M4A to WAV

This page decodes the AAC audio inside your M4A file and writes the samples straight out as a 16-bit PCM WAV, at the sample rate the recording was made at. Nothing is compressed a second time. The file will get considerably larger, and the exact figure appears before you press anything.

Why you have an M4A in the first place

M4A is what Apple devices produce by default. Voice Memos on an iPhone, recordings from QuickTime, tracks bought from iTunes, and the audio you get when you export from GarageBand all land as .m4a. It is also the extension you get when a video editor exports audio only, and what most podcast tools produce.

The container is the MP4 container, holding a single AAC audio track. The extension exists purely as a hint to software: .mp4 usually means there is video, .m4a says audio only, and the bytes inside are otherwise the same family. That is why some software opens an .m4a happily and other software refuses it while accepting an identical file renamed .mp4.

AAC is a good codec — generally better than MP3 at the same bitrate, which is why it is what streaming and broadcast standardised on. The problem is never quality. It is the software that will not take the file.

Where M4A stops being convenient

Plenty of tools accept only uncompressed PCM. Game engines, telephony and IVR systems, older samplers and hardware, some broadcast automation, most speech-analysis and acoustics packages, and a surprising number of upload forms for voice prompts all want a WAV and nothing else.

Windows is inconsistent too: the built-in players handle M4A, but plenty of older utilities, some transcription tools and various corporate systems do not. Android voice apps and Linux desktop tools are similarly hit and miss.

Converting to WAV removes the whole class of problem, because uncompressed PCM in a RIFF container is the one audio format that essentially everything written in the last thirty years can read.

What this conversion does and does not achieve

Decoding turns the compressed AAC frames back into a list of samples. A WAV stores that list with a 44-byte header and nothing else. So the conversion is a decode followed by a write, and there is no quality setting because there is no quality decision to make.

It does not improve the audio. What the AAC encoder discarded when the recording was made is gone permanently, and storing what remains in a bigger file preserves it exactly — artefacts included. This is why this page does not describe the result as lossless. No second lossy compression happens, which is a real and checkable guarantee, but it is a narrower claim than lossless and the difference matters.

The size is the cost. Uncompressed 16-bit PCM at 44.1 kHz in stereo is about 10.1 MB per minute; at 48 kHz it is about 11.0 MB. A twenty-minute voice memo that was 10 MB as an M4A comes out around 100 MB in mono. The page shows the exact figure and the multiple over your source before you start.

The sample rate trap, and why nothing is uploaded

There is a mistake that is easy to make in a browser and impossible to notice as a user. A browser decodes audio at whatever rate its audio context was created with, typically 48,000 samples per second. Decode a 44.1 kHz recording that way and the result has been resampled: the output is valid, it plays, the length is right, and the tool has quietly processed your audio while claiming not to have touched it.

This page reads the sample rate out of the MP4 structure first — the media header of the audio track — and decodes at exactly that rate, so no resampling happens. It reports the rate it used, and if the structure cannot be read it says which rate it fell back to instead of letting you assume the number came from your file.

One more thing gets reported honestly here. AAC encoders put a short run of warm-up samples in front of the real content, and whether the browser strips them is not something we control. When the length the container declares and the length that comes out of the decoder disagree by more than a few milliseconds, the page shows both numbers rather than picking one.

None of this needs a server. Decoding is your browser’s own, and the WAV is assembled in the tab: open your developer tools, watch the Network tab and convert a file. The Content-Security-Policy this site sends restricts connections to our own origin and your browser enforces it. No account, no queue and no watermark. The limits that exist come from your own device rather than from us, and the page states them up front: 300 MB per file and 30 minutes of audio on a desktop, half on a phone. After the first visit it works offline.