Initializing... drag & drop files here
Supports: RAF
This is a walk-through for Fujifilm shooters who need a TIFF out of a .RAF capture — for a print lab, a page layout, a retoucher, or a client who cannot open raw files. It covers the one setting that decides whether your TIFF is lossless, the two entries in that dropdown that silently produce a broken file, and the ceiling this path has that a raw developer does not.
If you have not decided that TIFF is the right target yet, the RAF and TIF comparison covers that question instead; this page assumes you already know you want one.
.RAF files onto the page or click "+ Add Files" to browse. Queue as many as you like — they all convert with the same settings, so sort out the compression choice before you start a big batch.TIFF is a container, not a single encoding. The same .tiff extension can wrap losslessly packed pixels, JPEG-compressed pixels, or no compression at all, and the "Compression Type" dropdown is where you choose. It opens on JPEG, which is lossy — that default is the single most common reason someone ends up with a TIFF that has already thrown pixels away. Match the setting to the job instead:
| What you are doing | Set Compression Type to | Why |
|---|---|---|
| Sending a file to a print lab or RIP | LZW | Lossless, and universally readable — this is the format print workflows have expected for thirty years |
| Handing a file to a retoucher | DEFLATE or LZW | Lossless, so the image survives repeated edit-and-save cycles with no artefact build-up |
| Placing the image in InDesign or Affinity Publisher | LZW | Lossless and placed natively by every layout app |
| Feeding an old scanner-era tool that rejects everything | NONE | Uncompressed. Enormous, but nothing can fail to read it |
| Squeezing a proof down for email | JPEG (the default) | Lossy, but small. Fine for a look-at-this copy, wrong for a master |
| Anything at all | Not CCITT Fax 4, not JP2K | Both produce a broken file — see the errors section below |
Two more things about that dropdown are worth knowing. "LOSSY" and "JPEG" are the same setting — both resolve to the same JPEG encoder mode, so they produce identical files; it is a duplicate label, not a second option. And the "Quality Preset" only bites on the lossy modes: the quality value reaches the TIFF writer, but the writer applies it to JPEG and WebP compression only. Under LZW, Deflate, ZSTD, PackBits or None it is ignored entirely, so do not expect moving it to change a lossless file.
| Camera | Raw pixel dimensions | Raw bit depth | Uncompressed 8-bit TIFF | Print width at 300 DPI |
|---|---|---|---|---|
| X-T5 (40.2 MP, X-Trans CMOS 5 HR) | 7728 x 5152 | 14-bit | about 119 MB | 25.8 inches |
| GFX100 II (102 MP, primary colour filter) | 11648 x 8736 | 14 or 16-bit | about 305 MB | 38.8 inches |
Those TIFF figures are simple arithmetic — three bytes per pixel for 8-bit RGB — and they are the number LZW or Deflate then compress down from. Photographic detail does not pack well losslessly, so a lossless TIFF of a full-resolution Fuji frame commonly lands in the same range as, or above, the raw file it came from.
This path develops the raw file once, with a neutral interpretation, and writes 8 bits per channel — there is no bit-depth control on the page. That is the right output for print, layout and handoff, and the wrong output for two jobs. If you still need to move exposure, white balance or highlight recovery, do it in a raw developer (Lightroom, Capture One, RawTherapee, or Fujifilm's own software) before anything is rendered, because that latitude only exists while the file is still raw. And if you are building a genuine preservation master, note that your camera recorded 14 bits per photosite (16 on some GFX bodies) and this TIFF carries 8 — export a 16-bit TIFF from a raw developer, or simply keep the original .RAF, which is the most faithful copy that will ever exist. For a delivery copy headed to the web rather than to print, RAF to JPG is a far more sensible target than a TIFF.
Because "TIFF is lossless" was never quite true — TIFF is a container that permits several encodings, including JPEG, and the Compression Type control on this page opens on JPEG. Nothing is wrong with your file; you got exactly what the default asked for. Set Compression Type to LZW, DEFLATE or NONE and re-convert, and the pixels that come back will be exactly the pixels that were written.
Both are genuinely lossless, so this is a compatibility question rather than a quality one. LZW is the older scheme and the one every TIFF reader ever written supports, which makes it the safe pick when the file is going somewhere you do not control — a print lab, a client, an old RIP. Deflate (the same algorithm as ZIP) usually produces a slightly smaller file and is fine in any modern application. If a tool refuses to open a Deflate TIFF, re-export it as LZW and it will.
No. There is no bit-depth control on this page and the output is 8 bits per channel, whatever the source held. That matters because Fujifilm X-series bodies record 14-bit raw and some GFX bodies record 16-bit, so a real amount of tonal headroom is discarded during the render. For screen, print and handoff it is not a practical problem — 8-bit RGB is what almost every output device consumes. For heavy post-processing or an archival master, use a raw developer that can write 16-bit TIFF, and keep the .RAF either way.
Do not count on it. Nothing in this conversion asks the encoder to strip metadata, but how much of a camera's raw metadata survives being demosaiced and re-written as a TIFF depends on the decoder, and Fujifilm-specific fields — the Film Simulation in use, dynamic-range settings, lens corrections — are stored in maker-note structures that a generic renderer has no obligation to carry across. If shot data has to travel with the image, keep the original .RAF alongside the TIFF, or record the settings separately.
Yes — queue as many as you like and they are processed together. The thing to plan for is not the count but the volume: at 40-80 MB per raw file, a hundred-shot batch is several gigabytes to upload, and each of those becomes a TIFF that is the same size or larger. If you are converting in bulk, decide the Compression Type before you start, because every file in the batch takes the same setting and re-running a large batch is the expensive mistake here.
Both. Most Fujifilm X-series bodies use the X-Trans colour filter array — a 6 x 6 non-Bayer pattern designed to suppress moire without an optical low-pass filter — while the GFX medium-format line uses a conventional array (the GFX100 II spec sheet lists a "102MP CMOS II HS with primary color filter"). Both save as .raf and both convert here. Worth knowing: demosaicing X-Trans is a harder problem than demosaicing a Bayer grid, and different renderers reach visibly different answers on fine detail, so an exact pixel match with your desktop developer is not something to expect.
Send the full grid. This page has no DPI control, and DPI is only a tag telling the printer how densely to lay pixels down — it never creates or removes any. Leave "Image resolution" on "Keep original" and the lab gets everything the sensor captured: at the standard 300 DPI, a 7728 px wide X-T5 frame prints 25.8 inches across without interpolation. Scaling down before you send is only worth doing if the lab has given you a maximum, or if the upload size is a genuine problem.
Your file travels over an encrypted connection, is developed and written to TIFF on our servers, and is deleted automatically after a few hours. It is never shared or made public, there is no sign-up, and nothing is watermarked. The practical ceiling on what you can convert is upload size and time rather than anything else, which is worth remembering with raw files that routinely run 40-80 MB each.