound network requests during processing.