How to Convert M4A to MP3 with ffmpeg

By the M4AtoMP3.org editorial team · October 21, 2026 · 10 min read

ffmpeg is the engine behind a huge share of the world’s audio and video tools — including, in a WebAssembly build, parts of the converter on this site. Using it directly gives you complete control over bitrate, channels, tags and naming, and it handles thousands of files as easily as one. The downside is that it is a command-line program with hundreds of options. This guide gives you the handful of commands you actually need for M4A to MP3, explains what each flag does, and covers the errors people most often hit.

Install ffmpeg

• Windows: download a “full” or “essentials” build from gyan.dev or BtbN on GitHub, unzip it, and add the bin folder to your PATH. Alternatively run winget install ffmpeg.
• macOS: brew install ffmpeg with Homebrew.
• Linux: sudo apt install ffmpeg (Debian/Ubuntu), sudo dnf install ffmpeg (Fedora with RPM Fusion) or your distribution’s equivalent.

Check that the MP3 encoder is available:

ffmpeg -hide_banner -encoders | grep mp3

You should see libmp3lame. On Windows PowerShell use findstr mp3 instead of grep.

The one command most people need

ffmpeg -i input.m4a -c:a libmp3lame -q:a 2 output.mp3

• -i input.m4a is the source file.
• -c:a libmp3lame chooses the LAME MP3 encoder.
• -q:a 2 asks for variable bitrate (VBR) quality level 2, which averages roughly 170–210 kbps and is transparent for most listeners on most music.
• output.mp3 is the destination; the extension tells ffmpeg to write an MP3 file.

That is all. ffmpeg reads the AAC or ALAC audio, decodes it, and encodes MP3.

Constant bitrate instead of VBR

Some car stereos, older hardware players, podcast hosts and DJ software expect constant bitrate (CBR). Use -b:a:

ffmpeg -i input.m4a -c:a libmp3lame -b:a 192k output.mp3

Common choices are 128k for speech, 192k for general music and 320k for archival copies from lossless sources. Our CBR vs VBR guide explains which to choose and why VBR is usually the better default.

VBR quality levels at a glance

Averages depend on the content; quiet speech produces lower bitrates than dense rock.

Voice recordings: mono and lower sample rate

Voice memos and lectures rarely need stereo. Downmixing to mono halves the data the encoder has to spend, so you can use a lower bitrate without losing clarity:

ffmpeg -i memo.m4a -ac 1 -c:a libmp3lame -b:a 64k memo.mp3

-ac 1 sets one audio channel. You can go further with -ar 22050 to resample to 22.05 kHz, which is still clear for speech, but most players handle 44.1 kHz better, so only do this if size really matters. The mono vs stereo and sample-rate article goes deeper.

Keep tags and album art

By default ffmpeg copies global tags (title, artist, album) but may skip the cover image and write ID3v2.4, which Windows Explorer and some cars do not read. A more compatible command:

ffmpeg -i input.m4a -map 0:a -map 0:v? -c:a libmp3lame -q:a 2 -c:v copy -id3v2_version 3 -write_id3v1 1 output.mp3

• -map 0:a includes the audio stream.
• -map 0:v? includes the embedded cover picture if there is one (the ? stops ffmpeg failing when there is not).
• -c:v copy copies the JPEG/PNG cover without re-compressing it.
• -id3v2_version 3 writes ID3v2.3, the most widely supported tag version.
• -write_id3v1 1 adds an old-style ID3v1 tag for very old players.

If the cover still does not appear, see our article on keeping metadata when converting.

Batch convert a whole folder

macOS and Linux (bash/zsh):

for f in *.m4a; do ffmpeg -n -i "$f" -c:a libmp3lame -q:a 2 "${f%.m4a}.mp3"; done

Windows PowerShell:

Get-ChildItem *.m4a | ForEach-Object { ffmpeg -n -i $_.FullName -c:a libmp3lame -q:a 2 ($_.BaseName + ".mp3") }

Windows Command Prompt:

for %f in (*.m4a) do ffmpeg -n -i "%f" -c:a libmp3lame -q:a 2 "%~nf.mp3"

The -n flag tells ffmpeg never to overwrite existing files, so you can re-run the loop safely after an interruption. To process subfolders too, use find on Unix or Get-ChildItem -Recurse in PowerShell. Our dedicated batch conversion guide includes a version that mirrors the folder structure into a separate output directory.

Trim, normalise or fade while converting

Because ffmpeg decodes the audio anyway, you can apply simple edits in the same pass:

• Trim: -ss 00:00:05 -to 00:10:00 keeps 5 seconds to 10 minutes.
• Fade in: -af "afade=t=in:d=2" adds a 2-second fade-in.
• Loudness normalise: -af loudnorm=I=-16:TP=-1.5:LRA=11 targets −16 LUFS, a common podcast level. See how to normalise audio volume.

Example combining trim and normalisation:

ffmpeg -ss 5 -i talk.m4a -af loudnorm=I=-16:TP=-1.5 -c:a libmp3lame -b:a 128k talk.mp3

Common errors and fixes

“Unknown encoder ‘libmp3lame’” — your build lacks LAME. Install a full build as described above.

“moov atom not found” — the M4A is incomplete, often from a recording that was interrupted or a download that did not finish. The audio index sits at the end of the file and is missing. Tools such as untrunc can sometimes rebuild it using a healthy file from the same device as a reference.

“Invalid data found when processing input” — the file is not really M4A (it may be a renamed WebM, OGG or even an HTML error page). Run ffprobe input.m4a to see what it actually is.

Output is silent — check that you did not use -vn together with -map 0:v, and that the input really has an audio stream (ffprobe lists it).

“Protected” or DRM errors — .m4p files from the old iTunes Store are encrypted; ffmpeg cannot and should not decrypt them.

Should you use ffmpeg or a GUI converter?

ffmpeg is unbeatable for automation and consistent results, and it is free. It is overkill if you convert one file a month. For occasional conversions, the browser M4A to MP3 converter gives you the same LAME-quality output with a file picker and bitrate menu, and nothing is uploaded. If you prefer a desktop app, VLC and Audacity both work.

Useful ffprobe checks before and after converting

ffprobe installs alongside ffmpeg and tells you what you are working with. Three commands worth knowing:

ffprobe -hide_banner input.m4a

shows the codec (aac or alac), sample rate, channels, bitrate and duration. If the codec is alac, the source is lossless and it is worth encoding at a higher quality such as -q:a 0 or -b:a 320k.

ffprobe -v error -show_entries format_tags -of default=nw=1 input.m4a

lists the tags (title, artist, album, date) so you can confirm they carried across after conversion — run it on the MP3 too.

ffprobe -v error -show_entries stream=codec_name,bit_rate -of csv=p=0 output.mp3

prints just the codec and bitrate, which is handy in scripts that verify each output.

Choosing settings by source

These are starting points. If a file sounds fine at a lower setting, the smaller file is the better file.

Speed tips

LAME is single-threaded, so one conversion uses one CPU core. On a multi-core machine you can convert several files in parallel. On Linux and macOS with GNU parallel installed:

ls *.m4a | parallel ffmpeg -n -i {} -c:a libmp3lame -q:a 2 {.}.mp3

That runs one ffmpeg per core and can turn an hour-long batch into ten minutes. Use -loglevel error to keep the terminal output readable when running many jobs at once.

Summary

For nearly every case, ffmpeg -i input.m4a -c:a libmp3lame -q:a 2 output.mp3 is the right command. Switch to -b:a 192k when something demands constant bitrate, add -ac 1 for voice, add the mapping and ID3v2.3 options to keep album art, and wrap it in a loop for whole folders.

← Back to all articles