Audacity cannot import M4A: the fix that always works

They hit "Audacity attempted to use FFmpeg but the libraries were not found" or "did not recognize the type of this file", spent an hour on the optional library installer, and failed. They want to edit their audio, not debug a dependency.

Convert the file to WAV or FLAC and import that instead. Audacity refuses .m4a because the audio inside it is AAC, and AAC support is not built in — it lives in an optional FFmpeg library that Audacity asks you to install separately. WAV and FLAC need no library at all, on any version, on any operating system, and they have needed none for twenty years. A converted copy imports on the first try, keeps every sample the AAC decoder produced, and takes about as long as the library installer takes to download. If you specifically want the library working, the last section explains how to do it properly.

The fix, in one line

Take the .m4a, convert it to WAV, 16-bit, at whatever sample rate the original has, and drag the result into Audacity. Nothing else changes. The waveform you get is the decoded audio from the M4A, sample for sample, which is precisely what Audacity would have shown you had the library been present.

The converter on this site will do it without an account, and without the file going anywhere — FFmpeg is compiled to WebAssembly and runs inside the browser tab, so the same library Audacity wants you to install is already loaded on this page, just not on your PATH. Any other route is equally valid: VLC's Media > Convert / Save, iTunes or the Music app's Create WAV Version, or a command line if you have FFmpeg there already.

What the error actually means

Two messages send people here, and they are the same problem at two different stages.

Audacity did not recognize the type of the file '…'. Try installing FFmpeg. For uncompressed files, also try File > Import > Raw Data.

Audacity attempted to use FFmpeg to import an audio file, but the libraries were not found.

The first appears when no FFmpeg library is configured at all. The second appears when Audacity has a path recorded but cannot load what is at the end of it — usually a version, an architecture or a filename it does not accept.

Audacity's own importers cover the formats whose decoders it is free to ship: everything libsndfile handles, plus MP3 and Ogg Vorbis, plus Opus from version 3.2 onwards. AAC is patent-encumbered, so Audacity has never bundled a decoder for it. The .m4a extension only names the container — an MPEG-4 box structure borrowed from video — and Audacity has no reader for that container either, which is why the file is rejected before the codec question even arises. An ALAC file with the same extension fails for the same reason, despite ALAC being lossless and open.

What Audacity opens with no library, and what it does not

Read that list from the other direction and the strategy is obvious. Converting into a format every Audacity install can already read removes the dependency permanently, rather than fighting it once per machine.

WAV or FLAC, and at what settings

Bit depth: 16-bit. Audacity works internally at 32-bit float regardless of what you give it, so importing 16-bit costs nothing while you edit. Writing 24-bit is only worth it if the source was captured at 24-bit, and an M4A never was — AAC does not store bit depth, so a 24-bit export just pads every sample with eight zero bits and adds half again to the file size.

Sample rate: leave it alone. Whatever the M4A holds, usually 44,100 Hz for music and 44,100 or 48,000 Hz for a phone recording, keep it. Resampling on the way in is an extra approximation for no benefit, and Audacity will show you the track's rate in the panel on the left of the waveform. If the material is destined for a video timeline, 48,000 Hz is the rate to match, and the guide on formats video editors accept covers why that matters more there than it does here.

WAV or FLAC? WAV if the recording is short, FLAC if it is long. A 40-minute interview at 128 kbps occupies about 37 MB as an M4A. The same audio is roughly 400 MB as a 16-bit 44.1 kHz stereo WAV, and typically 200 to 280 MB as FLAC, which Audacity imports just as readily. WAV also has a hard 4 GB ceiling written into its 32-bit length fields, which a multi-hour stereo recording will reach. WAV vs MP3 works through that arithmetic properly.

Why the library install keeps failing

Four causes account for nearly all of it, and none of them are anything you did wrong.

• Architecture mismatch. A 32-bit FFmpeg will not load into a 64-bit Audacity, and Windows builds have been 64-bit since version 3.1. On a Mac the equivalent trap is an Intel build of the library against an Apple silicon build of Audacity — an x86_64 dylib cannot be loaded by an arm64 process, and the dialog reports only that the libraries were not found.
• Version mismatch. Audacity links against a specific FFmpeg major version and ignores the rest. A newer avformat than the one your release expects is not a newer version of the right thing, it is the wrong thing.
• The download is blocked. The installers are hosted off Audacity's own site, and corporate proxies, filtered school networks and some antivirus products treat an unsigned codec DLL exactly as you would hope they treat an unsigned codec DLL.
• Permissions. The library has to land somewhere Audacity can read at launch. Installs into a per-user folder, or into an application bundle the OS considers sealed, fail silently and only surface as the same import error later.

The honest cost-benefit: the library buys you direct import of M4A, WMA and video files, which is worth having if you do this weekly. If you have one interview to edit today, thirty seconds of converting beats an hour of chasing a DLL, and you will still have a WAV afterwards that opens anywhere.

Installing it properly, if you would rather

Open Edit > Preferences > Libraries (on macOS, Audacity > Preferences, or Settings in newer builds). The FFmpeg section shows what Audacity currently sees. Download opens the page in Audacity's manual that names the correct build for your exact version — start there rather than searching, because the build has to match. After installing, come back and press Locate, point it at avformat-*.dll on Windows or libavformat.*.dylib on macOS, and restart Audacity before testing. Never take FFmpeg from a general-purpose "codec pack" site. Those bundle far more than a decoder, and the risk is real in a way the import error is not.

When it is not the library at all

If you converted the file and Audacity still refuses it, or if it imports as silence, one of these is usually the answer.

• The file is a placeholder. Files stored in iCloud Drive or OneDrive can appear at full name and near-zero size until they are downloaded. Check the size in the file manager first — a 0 KB or 2 KB "recording" has no audio in it yet.
• The extension is a lie. Renaming something to .m4a or .mp3 changes nothing inside the file, and Audacity reads the contents. Renaming versus converting explains what the two actually do.
• The recording is protected. Purchased .m4p tracks and DRM-wrapped audiobooks will not open in Audacity, in this converter, or in any other tool, and that is not a bug to be worked around.
• It is a video file. An .mp4 or .mov needs the same library. Pull the audio out first, then import that.

Getting the audio back out afterwards

Export is a separate question from import, and a happier one. Audacity has bundled the LAME encoder since 2.3.2, so File > Export > Export as MP3 works with nothing installed. WAV and FLAC export are built in as well. Only M4A, AAC and WMA export still route through the same optional library, so if you need an .m4a at the end, export a WAV from Audacity and convert that last file instead — one lossy encode at the end of the chain rather than one at each hop.

A working order for the whole job: convert the M4A to WAV, edit in Audacity, export a WAV, encode once to whatever you are delivering. Every intermediate step stays lossless, and the only generation of loss you add is the one you cannot avoid. What no step can do is put back the detail the original AAC encoder discarded — that was decided when the recording was first saved, and nothing downstream knows what was removed.