Initializing... drag & drop files here
Supports: X3F
.x3f files written by Sigma SD and DP cameras. Batch is supported — drop in a whole card's worth of frames and each one is decoded and converted in parallel.X3F is Sigma's proprietary RAW format, written by cameras that use the Foveon X3 direct image sensor. Foveon Inc. was founded in 1997 (a National Semiconductor / Synaptics spin-off co-founded by Carver Mead) and the X3 sensor first shipped in 2002 inside the Sigma SD9; Sigma acquired Foveon outright in November 2008. What makes the format unusual is the sensor itself: instead of a Bayer mosaic that puts one color filter over each photosite and interpolates the missing two, the X3 sensor stacks three photodiodes vertically at every pixel and reads red, green, and blue at the same location. That means no demosaicing, and in practice less risk of moiré and false color in fine detail.
That distinctive capture is exactly why an X3F is rarely the file you hand off. It is a sensor dump, not a picture — it carries the raw photodiode data plus embedded metadata (EXIF, white balance, a JPEG preview), and the image data is wrapped in Sigma's own structure rather than a standard TIFF. The practical reasons people convert:
| Property | Value |
|---|---|
| Full name | Sigma X3F (Foveon X3 RAW) |
| Type | Proprietary camera RAW (lossless sensor data) |
| Origin | Foveon Inc. (founded 1997), now part of Sigma Corporation |
| First camera | Sigma SD9 (2002) |
| Sensor | Foveon X3 — three stacked photodiodes capturing RGB at every pixel |
| File signature | Begins with the ASCII marker FOVb; not a standard TIFF preamble |
| Embedded data | Raw photodiode data, EXIF, white balance, JPEG preview |
| Cameras | Sigma SD9 / SD10 / SD14 / SD15, SD1 / SD1 Merrill, SD Quattro, DP1 / DP2, dp Quattro |
| Native software | Sigma Photo Pro (free from Sigma); also LibRaw, dcraw, RawTherapee |
| Best converted to | TIFF (archival), JPG / WEBP (sharing), PNG (lossless web) |
Sigma's own free Sigma Photo Pro is the reference application and gives the most faithful Foveon color rendering, since it knows the sensor's color science. Beyond that, open-source RAW engines LibRaw, dcraw, and RawTherapee can decode X3F, and several editors (such as recent versions of Corel PaintShop Pro) read it. Support is patchy compared with mainstream RAW formats, which is the usual reason people convert X3F to a standard image first — a JPG or PNG opens in any viewer, browser, or phone without special software.
You keep the rendered look but lose RAW editing headroom. Converting bakes the demosaic-free Foveon capture into a finished image, so the per-pixel color and lack of moiré are preserved in what you see. What you give up is the latitude to recover highlights, re-balance white balance, or push exposure later — that flexibility lives in the RAW data. If you want a future-proof master, export X3F to TIFF at 16-bit and Lossless? set to Yes; use JPG or WEBP only for the shareable copy.
X3F is proprietary to Sigma and far less common than Canon CR2 or Nikon NEF, so fewer applications bothered to support it. The format also wraps its sensor data in Sigma's own structure rather than a standard TIFF container, and the RAW payload is obfuscated, which slowed third-party decoders. Open-source projects like dcraw and LibRaw reverse-engineered support over time, but many lightweight viewers and web tools still can't preview an X3F — converting to a standard format sidesteps the whole problem.
A 16-bit TIFF. In our testing, exporting an X3F to TIFF with Bit Depth set to 16-bit and Lossless? set to Yes preserves the full tonal range from the Foveon capture with no compression artifacts, which is what you want for a long-term master or further retouching. The tradeoff is size — a 16-bit TIFF is large. For everyday sharing keep a parallel JPG or WEBP copy, and reserve the TIFF as the archival original.
Not for you. The later "Quattro" Foveon design changed how the layers are weighted — the top layer captures most of the luminance/detail (roughly 20 megapixels) while the two lower layers record color at about a quarter of that resolution — but the resulting file is still an X3F. The converter decodes both the classic and Quattro X3F variants the same way and outputs a normal raster image; you don't need to pick a sensor type.
Yes. Drop a whole batch onto the page and each X3F is decoded and converted independently, then offered as individual downloads or a single ZIP. This is the fastest way to turn a card full of Sigma RAWs into share-ready JPGs or archival TIFFs in one pass. The realistic limit is total upload size and your connection speed rather than a fixed file count, since decoding runs on our servers.
Yes. Files are uploaded over an encrypted (TLS) connection, processed on our servers, and deleted automatically after a few hours. There's no sign-up, no watermark on the output, and files are never shared or made public. Because RAW files are large, the main thing to plan for is upload time on a slow connection — not a storage limit.