CAVS to AIFF Converter

CAVS to AIFF: Read This Before You Upload

A .cavs file is a raw Chinese AVS video elementary stream, and an elementary stream carries exactly one media type. By definition there is no audio track inside it — FFmpeg registers the format as raw video, with a demuxer called cavsvideo feeding a decoder called cavs (AVS1-P2, JiZhun profile). So if your goal is to pull a soundtrack out of a bare .cavs and save it as AIFF, the honest answer is that there is usually nothing to pull, and the conversion will report an error rather than hand you a silent file. This page explains why, shows what to do instead, and covers the case where your file is not really a bare CAVS at all.

How to Convert CAVS to AIFF

1. Upload Your CAVS File: Drag and drop the .cavs onto the page, or click "+ Add Files" to browse. Several files can be queued and processed in one batch with the same settings.
2. Pick the Audio Codec: Audio Codec offers PCM 16-bit Big Endian — the default, and the byte order AIFF has used since its Motorola-era origins — plus PCM 16-bit Little Endian, PCM A-law and PCM mu-law. There is no bitrate control here because PCM is uncompressed: the size follows from the sample rate, bit depth and channel count.
3. Set Audio Channel, Audio Sample Rate or Trim (Optional): Leave Audio Channel and Audio Sample Rate on "Original" to mirror whatever the source actually contains, or pin them to Mono/Stereo and a fixed rate such as 44100 Hz or 48000 Hz. Trim takes a start time and duration if you only want part of the timeline.
4. Convert and Download: Click "Convert" and save the AIFF. No sign-up, no watermark.

Walk-through: Why a Bare CAVS Has No Audio to Extract

AVS — the Audio Video coding Standard — is the digital media compression family developed by China's AVS Working Group, established in June 2002. The video part you have, AVS1-P2 (the JiZhun or "base" profile), was promulgated as the Chinese national standard GB/T 20090.2-2006 in February 2006, and a later broadcast-oriented revision, AVS+, followed as GB/T 20090.16-2016. It is a block-based, motion-compensated design in the same broad family as MPEG-2 and H.264, and it shows up mainly in Chinese digital-TV, IPTV and set-top-box equipment.

The word doing the work here is elementary. An elementary stream is the encoder's raw output for a single track: compressed video frames, sequence headers, and nothing else. It has no container, which is also why it carries no timestamps, no track table and no metadata. In an AVS broadcast chain the sound is encoded separately and joined to the video only when the transport or program stream is muxed at the end. So:

• A spec-correct .cavs holds video and only video. Our converter runs the file through a straight decode-and-encode pass, and there is no silence-synthesis step anywhere in it. With no audio packets arriving, there is nothing to encode, so the job ends in an error instead of writing a silent AIFF. No setting on the page changes that — you cannot configure a soundtrack into existence.
• A file named .cavs that plays with sound is not a bare elementary stream. It is almost certainly a mis-named transport or program stream capture with the audio muxed in. That file converts to AIFF normally. Quick test: open it in VLC. Sound means there is audio to extract; silent video means there is not.

Where Your Audio Actually Lives

AIFF Output Options

Common Errors and How to Fix Them

• "The conversion failed / no output file." The most likely cause by far: the source is a genuine video-only CAVS, so the audio encoder received no packets. Confirm the file plays with sound in a media player before converting it to any audio format.
• "I wanted the picture, not the audio." AIFF is an audio-only container, so the video is discarded by design. Use CAVS to MP4 to keep the footage in a container ordinary players open.
• "My player won't open the .cavs at all to check it." Most mainstream players do not ship an AVS1-P2 decoder. FFmpeg does — its cavs decoder handles the JiZhun profile — which is why a server-side converter is the practical route. VLC is FFmpeg-based, so it is the easiest thing to test with.
• "The AIFF is enormous." That is PCM behaving normally: uncompressed 16-bit/44.1 kHz stereo runs roughly 10 MB per minute regardless of what the audio sounds like. Trim first, or choose a compressed target instead.
• "A long broadcast capture takes forever." The wait is the upload over your own connection, not the conversion. Trim to the segment you need before uploading.

When This Doesn't Work

If your file really is a video-only CAVS, no tool anywhere can produce audio from it — the data was never written into the file. The two genuine routes are to find the separate audio stream that travelled with the video (frequently AC-3 in AVS broadcast chains) and convert that, or to convert the complete muxed container — a transport stream or program stream — which holds both tracks. It is worth knowing that many "CAVS to AIFF" tools accept the upload and fail silently or hand back an empty file without explaining any of this. The empty result is the format talking, not the converter misbehaving.

Frequently Asked Questions

Why did my CAVS to AIFF conversion fail or produce nothing?

Because a bare .cavs is an AVS1 video elementary stream with no audio track inside it. The converter decodes what it is given and encodes it to PCM; with no audio packets there is nothing to encode, so the job errors out rather than fabricating silence. Convert the muxed broadcast container or the separate audio stream instead — those genuinely contain the soundtrack.

What exactly is a .cavs file?

A raw video bitstream encoded with AVS1, China's national video coding standard. The video part, AVS1-P2 (JiZhun profile), was promulgated as GB/T 20090.2-2006 in February 2006 by the AVS Working Group, which was established in June 2002. Despite the word "Audio" in the standard family's name, the .cavs extension refers to the video elementary stream specifically — the audio parts of AVS are separate specifications carried in separate streams.

How can I tell whether my file actually has audio before uploading?

Open it in VLC, which uses the same FFmpeg-based decoders. If you hear sound, the file is muxed — despite the extension — and it will convert to AIFF cleanly. If it plays as silent video, it is a true elementary stream and no audio target will produce anything. Checking first saves you an upload and a failed job.

What does this converter write into the AIFF, exactly?

Uncompressed PCM. The default codec is PCM 16-bit Big Endian — 16-bit is CD depth, and big-endian is the byte order AIFF has used since it was designed for Motorola 68000 Macs. A-law, mu-law and little-endian PCM are available if a specific tool needs them. Because PCM is uncompressed there is no bitrate setting: the file size is fixed by sample rate times bit depth times channels, which works out to roughly 10 MB per minute at 16-bit/44.1 kHz stereo.

Is AIFF lossless?

Yes, as written here. PCM stores the decoded samples verbatim, so nothing is discarded in the AIFF stage and the file can be edited and re-exported without generational loss. The caveat is upstream, as always: if the audio you extracted was lossy in the source container, the AIFF preserves that already-degraded waveform exactly rather than improving it.

Why does hardly anything play a .cavs file?

Because AVS1-P2 decoders shipped mainly in Chinese digital-TV, IPTV and set-top-box hardware, and desktop media software outside that market rarely bundled one. FFmpeg's open-source cavs decoder is the main reason these files can be handled at all on ordinary machines, and it is what VLC and this converter both rely on.

I want the video from my CAVS file — what should I use?

A video target rather than an audio one. CAVS to MP4 re-encodes the picture into a container that ordinary players, phones and browsers open, and the general Video Converter covers other output formats. Use this page only when you have a real audio source to work from.

How are my uploaded files handled?

Your file is uploaded over an encrypted connection, processed on our servers, and deleted automatically a few hours later. There is no sign-up, no watermark, and files are never shared or made public. On a large broadcast capture the realistic bottleneck is upload time on your own connection rather than the conversion itself.