Initializing... drag & drop files here
Supports: RAF
RAF is Fujifilm's proprietary RAW image format used across the X-Series (X-T5, X-H2, X-Pro3, X100V, X100VI, etc.) and GFX medium-format lines. The header begins with the literal bytes FUJIFILMCCD-RAW, followed by camera identifier, embedded JPEG preview, metadata, and the raw sensor pixels — typically captured from Fujifilm's X-Trans color filter array (a 6x6 non-Bayer pattern) on X-Series bodies and a conventional Bayer array on GFX. Because RAF stores minimally processed sensor data, files are large (often 25-90 MB at 26-102 MP) and only open in dedicated software like Fujifilm X RAW Studio, Lightroom Classic, Capture One, RawTherapee, or DxO PhotoLab. PDF is the universal envelope for sharing — it renders identically in every browser, OS, and email client.
| Property | RAF (Fujifilm) | CR2/CR3 (Canon) | NEF (Nikon) | ARW (Sony) | DNG (Adobe) |
|---|---|---|---|---|---|
| Header / signature | FUJIFILMCCD-RAW |
TIFF-based | TIFF-based | TIFF-based (Sony IFD) | TIFF/EP-based (open spec) |
| Sensor pattern | X-Trans (X-Series) or Bayer (GFX) | Bayer | Bayer | Bayer | Whatever the source camera uses |
| Compression | Uncompressed, Lossless, Lossy (X-T3+, X-Pro2+) | Lossless or C-RAW (CR3) | Lossless or compressed | Compressed by default | Lossless or lossy |
| Embedded JPEG preview | Yes | Yes | Yes | Yes | Yes |
| Spec status | Proprietary | Proprietary | Proprietary | Proprietary | Open (Adobe) |
| Typical file size (24-MP) | 25-55 MB | 25-40 MB | 25-45 MB | 25-50 MB | 25-50 MB |
| Compression Type | Use case | Quality | Output size |
|---|---|---|---|
| Screen (default) | Email proofs, web review, Discord/Slack sharing | Good | Smallest |
| Ebook | iPad/Kindle/tablet portfolio, gallery submissions | Better | Small |
| Default | General-purpose proofing | Balanced | Medium |
| Prepress | Print-shop handoff, magazine submission | High | Large |
| Printer | Final print production, large-format output | Highest | Largest |
RAF stands for "Raw Format" and is Fujifilm's proprietary container for unprocessed sensor data. The file starts with the ASCII bytes FUJIFILMCCD-RAW, then a camera identifier and pointers to an embedded JPEG preview, metadata, and raw pixels. It's Fujifilm-specific mainly because X-Series bodies use the X-Trans 6x6 color filter array — a non-Bayer pattern that needs custom demosaicing. RAW converters that don't ship X-Trans-aware demosaicers (Adobe used the same engine for years before improving it; Capture One has its own; Iridient uses RHATS) produce visibly different results from the same RAF.
Each RAF is rendered to an image and embedded in the PDF, so what you see is what gets baked in. If you want your in-camera Provia, Velvia, Astia, Classic Chrome, or Acros look preserved exactly, render the RAFs through Fujifilm X RAW Studio first (it uses the camera's image processor for true color science), export TIFFs or JPEGs, then merge those. Generic web converters demosaic with libraw/dcraw and won't match Fujifilm's in-camera rendering pixel-for-pixel.
Any Fujifilm body that writes .raf — that includes the entire X-Series (X-T5, X-T4, X-T3, X-T30 II/III, X-H2/H2S, X-S20, X-Pro3, X-E4/E5, X100V/VI, X-M5) and the GFX medium-format line (GFX100 II, GFX100S II, GFX100RF, GFX50S II, GFX50R, etc.). Whether the file is uncompressed, lossless-compressed, or lossy-compressed (the X-T3 and X-Pro2 introduced compression options) doesn't matter — the merger handles all three.
"Single PDF" is the default and produces one combined document — best for portfolios, proofs, and contact sheets where a client scrolls through pages. "Individual PDFs" makes one PDF per RAF and zips them — useful when you want each shot as its own file (e.g., one PDF per product SKU, or per client to drop into separate folders).
Contained (default) fits the entire RAF inside the page with letterboxing — nothing is cropped, but you'll see white space when the photo's aspect ratio doesn't match the paper. Cover fills the page edge-to-edge, cropping whichever dimension overflows. For portfolios and proofs use Contained so the framing you shot is preserved; use Cover for wallpaper-style or full-bleed layouts.
It depends almost entirely on Compression Type and Image Quality. As a ballpark for 20 X-T5-era 26-MP RAFs: Screen at quality 75 yields roughly 8-15 MB; Ebook around 12-25 MB; Default 20-40 MB; Prepress 60-120 MB; Printer can exceed 200 MB. The original RAFs themselves are not embedded — only the rendered images — so total output is far smaller than the sum of the input RAFs.
Yes, and that's the most common reason to merge. A Screen- or Ebook-compressed PDF of 20-50 RAF previews typically fits under 25 MB, which clears Gmail's 25 MB attachment cap, Outlook.com's 20 MB cap, and Slack's 1 GB free limit. Discord's free tier is 10 MB per upload — for that, drop quality to ~60 and pick Screen, or split into two PDFs.
Generic web converters typically extract and use the embedded JPEG preview when one exists (every modern Fujifilm body writes one), then resize and re-encode it to fit the page. That's why merging is fast even for large RAFs — the tool isn't running a full X-Trans demosaic. If you need true raw-developed output, develop the RAFs in Fujifilm X RAW Studio or Capture One first and merge the resulting TIFFs/JPEGs with Merge JPG to PDF instead.
For a single file, Convert RAF to PDF is more direct. If you'd rather end up at JPEG or TIFF for editing, see Convert RAF to JPG or Convert RAF to TIFF — TIFF preserves more bit-depth headroom for further work.