CAVS to WebP Converter

Extract a WebP Still from CAVS: What This Covers

A .cavs file holds raw Chinese AVS (AVS1) coded video with no container around it, so no image editor and no browser will open it directly. This tutorial covers pulling a still out of that stream and saving it as WebP — the modern web image format — including the one decision that actually changes the result, the Lossless? switch, and what to do when the extraction lands on the wrong frame.

How to Convert CAVS to WebP

1. Upload Your CAVS File: Drag and drop your .cavs onto the page, or click "Add Files" to browse for it. Several streams can be queued and run with the same settings.
2. Choose the Frame: Specific Frame is preselected. Leave Time (seconds) blank to take the first frame of the stream, or type a moment such as 12.250 to seek to twelve and a quarter seconds. Multiple Screenshots with a Capture Rate samples a run of stills instead.
3. Set Lossless and Quality Preset: Lossless? starts on No (Recommended) for a small lossy WebP; switch it to Yes for a pixel-exact copy. Quality Preset starts on "Very High (Recommended)", and Image resolution scales the still if you want it smaller.
4. Convert and Download: Click "Convert" and save the .webp. No sign-up, no watermark.

Walk-through: Lossy or Lossless?

WebP is really two compression engines behind one extension, and the Lossless? switch picks which one runs. Lossy mode uses predictive block coding related to the VP8 video codec; lossless mode uses a completely different entropy-coded scheme. Google measures lossy WebP at 25-34% smaller than comparable JPEG at equivalent SSIM quality, and lossless WebP at 26% smaller than the equivalent PNG.

When Lossless? is No, the Quality Preset decides how aggressive the lossy encode is. Each preset resolves to a fixed quality value:

Rules of thumb worth applying:

• Posting the still on a web page: leave Lossless? on No and Quality Preset at High or Medium. Nobody will see the difference and the file is a fraction of the size.
• Text, captions or graphics burned into the frame: set Lossless? to Yes. Lossy encoders ring around hard edges, which is exactly where broadcast overlays live.
• Keeping a master you will re-edit: set Lossless? to Yes, or use CAVS to PNG if the receiving software is old.
• Hitting a size budget: leave Lossless? on No and lower the preset one step at a time rather than jumping straight to Lowest.

Common Errors and How to Fix Them

• "The still is black or blank" — The timestamp landed past the end of the clip, or on a fade or a cut. Lower the Time (seconds) value so it sits inside the footage, or run Multiple Screenshots first to see which moments have content.
• "I got one picture, I wanted the whole clip" — This is frame extraction, not video conversion. For moving footage, use CAVS to MP4, or CAVS to GIF for a short looping animation.
• "My WebP will not open in an older program" — WebP is comparatively recent. Save the frame as CAVS to JPG or CAVS to PNG when the destination software predates it.
• "The frame looks soft or blocky" — That softness is in the source. AVS1 applies an in-loop deblocking filter, so heavily compressed broadcast material decodes soft no matter what the encoder does afterwards.
• "Uploading takes a long time" — Broadcast captures get large. The practical constraint here is upload size and time rather than the frame extraction itself, so trim a long recording before uploading if you only need one moment from it.

When This Doesn't Work

Frame extraction needs a decodable AVS1 bitstream. A truncated download, a file cut off mid-transfer, or a stream damaged during demuxing can leave the decoder unable to reach your chosen timestamp — the conversion then fails or returns an unexpected frame. Re-export or re-download the source and try a timestamp earlier in the clip. If the file opens as readable text in an editor it is an AviSynth .avs script rather than Chinese AVS video, and it contains no picture data to extract at all. And if what you actually need is every frame of the clip rather than a few stills, that is a video job: convert to a playable format first and pull frames from the result.

WebP Support and the Alternatives

Frequently Asked Questions

Does this produce an animated or looping WebP?

No. This is still-frame extraction: one WebP file per captured frame, never a single animated WebP. The format is capable of animation, but that is not what this conversion does. For a moving result, use CAVS to GIF or CAVS to MP4.

Should I pick WebP over PNG or JPG for an extracted frame?

WebP if the still is going on the web, because it wins on size in both modes — Google measures lossy WebP at 25-34% smaller than comparable JPEG and lossless WebP at 26% smaller than PNG. In our testing, one 720×576 extracted frame came out roughly a third smaller as lossless WebP than as PNG, consistent with that published average. Choose JPG or PNG when the file has to open in software that predates WebP.

Will transparency in the frame be preserved?

There is none to preserve. WebP does support an alpha channel in both lossy and lossless modes, but a decoded AVS video frame is fully opaque — video codecs of this generation carry no transparency — so the output is a solid image. Alpha only becomes relevant if you add it yourself in an editor afterwards.

Which frame do I get if I leave the time blank?

The first frame of the stream. Specific Frame is preselected and Time (seconds) starts empty, so a straight upload-and-convert gives you frame zero. The field accepts decimals, so 2.100 targets 2 seconds and 100 milliseconds in — useful for stepping around a moment of motion in hundredths until you land on a clean frame.

What does Multiple Screenshots actually produce?

A set of independent still WebP files, one per capture, taken across the clip at the Capture Rate you choose. The rate list runs from one frame every 0.1 seconds up to one frame every 10 seconds, with one frame per second preselected. They are not combined into a sheet and not animated — the point is to review several moments and keep the best.

What resolution will the WebP be?

The same as the video by default, because the frame is read from the decoded stream rather than from a screen. A 720×576 standard-definition capture gives a 720×576 image; a 1080p source gives 1920×1080. Image resolution scales it down by percentage, to a preset, or to explicit dimensions — scaling up only interpolates and adds nothing real.

What is CAVS, and why can nothing on my computer open it?

CAVS is first-generation Chinese AVS video — national standard GB/T 20090.2-2006, promulgated in February 2006 by the AVS Workgroup, which was founded in June 2002, and later issued by the IEEE as IEEE 1857-2013. A .cavs file is that coded video as a bare elementary stream with no container, no index and no metadata, so ordinary players and image tools have nothing to work with. Extracting a frame to a common image format is usually the quickest way to see what is in one.

Does the file have audio I am losing by extracting a still?

No. A raw .cavs is a video elementary stream and carries no audio track at all, so an image export discards nothing that was in the file. If you were expecting sound with this footage, it is still inside the container the video was demuxed from — usually a .ts or an .mp4.

What happens to my file after I convert it?

Your .cavs 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.