Convert X3F to APNG Online

Turn Sigma X3F Frames Into an Animated PNG

X3F is Sigma's RAW format, the container written by Foveon-sensor bodies — the SD and DP lines, the Merrill generation and the Quattro cameras. (Sigma's later Bayer-sensor fp bodies write DNG instead, so those files do not belong on this page.) APNG is an animated still-image format that keeps full colour losslessly. Put those together and this page does one specific job: it takes X3F frames, renders each one, and assembles them into a single looping PNG you can post or embed without Sigma Photo Pro on the other end. Files upload over an encrypted connection, are processed on our servers, and are deleted automatically after a few hours — no sign-up, no watermark.

X3F at a Glance

How to Convert X3F to APNG

1. Upload Your X3F Files: Drag and drop your Sigma RAW files or click to browse. Only .x3f input is accepted here. Upload order is playback order, so add a bracketed set, a focus-stack run or a timelapse in the sequence you want it to run.
2. Choose Combine?: Single animation is preselected and merges every uploaded X3F into one APNG, one frame per file. Individual files converts each X3F on its own instead.
3. Set the Framerate: The Framerate dropdown opens on 10 FPS (Recommended) and spans 1 FPS (Slideshow) to 50 FPS (Max Speed). Each frame is written with an exact 1/N-second delay, so 2 FPS gives a half-second hold per shot — usually what you want for a stills sequence — and 20 FPS flips through a stack quickly.
4. Set Resolution and Convert: Keep original is preselected. Resolution Percentage, a Preset Resolution (height-based, width follows the source aspect) or an explicit Width × Height will all shrink a full-frame Sigma capture down to something postable. Click Convert and download the APNG.

One File or Many: What Comes Out

Before assembly, each rendered frame is normalised to a PNG capped at 4,096 pixels on its long edge, so a full-resolution Sigma file is scaled down whether or not you set a resolution yourself. That ceiling is a practical benefit here: it stops a handful of RAW frames from becoming a hundred-megabyte animation.

When This Conversion Makes Sense

• Exposure brackets — flipping through a three- or five-shot bracket at 2 FPS shows the tonal range far better than a contact sheet does.
• Focus stacks — stepping through the stack in order makes the plane of focus travel through the subject, which is a good way to check coverage before merging.
• Timelapse and interval sets — a Sigma body shooting on an intervalometer produces exactly the kind of ordered sequence this page assembles.
• Product turntables — a set of shots around an object becomes a loop that plays inline anywhere, with no video player and no autoplay restrictions.
• Not for a single frame — one photo does not need an animation container; X3F to PNG or X3F to TIFF is the better output for one image.

Frequently Asked Questions

Do I need Sigma Photo Pro installed for this to work?

No. The X3F is decoded on our servers with an open-source RAW pipeline, so nothing needs to be installed on your side. Sigma's own software still renders Foveon colour differently from any third-party decoder, and LibRaw's published camera list flags the early SD9, SD10, SD14 and SD15 bodies as raw-decode-only — so if a frame from one of those cameras comes back with a colour cast, render it in Sigma Photo Pro first, export PNGs, and assemble those with PNG to APNG.

I uploaded one X3F and the APNG does not move — why?

Because a single image is a single frame. The assembler collapses it and writes an ordinary static PNG under the .apng name: fully valid, transparency intact, but no animation chunks. Add at least two X3F files with Combine? set to Single animation to get something that actually loops.

Does the full Foveon detail survive into the APNG?

The pixel detail does, up to the resolution you end up with; the RAW latitude does not. Each X3F is demosaiced and rendered to an 8-bit-per-channel sRGB frame before assembly, and every frame is capped at 4,096 pixels on the long edge, so the highlight and shadow headroom that made you shoot RAW is baked in at that point. APNG itself adds no further loss — it stores those rendered frames without re-compressing them lossily. Keep the X3F as your negative and export X3F to TIFF when you need an editable master.

Why APNG rather than GIF for photographic frames?

Because GIF quantises every frame to 256 colours. On skies, skin and any smooth gradient — exactly the tonal transitions a Foveon sensor is bought for — that shows as banding or dithering noise. APNG stores each frame in full colour with lossless compression, so the frame you see is the frame that was rendered. The cost is size: expect an APNG to be substantially larger than the equivalent GIF.

How large will the finished APNG be?

Larger than you expect, because size scales with pixels multiplied by frames and nothing is thrown away. A ten-frame sequence at a couple of thousand pixels wide will run to megabytes. The two controls that actually move the needle are Image Resolution and the number of frames you upload; the Framerate changes the playback speed, not the byte count, once the frames are fixed.

Does APNG keep transparency, and does an X3F even have any?

APNG carries a full 8-bit alpha channel and this pipeline preserves alpha wherever the source has it. A camera RAW does not — a Foveon capture is fully opaque — so in practice every frame here is solid. Transparency becomes relevant only if you render your frames elsewhere with a cut-out background and bring those in as PNGs instead.

Which viewers will play the result?

Chrome from version 59, Firefox from version 3, Safari from version 8, Edge from version 79 and Opera from version 46, which caniuse puts at about 95.94% of global traffic. Anything that does not support the animation, including a lot of desktop image software, falls back to showing the still image — that behaviour is written into the W3C PNG Third Edition Recommendation of 24 June 2025, which folded APNG into the PNG standard as section 4.9.

Can I control how long each photo is held on screen?

Yes, through the Framerate dropdown, which sets an exact per-frame delay of 1/N second for every frame. 1 FPS (Slideshow) holds each shot for a full second, 2 FPS for half a second, 10 FPS for a tenth. There is no per-frame timing control — every frame in the animation gets the same delay.