Initializing... drag & drop files here
Supports: MOS
This is an unusual pairing and it is worth stating plainly before you start. MOS is a raw photograph from a Leaf medium-format digital back — one very high-resolution still, undeveloped sensor data in a TIFF-derived container. HEVC is a video codec. Converting between them means developing the raw into an image and then encoding that single frame as a short, silent, motionless clip. There is a legitimate reason to do it — an H.265 ingest pipeline that will only accept a .hevc file — but if you simply want to see or share the photograph, MOS to JPG is the page you want, and if you want the still as a clip that plays widely, MOS to MP4 is a far more forgiving target.
| Property | Value |
|---|---|
| Full name | Leaf camera raw |
| Origin | Leaf — formerly a division of Scitex, then Kodak, now a subsidiary of Phase One |
| Container | TIFF-derived |
| Payload | Undeveloped colour-filter-array sensor mosaic plus development metadata |
| Typical resolution | 22 MP (Aptus-II 5) to 80 MP (Aptus-II 12) |
| Motion | None — a single still frame |
| Audio | None |
| Alpha channel | None |
| Decoded by | Capture One, Leaf Capture, and LibRaw-based tools |
| Succeeded by | IIQ, used by the later Leaf Credo backs |
| Property | Value |
|---|---|
| Codec standard | ITU-T H.265 / ISO/IEC 23008-2, "High efficiency video coding" |
| First approved | April 2013; the edition currently in force is dated January 2026 |
| Container actually written | Matroska — despite the .hevc file name |
| Video codec | H.265, fixed for this target |
| Audio track | None; the audio controls are hidden entirely for image sources |
| Frame rate | 1 frame per second |
| Clip length | Set by Image Duration; 5 seconds by default |
| Background fill | White by default |
| Rate control available | Quality Preset (default), Constant Quality, Constraint Quality |
| Opens reliably in | VLC, mpv and other ffmpeg-based players |
.mos onto the page or click "+ Add Files". A medium-format raw is large, so the upload is the part you will wait for.Three behaviours surprise people on this page, so they are worth spelling out. First, every uploaded image is normalised before encoding and capped at 4096 pixels on its long edge — so an 80-megapixel Leaf frame, which is roughly 10000 pixels wide, arrives at the encoder already downscaled. "Keep original" means the frame's own aspect ratio at that capped size, not the back's native pixel count.
Second, the encode runs at 1 frame per second with B-frames disabled. There is no motion to predict, so the encoder is writing what is effectively a single coded picture repeated for the duration you chose.
Third, the rate-control choices are deliberately narrower here than on a video-source page. Every bitrate-based and target-file-size mode is hidden for image sources; you get Quality Preset, Constant Quality, and Constraint Quality, and nothing else. Quality Preset is the right default because H.265 is a CRF-based codec and responds to it properly. If you need a hard ceiling on bitrate for an ingest specification, Constraint Quality is the mode that gives you a Max bitrate field.
If you upload several .mos files at once, the Merge strategy control decides what happens: Merge images — the default — strings them into one clip, while Video per image gives you a separate file per frame.
Because the file is a Matroska container that happens to be named .hevc. It is not a raw elementary H.265 stream, and Matroska is not a format QuickTime or most stock mobile galleries recognise. Players built on ffmpeg — VLC and mpv among them — open it directly with working seek. The simplest fix is to rename the file to .mkv, which lets a lot more software identify it correctly. If you need something that plays without any of this, use MOS to MP4.
No. A still photograph carries no audio, and the converter reflects that: for image sources the audio codec controls are hidden from the page entirely rather than being set to silence. The output is video-only. If your pipeline requires an audio track to be present, you will need to mux one in afterwards with a desktop tool.
Because there is exactly one picture to show. Encoding it at 30 or 60 fps would mean emitting dozens of identical frames every second for no benefit; at 1 fps the encoder writes far less data and every player still displays the image for the full duration. If a downstream system insists on a specific frame rate, re-encode the resulting clip in that system's own ingest step.
Not meaningfully. H.265 earns its reputation by predicting how blocks move between frames, and a single-frame clip gives it nothing to predict. You still pay the costs — slower encoding than H.264 and much patchier playback support — without collecting the benefit. For a still-as-clip, H.264 in an MP4 is the more sensible engineering choice; for the photograph itself, a JPG or PNG is.
Two limits apply in sequence. Every image is capped at 4096 pixels on the long edge before encoding, which already takes an 80-megapixel Leaf frame well below its native size, and then whatever you choose under Video resolution applies on top of that. This is not a setting you can raise. If full resolution matters, the video path is the wrong one — export the frame with MOS to PNG or MOS to TIFF, which keep the back's pixel dimensions.
From the Background Color option, which defaults to White for image sources. It only becomes visible when the output frame's aspect ratio differs from your photograph's — medium-format backs are close to square or 4:3, so asking for a 16:9 fixed resolution pads the sides. Either pick a different Background Color, choose a fixed resolution that matches your frame's proportions, or leave Video resolution on Keep original so no padding is needed.
Not on this page. For image sources every file-size and bitrate-based rate-control mode is hidden, leaving Quality Preset, Constant Quality and Constraint Quality. Constraint Quality is the closest thing to a size control — it exposes a Max bitrate field, and since you also control the duration, bitrate multiplied by seconds gives you a workable ceiling on the finished file.
It is uploaded over an encrypted connection, developed and encoded on our servers, and both the uploaded raw and the finished clip are deleted automatically after a few hours. There is no sign-up, no watermark, and nothing is shared or made public. Keep the original .mos yourself — the encode is a delivery artefact, not an archive, and none of the raw's editing latitude survives it.