Initializing... drag & drop files here
Supports: X3F
X3F is Sigma's proprietary RAW format for cameras built around the Foveon X3 sensor — the only consumer image sensor that captures red, green, and blue at every pixel by stacking three photodiodes vertically rather than using a Bayer color filter array. The format debuted with the Sigma SD9 in 2002 and ran through the SD, DP, dp Quattro, and SD Quattro lines until Sigma paused Foveon production in the late 2010s. Files start with the ASCII signature "FOVb" and store brightness and color data 1:1 without demosaicing interpolation.
Outside SIGMA Photo Pro, Lightroom (with Sigma's plugin), or RawTherapee, almost nothing opens X3F natively — which makes sharing a frustrating round-trip. Merging X3F files into a PDF rasterises each shot into a universally viewable page, with no extra software needed on the recipient's end.
| Property | X3F | DNG (Adobe RAW) | JPEG |
|---|---|---|---|
| Owner / introduced | Sigma / Foveon, 2002 | Adobe, 2004 | Joint Photographic Experts Group, 1992 |
| Sensor architecture captured | Foveon X3 (3 stacked layers per pixel) | Any (Bayer, X-Trans, Foveon via converter) | N/A — already demosaiced |
| Compression | Lossless (per-channel) | Lossless or lossy | Lossy DCT |
| Typical file size (per frame) | 30-80 MB (Quattro), 10-30 MB (Merrill) | 20-60 MB | 2-10 MB |
| Native software | SIGMA Photo Pro, Lightroom (plugin) | Lightroom, Photoshop, Capture One, darktable | Universal |
| Bit depth | 12 or 14 bit per layer | 14-16 bit | 8 bit |
| Best for | Editing Foveon highlights/shadows | Cross-platform RAW archive | Sharing finished images |
The PDF engine uses Ghostscript-style preset names. Pick by destination, not by quality alone — "best" means smallest, not most lossless:
| Preset | Target use | Image DPI | Typical PDF size (20 X3F frames) |
|---|---|---|---|
| Screen (Best) | Email, web preview, phone viewing | 72 dpi | 5-15 MB |
| Ebook | Tablets, casual review | 150 dpi | 15-40 MB |
| Default | General-purpose handoff | 150 dpi | 20-60 MB |
| Prepress | Press-ready proofs, color-critical | 300 dpi | 60-150 MB |
| Printer | Inkjet/laser printing at home or studio | 300 dpi | 60-150 MB |
For a shoot you might print, use Prepress with Quality Percentage 90-95. For client proofs over email, Screen at quality 60-75 keeps you under most attachment caps. Need raster output instead of a PDF? Use convert-x3f-to-jpg or convert-x3f-to-tiff and merge those with merge-jpg-to-pdf or merge-tiff-to-pdf.
XConvert decodes the Foveon RAW with the embedded camera profile — close to the camera's JPEG render, not SIGMA Photo Pro's interpretation. SIGMA Photo Pro applies its own X3 Fill Light, color wheel, and noise reduction tuning that aren't part of the RAW data itself. For client-ready proofs you should still run frames through SIGMA Photo Pro or Lightroom's Sigma plugin first, export TIFFs, then merge those into a PDF using merge-tiff-to-pdf.
Yes. The X3F container has stayed consistent across both Foveon generations — Merrill writes 4,800x3,200 full-resolution data, Quattro writes its top layer at 5,424x3,616 and the lower layers at half pitch. Both decode to a flat RGB image before they hit the PDF, so the merger does not care which generation captured them. SD9 and SD10 files (the original 2002-2003 generation, 3.4 MP "Foveon X3") also work.
X3I bundles seven X3F frames captured at staggered exposures for higher dynamic range. The merger treats each X3I as a single composite frame using the embedded preview — it will not unpack the seven sub-frames. To get the SFD-merged result you'll need SIGMA Photo Pro to export a TIFF first, then merge.
A 50-frame SD Quattro H batch at Prepress quality (300 dpi) typically produces a 200-400 MB PDF — Quattro frames are huge. If that's too large for handoff, drop to Ebook or Screen preset, or split into two batches. To shrink an existing PDF without re-merging, run it through compress-pdf.
Yes — set Combine? to "Individual PDFs" and you get one PDF per uploaded X3F. Useful when each frame needs to be printed or signed-off individually, or when you're routing files to different folders downstream.
A4 (the default) works for most international handoffs. Letter is the US-paper equivalent. Pick Original if you want each PDF page to match the X3F's native aspect ratio (3:2 for SD/SD Quattro, 1.74:1 for SD Quattro H stretched mode) — the page sizes itself to the image rather than the other way around.
Contained preserves the entire Foveon frame inside the page margins (default — best when no detail can be lost). Cover fills the page edge-to-edge and crops whatever doesn't fit the page aspect ratio — useful for full-bleed photo book mockups but it will trim 10-15% off a 3:2 frame placed on A4. For critical-edge shots (architecture, product), keep Contained.
Sigma cameras embed orientation metadata that XConvert reads, so portrait-orientation frames usually flip correctly on their own. If a frame still comes through rotated, use convert-x3f-to-jpg with the rotate option first, then feed the JPEGs into merge-jpg-to-pdf. Other RAW formats (Canon CR2/CR3, Nikon NEF, Sony ARW) have their own merger pages — see merge-cr2-to-pdf, merge-cr3-to-pdf, merge-nef-to-pdf, merge-arw-to-pdf.
Files transit through XConvert's processing service to render the Foveon RAW (the decode is too heavy for browser-side WASM today), but they are not stored long-term and are not used to train any model. The merged PDF is delivered back to your session and the source X3F files are deleted from working storage shortly after the job completes.