Extract Audio from MPG to AU — Free Online Tool
Extract audio from MPG video files and save it as AU format — a raw PCM audio file using Sun's legacy Unix container. The conversion decodes the MPG's MP2 or MPEG audio stream and re-encodes it as 16-bit big-endian PCM (pcm_s16be), producing uncompressed audio ideal for Unix-based workflows and legacy audio tools.
to
FFmpeg Command
Copy this command to run the same conversion locally with FFmpeg on your desktop. Download FFmpeg
Drop your MPG file here
or click to browse
Free — no uploads, no signups. Your files never leave your browser.
Settings
Note: Browser-based encoding uses approximate quality targets. For precise CRF compression, copy the FFmpeg command above and run it on your desktop.
Estimated output:
Conversion Complete!
DownloadHow It Works
MPG files typically carry audio encoded in MP2 (MPEG-1 Audio Layer II) or occasionally AAC, wrapped in an MPEG-1 or MPEG-2 program stream container. During this conversion, FFmpeg discards the video stream entirely and decodes the compressed MPG audio into raw PCM samples. Those samples are then written into a Sun AU container using 16-bit signed big-endian PCM encoding — the AU format's default and most compatible codec. Unlike remuxing, this is a full transcode of the audio: the lossy MPG audio is decoded to an intermediate PCM representation and then stored uncompressed in AU. The result is a lossless-quality snapshot of whatever audio survived the original MPG encoding, with no further generation loss introduced by this conversion.
What Each Flag Does
| Flag | What it does |
|---|---|
ffmpeg
|
Invokes the FFmpeg binary — the underlying engine that both the browser-based tool (via WebAssembly) and your local desktop installation use to perform this conversion. The exact same command runs in both environments. |
-i input.mpg
|
Specifies the input MPG file. FFmpeg reads the MPEG-1 or MPEG-2 program stream, identifying all contained streams including the video track (mpeg1video or mpeg2video) and the audio track (typically MP2) that will be extracted. |
-vn
|
Stands for 'video none' — instructs FFmpeg to exclude all video streams from the output. Since AU is a pure audio format, this flag is essential; without it, FFmpeg would attempt to write video into the AU container and fail. |
-c:a pcm_s16be
|
Sets the audio codec for the output to 16-bit signed big-endian PCM — the native and default encoding of the Sun AU format. This fully decodes the compressed MP2 audio from the MPG and re-encodes it as uncompressed samples in the byte order AU expects. |
output.au
|
Specifies the output filename with the .au extension, which tells FFmpeg to use the Sun AU muxer. The resulting file will be a valid AU container with a standard header followed by raw 16-bit big-endian PCM audio data. |
Common Use Cases
- Extracting dialog or music from VCD or DVD-sourced MPG files for use in Unix/Linux audio processing pipelines that expect raw PCM in AU format
- Feeding extracted MPG audio into legacy Sun workstation software or audio tools like SoX that natively read AU files
- Archiving the audio track from broadcast MPEG-2 recordings as uncompressed AU for long-term preservation without further lossy encoding
- Preparing audio samples from MPEG video for use in early internet or multimedia applications built around the .au file format
- Stripping and converting MPG audio to AU as an intermediate step before further processing in tools like Audacity or command-line signal processors that prefer raw PCM containers
- Developers testing or debugging MPEG audio decoding pipelines who need a simple, headerless-adjacent PCM output in AU format for byte-level inspection
Frequently Asked Questions
No additional quality loss is introduced by this specific conversion. The MPG audio (typically MP2) is lossy and was already degraded when originally encoded. FFmpeg fully decodes it to raw PCM and writes it into the AU container as pcm_s16be — an uncompressed format. You are essentially capturing the best possible audio that exists in the MPG file, with no second generation of lossy compression applied.
MP2 audio in MPG files is compressed, typically at 128–192 kbps, which is much smaller than uncompressed PCM. The AU format stores audio as raw 16-bit PCM samples, so a stereo stream at 44.1 kHz requires about 1.4 MB per second. Expect the AU file to be significantly larger than the compressed audio it was derived from — this is normal and expected for any conversion to an uncompressed format.
The output uses pcm_s16be — 16-bit signed PCM with big-endian byte order, which is the Sun AU format's standard and most widely supported encoding. Most audio tools on Linux, macOS, and even Windows can read this format, including SoX, Audacity, and ffplay. However, some consumer applications may not recognize the .au extension; if compatibility is an issue, converting further to WAV (which uses little-endian PCM) is straightforward.
No. The Sun AU format has an extremely minimal header — it stores only technical parameters like sample rate, channel count, and encoding type. It has no standardized metadata fields for ID3-style tags or other descriptive information. Any title, artist, or copyright metadata embedded in the MPG file will be lost in the AU output.
Replace '-c:a pcm_s16be' with '-c:a pcm_mulaw' or '-c:a pcm_alaw' in the command. For example: 'ffmpeg -i input.mpg -vn -c:a pcm_mulaw output.au'. Mu-law and A-law are logarithmic PCM encodings historically used in telephony; they reduce file size compared to pcm_s16be but at lower fidelity. The AU format also supports pcm_s8 and pcm_u8 for 8-bit output if you need even smaller files.
Yes. On Linux or macOS, wrap the command in a shell loop: 'for f in *.mpg; do ffmpeg -i "$f" -vn -c:a pcm_s16be "${f%.mpg}.au"; done'. On Windows Command Prompt, use: 'for %f in (*.mpg) do ffmpeg -i "%f" -vn -c:a pcm_s16be "%~nf.au"'. This processes every MPG in the current directory and saves a corresponding AU file. The in-browser tool handles one file at a time, so the command line is the recommended approach for batch jobs.
Technical Notes
The Sun AU format (.au, .snd) was developed by Sun Microsystems in the late 1980s and became the first widely distributed internet audio format, largely because of its simple fixed header and self-describing structure. It supports several PCM encodings: pcm_s16be (the default used here), pcm_s8, pcm_u8, pcm_alaw, and pcm_mulaw. Notably, the format uses big-endian byte order — the opposite of WAV's little-endian convention — which is why pcm_s16be is the natural fit. MPG files sourced from VCD typically carry 44.1 kHz stereo MP2 audio at 224 kbps, while MPEG-2 broadcast streams may use 48 kHz. The sample rate and channel configuration from the MPG source are preserved in the AU output without resampling. There is no support for multiple audio tracks — if the MPG contains more than one audio stream, FFmpeg will select the default (usually the first) stream. The AU format does not support chapter markers, subtitle streams, or embedded images, so none of those elements survive the conversion. For files larger than 2 GB, the AU format has no inherent size limitation in the header (unlike early WAV), making it theoretically suitable for very long recordings, though practical tool support for large AU files varies.