Initializing... drag & drop files here
Supports: AU
Move a legacy Sun .au (or .snd) recording into OGG, the open Xiph.Org container that game engines, open-source tooling and the modern web all understand. The output is encoded with Vorbis by default — the royalty-free codec .ogg is best known for — turning a Unix-workstation relic with no metadata and patchy player support into a small, taggable, widely decoded file. What the conversion cannot do is add fidelity, and with AU that matters more than usual: the classic Sun payload is telephone-grade 8-bit µ-law, and no encoder brings back what companding threw away.
.au or .snd file onto the page, or click "+ Add Files" to browse. Sun/SPARCstation system sounds, early Java applet audio and old Unix speech archives all work, and several files can be queued and converted with the same settings..ogg file individually or as a ZIP. Files upload over an encrypted connection, are converted on our servers and are deleted automatically a few hours later. No sign-up, no watermark.Ogg is a container, not a codec — the Audio Codec control decides what actually gets written into it. Vorbis is the default and the right answer for most people.
| Codec | What it is | Choose it when |
|---|---|---|
| Vorbis (default) | Xiph.Org's lossy codec from 2002, the one the .ogg extension is synonymous with |
You want a normal .ogg that drops straight into game engines, open-source tooling and web players |
| Opus | The newer IETF-standard codec (RFC 6716, 2012); more efficient than Vorbis at every bitrate, especially for speech | The target is fresh voice, streaming or web audio — or go to AU to Opus for a .opus file directly |
| FLAC | Lossless compression inside an Ogg wrapper | You want lossless but specifically need Ogg framing; for a normal lossless archive use AU to FLAC instead |
AU files come in two very different flavours and the outcome depends entirely on which one you hold. Vorbis is lossy: it stores audio efficiently, never adds detail the original capture lacked.
| Your AU payload | What it really is | What the OGG gives you | Sensible setting |
|---|---|---|---|
| 8-bit µ-law / A-law @ 8 kHz | Sun's telephone-grade default; usable bandwidth capped near 4 kHz | The same intelligible speech at a fraction of the bits — no highs or dynamic range return | 48–64 kbps, mono |
| 16-bit linear PCM @ 44.1 kHz | An effectively lossless CD-grade master | A clean first-generation Vorbis encode; this is where the codec shines | 96–128 kbps stereo is transparent for most listeners |
| 24/32-bit linear PCM | A high-headroom master from a later tool | A good listening copy, but the extra depth is discarded | 160–192 kbps, or archive losslessly instead |
| You need every sample kept | — | Vorbis discards data by design | Use AU to FLAC or AU to WAV |
One pipeline detail worth knowing before you convert: we restrict Vorbis encoding to 22,050, 32,000, 44,100 and 48,000 Hz, because libvorbis becomes unreliable at very low sample rates paired with higher bitrates. A classic 8 kHz µ-law AU is therefore resampled up to 22,050 Hz on the way in, whether you leave Audio Sample Rate on ORIGINAL or pick 8000 HZ yourself. That resampling costs nothing in quality — it simply cannot add any either.
No, and no format can. If the AU holds the classic 8-bit µ-law or A-law payload (Sun's companded default at 8 kHz mono), Vorbis reproduces that intelligibility efficiently but cannot restore the highs and dynamic range the compander discarded when the recording was first made. If it holds linear PCM, you get a clean first-generation encode of an effectively lossless master — that is the case where OGG shines. What you gain either way is size, tagging and software support, not fidelity.
Because that is the lowest rate we let libvorbis encode at. Vorbis is technically capable of 8 kHz, but the encoder behaves unpredictably at very low sample rates when a higher bitrate is requested, so our pipeline snaps any requested or detected rate to the nearest of 22,050, 32,000, 44,100 or 48,000 Hz — and 8,000 lands on 22,050. The audio is resampled, not re-synthesised: it still contains only what the 8 kHz original captured, and it sounds the same. If keeping the literal 8 kHz rate matters to you, AU to Opus accepts 8, 12, 16, 24 and 48 kHz, and AU to FLAC preserves the source rate exactly.
For anything that expects a .ogg, keep Vorbis. It is what game engines such as Unity, Unreal and Godot consume, what a great deal of open-source software assumes when it sees the extension, and what the widest range of existing .ogg tooling reads without complaint. Opus is the better codec on the numbers — more efficient at every bitrate and dramatically better for speech below 64 kbps — so pick it when the destination is new voice or streaming work rather than an existing Ogg pipeline. For a plain .opus file, use AU to Opus instead.
Less than instinct suggests, because the source usually cannot justify much. A µ-law AU is a 64 kbps stream to begin with and holds nothing above roughly 4 kHz, so 48–64 kbps mono captures all of it and higher settings only inflate the file. A 16-bit/44.1 kHz linear-PCM AU is a genuine master, and 96–128 kbps stereo Vorbis is transparent for most listeners on most material. When in doubt, leave Quality Preset alone and let the encoder scale to the source rather than forcing a number.
VLC, foobar2000, Audacity, Winamp, Clementine and most Linux desktop players open Ogg Vorbis natively, and Firefox, Chrome and Edge play it in the <audio> element. Android has supported Vorbis since its early releases. The notable holdout is Apple: Safari and the stock iOS/macOS players do not play Ogg Vorbis, so if the file is destined for an iPhone or Apple Music, convert to AU to M4A or AU to MP3 instead.
More than most legacy conversions, yes — this is a genuine modernisation rather than swapping one dated wrapper for another. A Sun/NeXT workstation relic carrying audio/basic, with no tag fields and support limited to VLC-class tools, becomes an open, royalty-free file that current software actively reads and that can carry title, artist and album metadata. You trade a small amount of fidelity (nothing at all, on a µ-law source) for a file the modern ecosystem understands.
No. Vorbis discards data by design, which makes it the wrong choice for preservation. If your .au holds linear PCM and you care about the master, keep the original file, or convert with AU to FLAC for lossless compression with tagging, or AU to WAV for a plain uncompressed working copy — both preserve every sample exactly. For a µ-law AU there is little to preserve beyond the speech itself, so a low-bitrate OGG is usually all anyone needs.
Yes, OGG to AU reverses the direction, but it is not a round trip to the original quality. Once audio has been through Vorbis, rebuilding an .au around it wraps already-lossy audio in an older container — and our AU output defaults to 8-bit G.711 µ-law unless you choose a PCM encoding, which would compand it again on top. Keep the source .au if you may need its fidelity later.
Your file is uploaded over an encrypted connection, converted on our servers, and deleted automatically a few hours after the job finishes — never shared, never made public, with no sign-up and no watermark. The practical limit on long recordings is upload size and time rather than the encode: a 60-second 8 kHz mono µ-law AU is only about 480 KB, while a 60-second 16-bit/44.1 kHz stereo PCM AU is about 10.6 MB before Vorbis compresses it.