RAF to MKV Converter

Convert RAF files to MKV format online. Free, fast, no watermarks.

Initializing... drag & drop files here

Supports: RAF

OptionsAdvanced Options - Our defaults are optimized for the best results. We recommend you keeping the defaults unless you have a specific need.
Show All Options
Merge strategy
Select Merge images to combine all uploaded files into a single video. Use Video per image to create a separate video for each individual file.
Image Duration
Duration
How long each still image is held on screen. This applies to static images — including a static WebP, AVIF or HEIC. It is ignored for animated sources, which keep their own frame timing: GIF, APNG, and an animated WebP, AVIF or HEIC.
Background Color
Background Color
File Compression
Preset
Video resolution

Which Codec Should Your RAF Slideshow Use?

Converting Fujifilm raw stills into an MKV gives you something no other target here does: a real, correctly-named Matroska container plus a codec dropdown with twenty-one entries. That choice is the whole decision on this page — the container is the easy part. The short answer is to leave it on H.264, which is what it ships with. The long answer is that several of the listed codecs will quietly resize your photograph, and one of them will fail outright at the default settings. This page explains which, and why.

Codec Comparison for a Full-Resolution RAF Frame

Every image is normalised to a maximum long edge of 4096 pixels before encoding, so a full-resolution 3:2 Fujifilm frame — 7728×5152 on an X-T5 — arrives at the encoder as 4096×2730. Here is what each codec then does with it:

Codec On our supported MKV list What happens to a 4096×2730 still Verdict
H.264 (default) Yes Kept at 4096×2730 The safe choice; plays everywhere
H.265 Yes Kept at 4096×2730 Smaller at equal quality; needs a newer decoder
AV1 Yes Kept at 4096×2730 Smallest files, slowest encode
VP9 Yes Kept at 4096×2730 Solid open alternative to H.265
VP8 Yes Kept at 4096×2730 Older; use Quality Preset, not Constant Quality
Theora Yes Kept at 4096×2730 Legacy open codec, rarely the right answer
MPEG-2 Yes Fails — see the warning below Only with an explicit smaller frame size
MPEG-4 Yes Clamped to 2048 px per axis, including on "Keep original" — and per-axis, so the aspect ratio breaks Legacy players only
Xvid, DivX, MS MPEG-4 No Also reduced toward 2048 px Legacy players only
MPEG-1 No Capped just under 4096 px Historic interest only
H.263 No Snapped to 1408×1152 — aspect ratio not preserved Avoid
H.261, H.263+ No Capped at 352×288 Avoid
WMV 1, WMV 2, FLV, Flash Video, MJPEG, HuffYUV No Varies by codec Untested combinations

Matroska itself is broader than that list: its codec specification defines identifiers for MPEG-1, MPEG-4 ASP and MJPEG among others, plus a catch-all V_MS/VFW/FOURCC mapping for Video-for-Windows codecs carried over from AVI. So a "No" above means the combination is outside what we validate, not that the format forbids it — but the practical risk is identical either way, because the codec you pick is passed straight through without being checked against the container. A fringe choice can mux cleanly and still fail to decode in the player you hand it to.

The one outright failure is worth spelling out: picking MPEG-2 and leaving the resolution on "Keep original" produces a hard error, because the pre-encode normalise step lands the frame on exactly 4096 pixels and the MPEG-2 encoder rejects any dimension that is a multiple of 4096. We reproduced it directly — ffmpeg reports "Width or Height are not allowed to be multiples of 4096" and writes a zero-byte file, while the same encode at 3840×2560 succeeds. If you need MPEG-2, set an explicit Width x Height first.

When to Stay on H.264

  • The file has to play on someone else's machine, phone, TV or in a browser without being asked about first.
  • You want the shortest encode time on a large batch of raws.
  • The clip is going into an editor or a platform that has not committed to newer codecs.
  • You simply want it to work. H.264 in Matroska is about as universally handled as video gets.

When to Switch Codecs

  • Storage is the constraint: pick H.265 or AV1. Held still frames compress extremely well and the newer codecs widen the gap further.
  • You are feeding something open-source or web-oriented: VP9 is well supported by open tooling; if the destination is genuinely the web, RAF to WebM is a better fit than an MKV.
  • You are targeting old hardware: MPEG-4 will play on very old set-top boxes and DVD-era software, at the cost of a silent downscale to 2048 pixels.
  • Someone specifically asked for an elementary-codec filename: RAF to HEVC and RAF to AV1 write the same Matroska bytes under a codec-named extension, which then needs renaming to open. This page avoids that problem entirely.

How to Convert RAF to MKV

  1. Upload Your RAF Files: Drag the .raf files onto the page or click "+ Add Files". Only .raf inputs are accepted, and upload order becomes shot order when files are merged.
  2. Set the Merge Strategy and Image Duration: Merge strategy ships on "Merge images", building one clip from every file; switch to "Video per image" for one MKV per photograph. Image Duration ships at 5 seconds per still and ranges from 1/60 second to 10 seconds.
  3. Pick the Video Codec and Resolution: Video Codec ships on H.264. Under Video resolution, "Keep original" is selected; use Preset Resolutions to set a height with width following the source aspect, or Width x Height for an exact frame — required if you choose MPEG-2.
  4. Convert and Download: Click Convert and save the .mkv. It opens directly, with no renaming. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. No sign-up, no watermark.

What Every MKV From a Still Image Looks Like

Regardless of which codec you choose, these are fixed for image sources:

Setting Value Adjustable
Container Matroska No — that is the point of this page
Frame rate 1 fps No
B-frames Disabled No
Pixel format yuv420p — 8-bit, 4:2:0 No
Audio track None No — a photograph has no audio, so the controls are hidden
Clip length 5 seconds per still by default Yes, via Image Duration
Long-edge cap before encoding 4096 px No
Background and letterbox fill White Yes, via Background Color
Rate control modes offered Quality Preset (selected), Constant Quality, Constraint Quality Yes

Bitrate and target-file-size modes are hidden for image sources, because a still has no source bitrate to scale from. Size is governed by resolution, clip length and the quality mode you pick.

Frequently Asked Questions

Why did my MPEG-2 conversion fail when everything else worked?

Because of an unlucky collision between two limits. Images are normalised to a 4096-pixel long edge before encoding, and the MPEG-2 encoder refuses frame dimensions that are exact multiples of 4096 — it reports "Width or Height are not allowed to be multiples of 4096" and produces nothing. Any full-resolution Fujifilm frame wider than 4096 lands on exactly that number. Set an explicit Width x Height (3840×2560 works) or a Preset Resolution before selecting MPEG-2.

Why is my MPEG-4 or Xvid output smaller than I asked for?

Those codecs are capped at 2048 pixels on each axis in this pipeline, and the resize is applied silently rather than refusing the job. If you set 1920×1280 you will get it; if you leave "Keep original" you will get roughly 2048 pixels wide instead of 4096. Choose H.264 or newer if resolution matters.

Why is there no audio in my MKV?

There is no audio in a photograph, so there is nothing to encode and the audio controls are hidden rather than offering an option that would do nothing. If you want a soundtrack over the slideshow, produce the MKV here and add audio in a video editor afterwards.

Why does the video run at 1 frame per second?

Because a held still image is identical in every frame, so storing 25 or 30 copies per second only inflates the file. The clip plays for the full Image Duration regardless; only the number of stored frames differs. There is no frame-rate control on this page for that reason.

Can I put a whole shoot into one MKV?

Yes, and that is the default behaviour. Merge strategy ships on "Merge images", so every raw you upload becomes one shot in a single MKV, in upload order, each held for the Image Duration you set. Twelve photographs at 5 seconds each gives a one-minute film. Pick "Video per image" if you would rather have one file per photograph.

Does the MKV keep the RAF's 14-bit colour and full detail?

No. The raw decoder develops Fujifilm raw at 8 bits per channel by default, and the encoder writes yuv420p, which is 8-bit with the colour channels at half resolution. Combined with the 4096-pixel long-edge cap, an MKV from this page is a delivery file, not an archival one. Keep the .raf masters, and use RAF to TIFF if you want a still render instead.

Is every codec in the dropdown actually safe to use?

No, and it is worth being explicit about that. Your codec choice is passed straight to the encoder rather than being checked against the output container, so a fringe pick such as HuffYUV, Flash Video or WMV can produce a file that muxes without complaint and then fails to decode in the player you hand it to. The rows marked "Yes" in the table above are the combinations we treat as supported for Matroska output; everything else is a gamble that is rarely worth taking.

How are my raw files handled during conversion?

Files 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. If the finished MKV is larger than you want, compress MKV will shrink it, and the RAF converter hub lists every other target Fujifilm raw can reach here.

Rate RAF to MKV Converter Tool

Rating: 4.8 / 5 - 54 reviews