FLV to AU Converter

FLV to AU: What This Converter Actually Produces

FLV is Adobe's Flash Video container — a relic of the YouTube-and-Flash era, still readable by VLC and ffmpeg long after Adobe ended Flash Player support on December 31, 2020. AU is older still: Sun Microsystems' audio format from the late 1980s, the one Java's audio APIs and classic Unix sound pipelines read natively, and the reason the alternate extension .snd exists. This page takes the soundtrack out of an FLV, discards the picture, and writes it as an .au file encoded in 8-bit G.711 mu-law — AU's original telephony encoding. Knowing that up front matters, because it explains both why the output can be larger than the audio inside the FLV and why it is not any higher in fidelity.

The AU Format at a Glance

How to Convert FLV to AU

1. Upload Your FLV File: Drag and drop or click "Add Files" to select FLV videos from your computer. Batch is supported — queue several clips and they convert one after another, and only the audio track is kept.
2. Set the Audio Sample Rate: Audio Sample Rate offers Original, 8000, 12000, 16000, 24000, 44100 and 48000 HZ. This is the main size control on the page: for Java AudioClip playback or a telephony prompt, pick 8000 HZ.
3. Set the Audio Channel (Optional): Audio Channel offers Original, MONO or STEREO. Mono halves the file, and mono is what classic Unix system sounds and phone prompts expect anyway.
4. Trim and Convert: Switch Trim from Unchanged to Time Range and enter a Start time and Duration — in seconds or as HH:MM:SS.sss — if you only need one segment, then click Convert. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. No sign-up, no watermark.

Walk-through: Sample Rate and Channels Are Your Only Size Controls

There is no quality slider, bitrate box, or codec picker on this conversion, and that is a property of the target rather than a gap in the tool: mu-law AU is a fixed one-byte-per-sample encoding, so its size is exactly sample rate multiplied by channel count multiplied by duration. That makes the arithmetic completely predictable, and it makes Audio Sample Rate and Audio Channel the only two levers you have.

One honesty note that matters here: an FLV's audio track is almost always MP3 or AAC, and both are lossy — perceptual detail was thrown away when the FLV was first encoded. Decoding that into mu-law AU makes the file larger, but it cannot put back what the original codec discarded, and mu-law companding shaves a little more off on the way through. What you get is a bulky, legacy-compatible copy of already-lossy audio, not new fidelity.

Common Errors and How to Fix Them

• "Output is silent" — the FLV may be video-only, or its audio track uses a codec the decoder skipped. Play the source in VLC first to confirm it actually has sound.
• "The AU file is huge for a short clip" — expected at high sample rates. Keeping Original on a 44.1 kHz stereo soundtrack costs about 5.3 MB per minute; switching to 8000 HZ and MONO brings the same minute down to roughly 480 KB.
• "My player will not open the .au" — most modern media players dropped AU support years ago. VLC, Audacity and Java-based tools still read it. If you only need something playable, convert FLV to MP3 instead.
• "The upload is slow or times out" — a large FLV archive has to upload before it can be processed, so upload time is the real bottleneck, not the encode. Trim to the segment you need first.
• "The picture is gone" — intended. This tool extracts audio only and discards the video stream entirely.

When AU Is the Wrong Target

If your goal is a small, broadly playable file to share or keep on a phone, AU is the wrong choice — it is a niche legacy format, honestly useful today only for Java audio resources, older Unix audio pipelines, telephony prompts, and coursework that distributes reference clips as .au because the 24-byte header is trivial to parse by hand. For everything else, convert FLV to MP3 produces a far smaller, universally playable file, and convert FLV to WAV gives you uncompressed 16-bit audio that every editor accepts. Corrupted or partially downloaded FLVs — a common fate for old streamed Flash content — may decode with glitches or stop early, and no converter can recover audio that was never fully written.

Frequently Asked Questions

Which encoding does the AU output use, and can I change it?

It is 8-bit G.711 mu-law, and no — there is no codec selector on this page. We checked the option set: the Audio Codec group renders with no selectable entries for an AU target, so mu-law is what you get every time. The AU container itself can hold other encodings, including 8-bit A-law and 16-bit big-endian linear PCM, and the header's encoding field records which one was used; this conversion simply always writes the mu-law variant that legacy "Sun audio" tools expect.

Why is my AU file bigger than the FLV's audio if the video was removed?

Because the FLV's audio is MP3 or AAC — lossy, compressed formats typically running at 96-192 kbps — while mu-law AU spends one byte per sample per channel, about 706 kbps at 44.1 kHz stereo. So a minute of soundtrack that occupied roughly 1 MB inside the FLV comes out as roughly 5.3 MB of AU. Those extra bytes are not extra fidelity; they are an uncompressed copy of audio that already lost detail. Choosing 8000 HZ and MONO produces an AU that is genuinely smaller, about 480 KB per minute.

Can converting an FLV to AU improve the audio quality?

No. The FLV's audio was encoded with a lossy codec, so the detail discarded at that stage is gone permanently. Writing it into AU gives you a bigger file containing the same lossy audio, and the mu-law encoding on this page is telephony-grade rather than hi-fi, so it removes a little more. If you need the best possible quality, start from the original master rather than an FLV.

Why does AU use big-endian byte order when modern formats are little-endian?

Sun's SPARC and NeXT's Motorola 68k workstations were big-endian, so when the format was specified the header words and any multi-byte samples were stored big-endian to match native CPU memory layout. Modern x86 and ARM machines are little-endian, so their audio libraries byte-swap transparently on read. It only becomes a concern in low-level code that parses the 24-byte header by hand — which is exactly why AU still turns up in teaching material.

Does the sample rate change how the audio sounds, or just the size?

Both, and in that order. Mu-law is a fixed encoding, so the sample rate sets the file size directly and also sets the highest frequency the file can represent — roughly half the sample rate. At 8000 HZ you get telephone-band speech, which is fine for a voice prompt and poor for music. At 44100 HZ you keep the full range the FLV's soundtrack had, at about eleven times the bytes.

Is AU still worth using in 2026, or should I pick something else?

For a new project, pick MP3, WAV or FLAC — AU is niche today. It survives where it is specifically required: Java audio APIs still ship AU support, older telephony stacks expect 8 kHz mu-law, and many DSP and computer-science courses distribute reference clips as .au because the short header is easy to parse. If a tool, textbook or vendor explicitly asks for "AU" or "Sun audio", this conversion feeds a Flash-era recording into it. Otherwise FLV to MP3 is the better target.

Can I extract only part of the FLV's soundtrack?

Yes. Switch Trim from Unchanged to Time Range and set a Start time and a Duration, in plain seconds or as HH:MM:SS.sss. Because mu-law size is strictly proportional to duration, trimming is the most effective way to control the output size after sample rate and channel count — cutting a five-minute recording to thirty seconds cuts the file by the same factor.

How are my files handled, and how long are they kept?

Files are uploaded over an encrypted connection and processed on our servers. They are deleted automatically after a few hours and are never shared or made public. There is no sign-up and no watermark on the output. The practical limit on large FLV archives is upload time, since the decode and re-encode both run on our servers.