CAVS to MPEG-2 Converter

CAVS to MPEG-2: Getting a Raw AVS Stream into a Format Things Can Read

A .cavs file is a bare Chinese AVS video stream — AVS1-P2 in its Jizhun (Main) profile, the video half of China's national audio-video coding standard GB/T 20090.2-2006. It is not a container. There is no audio track inside it, no index, and no metadata block telling a player how long it runs. That is why so little software opens one, and why converting it to MPEG-2 is a sensible rescue: MPEG-2 is the codec DVD authoring tools, DVB and ATSC ingest equipment, set-top boxes and older signage players were all built around, wrapped this time in a real program stream with timing information.

How to Convert CAVS to MPEG-2

1. Upload Your CAVS File: Drag and drop the .cavs onto the page or click "+ Add Files". Only .cavs is accepted here; several can be queued and they share the same settings.
2. Choose a File Compression Mode: "Quality Preset" is preselected on "Very High (Recommended)". For a predictable result on this route, switch to "Specific file size", "Constant Bitrate" or "Variable Bitrate" and state a number — those three are honoured directly, while "Constant Quality" and "Constraint Quality" are discarded on an MPEG-2 target and replaced with the minimum bitrate for the frame size.
3. Set the Video Resolution: "Keep original" keeps the frame size the AVS stream carries. Use "Width x Height" for exact broadcast dimensions such as 720 x 480 or 720 x 576, "Preset Resolutions" to pick a target height with the width following the source aspect, or "Resolution Percentage" to scale proportionally.
4. Convert and Download: Click "Convert" and take the .mpeg2 file. Video is written as MPEG-2 and the audio codec is fixed at MP2 — the only entries the panel offers on this target. No sign-up, no watermark.

What Is Actually Inside a .cavs File

That last row is the reason this conversion only runs one way. AVS1-P2 is not offered as an output format anywhere on this site, and encoders for it are scarce outside Chinese broadcast tooling, so a .cavs file is best treated as something to migrate off rather than a format to produce. Once it is MPEG-2 you can go anywhere from there — CAVS to MP4 is the better first stop if playback rather than broadcast compatibility is your goal.

Choosing the Output Frame Size

MPEG-2 will not encode below a floor derived from the frame size — roughly 0.1 bits per pixel at standard definition rising to 0.18 at 4K, and never below 2 Mbps — so the frame size you pick sets the floor of what the file can weigh:

For disc authoring, enter the exact pixel dimensions in "Width x Height" rather than using a height preset: the presets set the height and let the width follow the source aspect, which will not land on 720 pixels for you. And remember the converter produces the video stream, not a burnable disc — the VIDEO_TS structure is the authoring tool's job.

Common Problems with Raw AVS Streams

• The clip plays too fast or too slow. A raw stream has no container timing, so the reader falls back to what the AVS sequence header declares — and if that is missing or the stream was cut mid-sequence, it starts from a 25 fps assumption. A 29.97 fps recording read as 25 fps runs long and out of sync with any audio you add later.
• The output has no sound. Expected: AVS elementary streams carry video only. There is nothing for the MP2 audio encoder to work on, so the MPEG-2 file is silent. Add the soundtrack in an editor afterwards.
• The file will not open at all. Check that it really is an AVS elementary stream and not an AVS video track that was pulled out of a transport stream with the wrong tool, or a .cavs-named file that actually holds something else. A truncated download fails the same way.
• Interlaced content still combs. Much AVS broadcast material is interlaced, and nothing in this pipeline deinterlaces, so comb lines on moving objects are re-encoded into the MPEG-2 output. Deinterlace beforehand if the result must be progressive.
• The output is bigger than the input. AVS1-P2 was designed to reach roughly H.264-class efficiency, and MPEG-2 is a generation behind it. Re-encoding to MPEG-2 at a watchable quality genuinely costs more bits — that is the price of the compatibility you are buying.

Frequently Asked Questions

What is CAVS, and who used it?

CAVS is the common name for AVS1-P2, the video part of the Audio Video Coding Standard developed by China's AVS Workgroup and issued as national standard GB/T 20090.2-2006. It was built as a lower-complexity alternative to H.264 for Chinese digital television and set-top boxes, and single-chip AVS decoders shipped in volume from the late 2000s onward. Outside that ecosystem it is rare, which is exactly why files turn up that nothing on a Western desktop will open.

Why does my converted file have no audio?

Because the source had none. A .cavs file is a raw video elementary stream — the format has no provision for an audio track, and AVS audio is a separate specification carried in a separate stream. The MPEG-2 output therefore comes back silent even though MP2 is listed as the audio codec. If the original programme had sound, it lives in whatever transport stream the .cavs was extracted from; go back to that file and convert it instead.

Can I convert MPEG-2 back to CAVS?

No — AVS1-P2 is not available as an output format here, and encoders for it are rare outside Chinese broadcast tooling, so there is no practical path back into the format. Treat a .cavs file as an archive to migrate out of. If you need to keep the picture as close to the original as possible while you move, use a generous bitrate on the first conversion — every generation after that compounds.

Which compression mode should I pick for a CAVS source?

"Specific file size", "Constant Bitrate" or "Variable Bitrate". MPEG-2 has no constant-rate-factor mode, so the two quality-style options — "Constant Quality" and "Constraint Quality" — are converted into a plain bitrate at the minimum for your frame size, and whatever number you typed is dropped. The preselected "Quality Preset" works, but it scales the source's bitrate, which for a raw stream with no declared bitrate is a less reliable starting point than simply naming the bitrate or size you want.

Will the picture look worse after converting to MPEG-2?

Slightly, yes. This is a full decode and re-encode, and MPEG-2 needs more bits than AVS1-P2 for the same visual quality because it is an older design with a simpler transform and prediction toolset. Give it headroom: at standard definition 4-6 Mbps is comfortable, and dropping to the 2 Mbps floor will show blocking on detailed or fast-moving footage.

Is this the right way to get CAVS footage onto a DVD?

It is the first half. Set the resolution to 720 x 480 for NTSC or 720 x 576 for PAL using "Width x Height", keep the codec on MPEG-2 and pick a bitrate that fits your disc, and the output is a stream a DVD authoring application will accept. The authoring application still has to build the VIDEO_TS folder structure with its IFO and VOB files and burn the image — no converter produces a playable disc on its own.

What frame rate does the conversion use?

Whatever the stream declares. AVS sequence headers carry a frame-rate code, so a complete, well-formed .cavs normally converts at its true rate. Where a raw stream is truncated, spliced or missing its opening headers, the reader falls back to a 25 fps assumption, and the result runs at the wrong speed. If the converted clip's duration does not match what you expected, that mismatch is the reason.

Should I convert to MPEG-2 or straight to MP4?

Depends entirely on the destination. Choose MPEG-2 when something specifically requires it — DVD authoring, a DVB or ATSC ingest chain, older signage or set-top hardware. Choose CAVS to MP4 when a human is going to watch the file: H.264 in MP4 is smaller at the same quality and plays on phones, browsers and televisions without ceremony. CAVS to MKV is the third option if you want a flexible container for archiving.

How are my files handled, and how long do you keep them?

Your file is uploaded over an encrypted connection, decoded and re-encoded on our servers, and deleted automatically a few hours after conversion — no sign-up, no watermark, never shared or made public. Raw elementary streams are large for their duration because they carry no container overhead but also no compression advantage, so upload time is usually the longest part of the job.