Initializing... drag & drop files here
Supports: MOS
This is an image-to-video conversion: your Leaf MOS medium-format RAW captures are developed into viewable frames and encoded as an AV1 video, one clip per capture, concatenated in upload order. It is the route people take for a product turntable, a studio timelapse, or a stills reel that has to be small enough to sit on a web page. Two facts are worth having before you start — the output is silent, and it is a Matroska file carrying an .av1 extension, so most players want it renamed to .mkv.
| Property | Value |
|---|---|
| Full name | Creo Leaf Mosaic RAW |
| Structure | TIFF-based container holding undemosaiced sensor data |
| Written by | Leaf medium-format digital backs |
| Sibling format | Phase One IIQ ("Intelligent Image Quality RAW", also TIFF-based) |
| Metadata carried | EXIF, IPTC, XMP, ICC profile, plus Leaf maker notes |
| Colour | High bit depth per channel, no alpha channel |
| Typical pixel count | Medium-format — far larger than any video resolution you would deliver |
| What conversion does to it | Demosaic plus a standard tone curve to produce a viewable RGB frame |
| Property | Value |
|---|---|
| Codec | AV1, described by the Alliance for Open Media as "an open video codec"; the published bitstream spec is v1.0.0-errata1 |
| Encoder used here | SVT-AV1 |
| Container written | Matroska, saved with an .av1 extension — rename to .mkv for player compatibility |
| Audio | None. Still-image sources carry no audio, so the Audio Codec group is not shown and the result is silent |
| Quality control | Quality Preset (default), Constant Quality, or Constraint Quality. CRF runs 0–63, lower being better quality |
| Rate control | No bitrate or target-file-size modes exist on image-to-video conversions |
| Clip length | Number of captures × the Image Duration value |
| Background fill | The Background Color option, which defaults to White for image sources |
| Encoder frame ceiling | 8192 × 8192 |
| Transparency | Not supported — video formats cannot store an alpha channel |
1/60s (single frame at 60fps) up to 10 seconds per frame. Merge strategy defaults to Merge images (one video from the whole set); switch it to Video per image to get a separate clip per capture..av1 to .mkv if your player will not open it. Files upload over an encrypted connection, are processed on our servers, and are deleted automatically after a few hours. No sign-up, no watermark.No — it is a Matroska file with an unfamiliar extension. AV1 is a codec, not a container, so the video has to live inside something; our pipeline writes Matroska because it is the most flexible container for AV1. Most players decide how to open a file by extension, so renaming sequence.av1 to sequence.mkv makes VLC, MPV, and any other Matroska-aware player treat it correctly. Nothing inside the file changes.
Because the source is a set of photographs. Still images carry no audio, so there is no audio stream to encode and the Audio Codec group does not appear on this page at all. If your reel needs music or narration, add it in a video editor after conversion, or start from a source that already has an audio track.
Multiply the number of captures by the Image Duration value. Twelve MOS files at the default 5 seconds per frame gives a 60-second clip; the same twelve at 1/30s (single frame at 30fps) gives well under half a second, which is the setting you want when the captures are consecutive frames of motion rather than separate shots.
The AV1 CRF scale on this page runs from 16 to 63, and lower means better quality — that is a wider range than the 16–51 scale people remember from H.264, so numbers do not transfer directly. Mid-20s to low-30s is the usual visually-clean territory for AV1; go lower only if you are archiving, since the file grows quickly. If you would rather not think about it, leave Quality Preset on Very High (Recommended) and let it pick.
That is the Background Color option doing its job. Video frames are a fixed rectangle, so when a capture's aspect ratio does not match the output size — very common with medium-format sensors, which are not 16:9 — the difference has to be filled. The default fill for image sources is White. Pick any other colour from the dropdown if white is wrong for your presentation, or set Width x Height to a ratio that matches your captures so there is nothing to fill.
No. The MOS is demosaiced and given a standard tone curve so it becomes a viewable image — that is genuine RAW development, and it is what makes the video possible at all. What it is not is your development: white balance, curves, lens corrections and retouching held in your editor's sidecars are not read. For a graded result, export finished frames from your editor and build the video from those.
AV1 is one of the most computationally expensive codecs in wide use, and a medium-format sequence gives it an enormous number of pixels to chew through. The single most effective fix is to stop encoding at capture resolution: set Preset Resolutions (which sets a height and lets width follow the source aspect ratio) or Width x Height to the size you will actually publish at. Trimming the sequence to the frames you need helps too.
Choose AV1 when file size at a given quality is the priority and your playback target is a modern browser or a recent phone. Choose MOS to MP4 when the video has to drop into a presentation, a client email, an older TV, or an editing timeline — H.264 in an MP4 is universally accepted, decodes in hardware nearly everywhere, and encodes far faster. An AV1 file also needs the rename step; an MP4 does not.
Not as transparency. No video format stores an alpha channel, so anything transparent is flattened against the Background Color before encoding. RAW captures have no alpha to begin with, but if you pre-masked your frames in an editor and exported PNGs, expect the cut-out area to become solid colour here. Keep a cut-out sequence as APNG instead, which does carry alpha.
They are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours — no sign-up, no watermark, never shared or made public. Each RAW is rendered to a clean intermediate image before encoding, and the clips are concatenated strictly in the order you uploaded them, so re-running the same set with the same settings gives the same video.