Initializing... drag & drop files here
Supports: CR2
A .cr2 is a Canon raw still — ExifTool's format list describes it plainly as "Canon RAW 2 (TIFF-based)" — while WebM is a video container the WebM Project built on Matroska and published under a BSD-style licence in 2010. Going from one to the other is not a re-wrap. The raw is developed into an ordinary RGB frame, that single frame is held on screen for a duration you choose, and the result is encoded as a short, silent WebM clip. Nothing in the output moves, because there was never any motion in the source.
| Property | Value on this page |
|---|---|
| Container | WebM, a subset of Matroska |
| Video codec | VP9, preselected; VP8 and AV1 are the other two choices |
| Audio stream | None — a photograph carries no sound, so none is written |
| Frame rate | 1 fps; there is no motion to sample |
| Clip length | The Image Duration you set, from 1/60 s up to 10 seconds, starting at 5 seconds |
| Frame count | Duration multiplied by 1 fps, so a 5-second clip is 5 identical frames |
| B-frames | Switched off — they are useless between identical frames and delay the start time |
| Frame preparation | The raw is rendered to a clean PNG at no more than 4096 pixels on its long edge before encoding |
| Letterbox padding | Filled with the Background Color, white unless you change it |
| Property | VP9 | VP8 | AV1 |
|---|---|---|---|
| Preselected here | Yes | No | No |
| Opened up by Google | 2013 | 2010, later documented by the IETF as RFC 6386 | Developed instead by the Alliance for Open Media, 2018 |
| Named in the WebM Project's own description | Yes | Yes | No — it postdates the original pairing |
| Compression at equal quality | Better than VP8 | The weakest of the three | The strongest of the three |
| Encoding time on a single still | Short | Shortest | Noticeably longer |
| Browser reach | Broad across current desktop and mobile browsers | Broadest, it is the oldest | 94.28% of tracked browsers per caniuse, Safari only partial from 17.0 |
| Constant Quality slider | CRF 16–63 | CRF 16–63 | CRF 16–63 |
| Pick it when | You have no specific requirement | An old WebM decoder has to read the file | File size matters more than encode time |
.cr2 onto the page or click "+ Add Files". Queue several and Merge strategy decides between one combined clip ("Merge images") and a separate file per photograph ("Video per image")..webm. The file is uploaded over an encrypted connection, converted on our servers and deleted automatically a few hours later — no sign-up, no watermark.Because there is nothing to encode. A still photograph has no audio, so the converter writes no audio stream at all rather than padding the file with a blank one — that is why the Audio Codec control does not appear for this conversion. A video container with only a video track in it is entirely normal and every player handles it. If you need music or narration over the still, add it in an editor after downloading.
Because a photograph has no motion to sample. The image is looped, held for the duration you chose, and stopped there, so a five-second clip is five identical frames rather than 150. Encoding repeated frames costs almost nothing, and any editor will re-time the clip to 24, 25 or 30 fps if something downstream expects a conventional rate.
VP9 is preselected because it is the sensible middle ground: better compression than VP8, far faster to encode than AV1, and decoded by current browsers without fuss. Switch to VP8 only if something old has to read the file — it is the codec WebM launched with. Switch to AV1 if size is the priority and you do not mind waiting; note that caniuse puts AV1 at 94.28% of tracked browsers, with Safari offering only partial support from 17.0 onward.
The photograph's own dimensions unless you change them, with one caveat: the frame is prepared at no more than 4096 pixels on its long edge before it reaches the encoder, so a high-resolution modern raw is scaled down to fit. For anything web-facing you want far less than that anyway — pick a Fixed Resolution such as 1920 x 1080, or set Width x Height in pixels.
Because the frame you asked for is a different shape from the photograph. Rather than stretching or cropping the image, the converter centres it and pads the leftover space with the Background Color, which is white by default. Either choose a frame size with the same orientation as the photo, or change the padding colour to black or something that suits the project.
No. A CR2 holds unprocessed sensor data, which is why exposure, white balance and highlight recovery are still adjustable long after the shot. Making a video frame means demosaicing that data into ordinary 8-bit RGB and baking the current interpretation in permanently. Do your raw edits first, keep the .cr2 as the master, and treat the WebM purely as a delivery file.
Yes. Add them all before converting and leave Merge strategy on "Merge images" — each rendered photo is held for the duration you set and they play back to back in upload order. There are no transitions between them; each is a hard cut to the next still. Choose "Video per image" instead and you get one .webm per photograph, all sharing the same settings.
If the goal is to view, share, print or archive the shot, a video container is the wrong wrapper — use CR2 to JPG for something that opens everywhere, or CR2 to WebP for a much smaller web image. WebM makes sense when you genuinely need a video file: a slate or title card on a timeline, or a still that has to sit alongside other WebM assets. If your camera is a newer mirrorless body writing .cr3 files, use CR3 to WebM instead.