Initializing... drag & drop files here
Supports: X3F
.x3f) from a Sigma SD, DP, dp Quattro, fp, or fp L camera. Batch is supported — drop an entire shoot at once. Files stay in your browser session; nothing is sent to a server account.X3F is Sigma's proprietary RAW format for Foveon X3 sensors — the three-layer photodiode stack first shipped on the Sigma SD9 in 2002 and used since across SD, DP, dp Quattro, and fp/fp L bodies. Foveon files are large, only natively readable in Sigma Photo Pro and a handful of RAW processors, and impossible to share or stream as photos. Re-rendering a folder of X3F frames as a single AV1 video gives you a portable, browser-playable, royalty-free deliverable that's typically 30% smaller than HEVC for the same visual quality (per the Alliance for Open Media's compression benchmarks). Typical scenarios:
Need a different output? Render the same X3Fs as MP4 or WebM instead, or export individual frames as JPG or PNG for stills work. Once your AV1 is built, Compress AV1 can shrink it further for email or chat.
| Property | AV1 | HEVC (H.265) | H.264 |
|---|---|---|---|
| Released | March 2018 (AOMedia) | 2013 (ITU-T / ISO) | 2003 (ITU-T / ISO) |
| Licensing | Royalty-free (Alliance for Open Media) | Patent pools (MPEG LA, HEVC Advance, Velos) | Patent pools (MPEG LA), free for streaming since 2010 |
| Compression vs H.264 | ~50% smaller at same quality | ~25–50% smaller | Baseline |
| Compression vs HEVC | ~15–30% smaller | Baseline | Larger |
| Browser playback | Chrome 70+, Firefox 67+, Edge 121+, Safari 17+ (partial; native on Apple Silicon M3+) | Safari, Edge with HEVC extension, Firefox 134+ on Win/macOS | All modern browsers, universal |
| Hardware decode | Intel 11th gen+, AMD RDNA 2+, Nvidia RTX 30+, Apple M3+ | Most chips since ~2014 | Nearly every chip since 2005 |
| YouTube role | Default for 4K+ on capable devices since 2020 | Limited use | Default for ≤1080p on older devices |
| Best for | Royalty-free streaming, modern browsers, 4K archive | iOS/macOS ecosystem, broadcast | Maximum compatibility, older devices |
The Constant Quality mode maps to an AV1 CRF (Constant Rate Factor) value internally. Lower CRF = higher quality and larger file. Treat these as starting points for a 1080p slideshow at ~30 fps:
| Preset | Approx. CRF | Visual result | Use for |
|---|---|---|---|
| Lowest | ~55 | Visible blocking on detail, smallest file | Voicemail-grade proof clips, quick share previews |
| Low | ~45 | Noticeable softening on fine detail | Mobile previews, quick client check |
| Medium | ~35 | Good general quality, modest size | Web/social slideshows |
| High | ~28 | Visually transparent on most displays | Portfolio reels, client deliverables |
| Very High (Recommended) | ~22 | Near-source quality, larger file | Archive, 4K masters, time-lapse |
| Highest | ~15 | Effectively lossless to the eye | Mastering passes, regrading later |
CRF mappings vary by encoder build (libaom vs SVT-AV1); the values above are typical libaom-style ranges. AV1 CRF generally runs higher than x264 CRF for the same perceptual quality.
Because AV1 is the only practical way to package a sequence of Foveon X3F frames into a single file that streams in a browser, posts to YouTube without re-encoding, and stays small. X3F itself is a still-image RAW container; turning a folder of them into AV1 produces a slideshow or time-lapse that anyone can watch. The conversion demosaics each Foveon frame, scales it to your target resolution, and encodes the sequence with AV1 — you lose the editable RAW data but gain a portable deliverable.
Yes — any video codec, AV1 included, encodes to YUV 4:2:0 8-bit (or 10-bit if you target it specifically) and discards the per-pixel three-layer color samples that make Foveon distinctive. If you need to preserve the original RAW for re-editing, keep the X3F files and only treat the AV1 export as a deliverable. For stills that retain more of the Foveon character, export individual frames as PNG or TIFF instead.
Three things matter. First, licensing — AV1 is royalty-free under the Alliance for Open Media patent license; HEVC has three active patent pools (MPEG LA, HEVC Advance, Velos Media) that charge encoder/decoder fees and per-stream content royalties. Second, compression — AV1 is roughly 15–30% smaller than HEVC at matching quality per AOMedia's published comparisons. Third, browser playback — AV1 plays natively in Chrome 70+, Firefox 67+, and Edge 121+, while HEVC needs a paid Microsoft extension on Edge and partial Firefox support that only landed in Firefox 134.
Every Sigma camera with a Foveon X3 sensor, starting with the SD9 in 2002. That spans the SD-series DSLRs (SD9, SD10, SD14, SD15, SD1, SD1 Merrill), the DP-series compacts (DP1/DP2/DP3 and their Merrill variants), the dp Quattro series (dp0/dp1/dp2/dp3 Quattro), the SD Quattro mirrorless bodies, and the Bayer-sensor fp/fp L (which also write X3F when configured for Sigma Photo Pro workflows). Files from older bodies use the original Foveon X3 layout; Quattro bodies use a modified Quattro X3 layout — both are read.
The encoder sets the output frame rate from your Image Duration: 1 second per frame yields ~1 fps, 5 seconds per frame yields 0.2 fps (you'll usually want playback software to handle stills), and 1/24 second per frame yields a film-rate 24 fps clip suitable for time-lapses. For a smooth slideshow at standard playback rates, use 1/24, 1/30, or 1/60 second per frame; for a paced slideshow with each photo on screen, use 2–10 seconds per frame.
Conditionally. AV1 hardware decoding shipped on Apple Silicon M3 (Macs from late 2023+), iPhone 15 Pro and 15 Pro Max (A17 Pro chip), iPhone 16 series, and Apple TV 4K (3rd generation, 2022 onward with the A15 Bionic). Older Apple devices fall back to software decoding in Safari 17+, which works for 1080p but can struggle at 4K. If you need universal Apple compatibility today, render the same shoot as MP4 with H.264 instead.
Roughly: for a 1080p AV1 slideshow at the Very High preset, expect about 1–3 MB per minute of playback for talking-head-style stills, 3–8 MB per minute for high-detail nature/landscape stills, and 10–30 MB per minute for 4K time-lapses with lots of motion. AV1 sizes can vary 2–3× depending on scene complexity since CRF allocates bits where they're needed. If you need a hard target, use Constraint Quality mode to cap the maximum bitrate.
Two reasons. First, X3F is a Foveon RAW format — the converter has to demosaic the three-layer sensor data and apply white-balance and color-profile decoding before it has pixels to encode, which is heavier than reading a JPEG. Second, AV1 encoding itself is slow — libaom is roughly 5–20× slower than x264 at comparable quality, in exchange for the size savings. A 100-frame X3F → AV1 1080p conversion can take several minutes; 4K can take 10× longer. SVT-AV1-backed encoders are faster but still trail H.264.
Not from this page — the X3F-to-AV1 flow is image-to-video with no audio input. To add a soundtrack, convert to AV1 here first, then combine the AV1 with an audio file using a video editor (DaVinci Resolve, Shotcut, and Kdenlive all import AV1) or remux with ffmpeg (ffmpeg -i slideshow.mp4 -i music.mp3 -c copy out.mp4).