Initializing... drag & drop files here
Supports: CAVS
WAV is the standard way to hand raw, uncompressed audio to an editor, a DAW or a transcription tool — no codec, no quality setting, just linear PCM samples in a RIFF wrapper. The wrinkle on this page is what a .cavs file actually is: a raw Chinese AVS (AVS1) video elementary stream, which carries picture data and nothing else. This page explains both formats, how the extraction is configured, and what happens when the source has no audio to give.
| Property | Value |
|---|---|
| Standard | GB/T 20090.2-2006 (AVS1 video); reissued as IEEE 1857-2013 |
| Promulgated | February 2006 as a Chinese national standard |
| Developed by | AVS Workgroup, China (founded June 2002) |
| Payload | AVS1-P2 coded video, Jizhun profile |
| Container | None — a bare elementary stream, no index, no header |
| Audio | None; the AVS family specifies audio in separate parts of the series |
| Typical origin | Chinese digital TV equipment, or video demuxed out of a .ts or .mp4 |
| Succeeded by | AVS2, published by the IEEE as 1857.4-2018 |
| Property | Value |
|---|---|
| Container | RIFF, introduced by Microsoft and IBM in 1991 |
| Payload | Linear PCM by default; A-law and mu-law also selectable here |
| Compression | None in PCM mode — samples are stored verbatim |
| Typical rate | 44.1 kHz, 16-bit stereo = 1,411 kbps, about 10 MB per minute |
| Size ceiling | Roughly 4 GiB, because RIFF records chunk sizes in 32-bit fields |
| Best for | Editing masters, DAW imports, transcription, sample libraries |
| Weak spot | Size; it is the wrong format for sharing or streaming |
.cavs onto the page, or click "Add Files" to browse for it. Several files can be queued and run with the same settings..wav. No sign-up, no watermark.| Audio Codec | Bits per sample | Size vs 16-bit | Use it when |
|---|---|---|---|
| PCM 16-bit Little Endian (default) | 16 | baseline | Anything normal — this is the CD and DAW standard |
| PCM 24-bit Little Endian | 24 | 1.5× | Handing a master to a studio pipeline that works in 24-bit |
| PCM 32-bit Little Endian | 32 | 2× | Maximum headroom for further processing |
| PCM A-law | 8, logarithmic | 0.5× | Telephony interchange, European systems |
| PCM mu-law | 8, logarithmic | 0.5× | Telephony interchange, North American and Japanese systems |
Depth changes the file size linearly and nothing else: 44.1 kHz 16-bit stereo runs about 10 MB per minute, so the same minute at 24-bit is about 15 MB. Raising the depth cannot add detail that was not captured — it only preserves what is already there without rounding.
The extraction selects the audio stream and discards the picture. When the upload contains no audio stream at all — which is the definition of a raw .cavs — there is nothing to select, so the encoder stops with an error and no file is written. Nothing is silently produced and nothing is lost. The soundtrack that belongs with that footage is still inside the container the video was demuxed from: Chinese AVS broadcast and optical-media content is multiplexed, with the coded video and a separate coded audio track interleaved inside an MPEG transport stream or an MP4. Point the converter at that original file — TS to WAV or MP4 to WAV — and the audio is read and written out as PCM normally. A container that has merely been renamed to .cavs is detected by its contents rather than its extension, so those convert without any special handling.
Because there is no audio stream to write. In our testing, feeding a video-only input to the audio pipeline stops at the encoder with "Output file does not contain any stream", and no output file is created. An error is the more useful outcome — a silent WAV would look like a successful conversion and only reveal the problem later, after you had built something around it.
Into whatever container the video was demuxed from, where it still is. AVS material is multiplexed: coded video and a separate coded audio track live side by side inside a transport stream or MP4. Demuxing to a bare .cavs keeps only the video elementary stream. Convert the original file instead (TS to WAV or MP4 to WAV) and the audio comes with it.
PCM 16-bit little-endian — the same linear PCM as an audio CD and the format every editor expects. The Audio Codec control also offers 24-bit and 32-bit little-endian PCM for deeper masters, and A-law and mu-law, which are 8-bit logarithmic telephony encodings that halve the file size at the cost of dynamic range.
Not in the PCM modes. Linear PCM stores samples verbatim, so if the source track exists and you leave Audio Sample Rate and Audio Channel on Original, nothing is resampled, re-quantised or re-encoded. Two things do change the samples: choosing a different sample rate, which is a genuine resample, and choosing A-law or mu-law, which are lossy 8-bit encodings despite living in a WAV file.
Because it is uncompressed. A compressed source might carry audio at 128 kbps; the same audio as 44.1 kHz 16-bit stereo PCM is 1,411 kbps — roughly eleven times the data for the same sound. That is the trade WAV exists to make. If you want a smaller file after editing, compress the result with WAV to MP3 or export straight to a modern codec instead.
The format itself tops out near 4 GiB, because RIFF stores chunk sizes in 32-bit fields — that works out to a little under seven hours of 44.1 kHz 16-bit stereo. In practice you will run into how large a file you can comfortably upload long before you approach the format ceiling.
When the WAV is an intermediate rather than a deliverable. Extra bit depth buys headroom for further processing — gain staging, normalisation, noise reduction — where repeated rounding at 16 bits would accumulate. It cannot recover detail that was never captured, so extracting an 8-bit or heavily compressed broadcast track at 32 bits gives you a larger file containing exactly the same audio.
.cavs and a .avs file?They are unrelated despite the similar names. A .cavs file is coded Chinese AVS video. A .avs file is almost always an AviSynth script — plain text that tells a frameserver how to process other clips, holding no media of its own. Open the file in a text editor: readable instructions mean AviSynth, and binary data means a genuine AVS video stream.
Your file is uploaded over an encrypted connection, processed on our servers, and deleted automatically a few hours after the conversion finishes — never shared and never made public. There is no sign-up and no watermark.