M4A to AU Converter

How to Convert M4A to AU Online

1. Upload Your M4A File: Drag and drop or click "Add Files" to select an.m4a (MPEG-4 Audio with AAC or ALAC) file. Batch conversion is supported, so you can queue several tracks at once.
2. Pick Audio Channel: Default is Original, which preserves whatever channel layout the M4A carries. Choose Mono to halve the output size (common for system sounds, telephony, and Java alert clips) or Stereo to keep both channels for music playback.
3. Set Audio Sample Rate (Optional): Default keeps the source rate. Pick 8000 Hz for telephony-style μ-law clips, 22050 Hz for low-bandwidth voice, 44100 Hz to match CD-quality M4A music, or 48000 Hz for broadcast-grade audio. Lower rates cut file size proportionally.
4. Trim and Download: Toggle Trim on to extract a segment — enter a Start Time and Duration in seconds or HH:MM:SS.sss format. Click "Convert" and the AU file lands on our servers. No sign-up, no watermark, files are removed from our servers after processing.

Why Convert M4A to AU?

M4A is Apple's MPEG-4 audio container, almost always wrapping AAC-LC (the iTunes Store default since 2003) or Apple Lossless (ALAC). AU — sometimes called Sun audio or NeXT/Sun soundfile — is the legacy format Sun Microsystems shipped with the SPARCstation 1 in 1989 and is still the path of least resistance whenever the playback target is a JVM, a Solaris box, or an old SPARC/NeXT machine.

• Java audio playback — javax.sound.sampled.AudioSystem reads AU, WAV, and AIFF out of the box on every JRE. AAC inside an M4A container needs a third-party SPI (JAAD, JLayer-AAC) or FFmpeg shelling out; AU just works in stock Java applets, desktop apps, and server-side audio pipelines.
• Unix/Linux system sounds and alerts — Solaris, the legacy NeXTSTEP/OPENSTEP environment, and many older Linux desktop sound themes ship AU files at 8000 Hz μ-law for short beeps and bells.
• Telephony and IVR prompts — 8 kHz μ-law AU matches the G.711 wire format used by SIP and PSTN gateways, so the file plays without resampling. M4A would need to be transcoded inside Asterisk, FreeSWITCH, or Twilio Studio before it can be played to a caller.
• Scientific and DSP workflows — MATLAB's audioread, Octave, Praat (linguistics/phonetics), and NumPy's SciPy wavfile-style loaders accept AU's PCM payload directly. AAC's lossy decoder smears the spectrum, which is unacceptable for spectral analysis or speech research.
• Archival of voice recordings — uncompressed PCM AU is decoder-agnostic, so a 30-year-old AU file recorded on a SPARC will still open in 2026 without hunting down an AAC decoder license.
• Format requirements for legacy tools — Audacity, SoX, FFmpeg, and the NeXT Music Kit all read AU natively. Some old build pipelines and university courseware still emit AU-only output, and an M4A drop-in needs converting first.

M4A vs AU — Format Comparison

AU Sample Rate and Encoding Guide

A 3-minute M4A at 128 kbps is roughly 2.8 MB. Converted to 44.1 kHz 16-bit stereo PCM AU, the same clip is about 31 MB — eleven times larger because PCM stores every sample uncompressed. Drop to 8000 Hz mono μ-law and the AU shrinks to roughly 1.4 MB while remaining readable by every Java and Solaris tool that expects classic AU.

If you also need other conversions, see M4A to WAV, M4A to MP3, or WAV to AU.

Frequently Asked Questions

Why is the AU file so much larger than the M4A?

M4A uses AAC, a perceptual codec that throws away inaudible spectral content and typically compresses to 128–256 kbps. AU stores audio either as 8-bit μ-law (around 64 kbps) or as uncompressed PCM (705 kbps for 44.1 kHz/16-bit mono, 1411 kbps for stereo). Picking PCM at CD rates produces a file roughly 10–12 times the size of the source M4A; that is the cost of a lossless, decoder-free container. To minimise size, switch Audio Channel to Mono and drop Audio Sample Rate to 8000 Hz, which is what classic AU files used.

Does this converter produce μ-law AU or linear PCM AU?

The converter outputs PCM-encoded AU (FFmpeg's default pcm_s16be mapping for the.au extension), which matches the modern interpretation of the format and what Java's AudioInputStream decodes without extra SPIs. If you specifically need 8-bit μ-law for a telephony pipeline or an authentic SunOS-era clip, set Audio Sample Rate to 8000 Hz and Audio Channel to Mono — the resulting file is the same size as a μ-law file (1 byte per sample) and is readable by every tool that expects classic AU.

Will Java's javax.sound.sampled play the output?

Yes. The JRE bundles AU support since Java 1.3, so AudioSystem.getAudioInputStream(file) opens the output directly without third-party libraries. AAC inside an M4A would otherwise require a SPI such as the Java AAC Player (JAAD2) or shelling out to FFmpeg. If you are building Java applets, JavaFX MediaPlayer apps, or server-side audio tooling, AU is the lowest-friction format.

Does AU preserve the M4A's metadata (artist, title, album art)?

No. AU has only an optional free-form annotation field after its 24-byte header — there is no equivalent to iTunes-style ID3 or MP4 atom tags, and album art is not representable. If you need to keep that metadata, convert to FLAC or WAV with INFO chunks instead, or keep a copy of the M4A alongside the AU.

Can I trim long M4A files before converting?

Yes. Toggle Trim, enter a Start Time (e.g. 00:01:30) and a Duration (e.g. 00:00:45) in HH:MM:SS.sss format. Trimming happens before encoding, so the output AU is just the selected segment — useful for extracting a 5-second alert from a 30-minute podcast without paying the file-size penalty for the full track.

Why would I still use AU in 2026?

Three reasons keep AU alive: (1) Java's standard library reads AU but not AAC; (2) telephony and IVR systems still speak G.711 μ-law, which is byte-for-byte compatible with classic AU; (3) academic phonetics tools (Praat, MATLAB audio toolbox) prefer uncompressed inputs, and AU is a smaller, simpler container than WAV when you only need one PCM stream and no metadata.

Is the conversion lossless from M4A's perspective?

The PCM step is lossless — every sample the M4A decoder produces is written verbatim to the AU. However, M4A with AAC is itself a lossy format, so the audio you end up with reflects whatever AAC already discarded at encode time. If the M4A is actually ALAC (Apple Lossless), the round-trip M4A→AU is fully lossless. To tell which codec your M4A uses, run ffprobe -show_streams file.m4a and look at the codec_name field.

Does my browser need to upload the file to a server?

Yes — large audio conversions need server-side FFmpeg-class processing, so the M4A is uploaded over HTTPS, converted, and the AU is returned to your browser. The uploaded file is removed from our servers after the job completes; we do not retain audio. If you need fully offline conversion, ffmpeg's -i input.m4a output.au works locally too.

What sample rates and bit depths does AU actually support?

The format spec defines 27 encoding codes covering 8-bit μ-law and A-law, linear PCM at 8/16/24/32-bit, IEEE float at 32 and 64-bit, and several G.7xx ADPCM variants. Sample rate is stored as a 32-bit field, so anything from 1 Hz to 4 GHz is technically legal. In practice, the rates listed in the dropdown (8000, 11025, 12000, 16000, 22050, 24000, 44100, 48000 Hz) cover every real-world use case.