mirror of
https://github.com/storytold/photocraft.git
synced 2026-10-18 06:15:19 +02:00
8a6af66cf6
* TIFF: read and write CIE L*a*b* (CIELab, ICCLab); Lab documents stay Lab Photoshop saves Lab documents as CIELab TIFFs (PhotometricInterpretation 8, a* and b* signed), which PhotoCraft could not open, and PhotoCraft saved Lab documents as sRGB TIFFs. - codecs: a Lab channel layout (Lab, LabA) holding a* and b* as PhotoCraft's Lab documents do ((a + 128) / 255; 16 bits like the PSD importer), with its own Lab <-> sRGB conversion (same formulas as photocraft-color; codecs stays standalone) for formats without Lab, which now warn LabConverted. - TIFF read: CIELab and ICCLab (8 and 16 bits, alpha) are handed to the tiff crate as RGB and labelled Lab; a* and b* are converted from the file's encoding (signed a / 256a, or offset). - TIFF write: Lab and LabA as CIELab, as Photoshop writes it; float Lab is written with 16 bits (DepthReduced). - io: a Lab TIFF opens as a Lab document; a Lab document with one plain layer is written to TIFF in Lab (other formats still use sRGB). Photoshop 2025's Lab TIFFs open with the colours Photoshop shows (its own RGB conversion matches within one level at 8 bits and exactly at 16 bits), and Photoshop reads PhotoCraft's Lab TIFFs back with the same L*a*b* values. Fixes #4936 * DNG: decode lossy (JPEG-compressed) DNG tiles (FILE-215-11) DNG 1.4 lossy JPEG (compression 34892) stores each tile or strip as a baseline-DCT JPEG of 8-bit samples: a gray stream for CFA data, a YCbCr stream for three-sample LinearRaw. The decoder arm from #2021 read the streams but widened every sample by 257, so the values no longer matched the file's levels: an 8-bit WhiteLevel (or its default, 255) clipped everything to white and a 256-entry LinearizationTable was indexed past its end. It also required the JPEG frame to match the clipped row count, which rejected the padded edge tiles real tiled files carry. The arm is now `lossy_jpeg()` in sensor.rs. The decoded 8-bit values are the stored samples, and BlackLevel, WhiteLevel (default 255 for 8-bit data) and the LinearizationTable apply to them exactly as for every other compression. A tile may be coded at the full tile size (edge tiles padded, as the spec says) or at the image's row count; a strip at its row count; the width must be the segment's. BitsPerSample must be 8 and the JPEG's component count must equal SamplesPerPixel (1 or 3); anything else is an error, never a silent conversion. The frame header is checked before any pixel is decoded and the JPEG decoder's own size cap is the segment the IFD declares (already charged against the limits), so a hostile SOF cannot size a buffer past the tile. testgen writes true 8-bit lossy fixtures (`LossyJpegTiles`, and strips via the new `LossyJpegStrips`), accepts pre-encoded `lossy_segments` for hostile-input tests, exposes `baseline_jpeg()`, and leaves WhiteLevel out when `white` is 0. The DNG tests cover CFA and LinearRaw RGB from tiles and strips at 37x23 (partial edge tiles, a short last strip), the default white level, black/white scaling and a LinearizationTable against the uncompressed path, and hostile inputs: a truncated tile, frame-size and component-count mismatches, a 65535x65535 SOF under default and tight limits, a tile over the allocation budget, and 12/16-bit or 4-sample files. The hostile mutation set gains a lossy CFA and a lossy LinearRaw file. The parity doc's DNG row is updated; the scorecard row is left for the coordinating branch. * Raw: measure the Nikon D90 image area (4288x2848 at (10, 10)) Aligning the decoded D90 sensor plane with the camera's own JPEG puts the image at (10, 10) in the 4352x2868 sensor plane, giving the camera's 4288x2848 frame (10-pixel margins; the right border reads the camera's flat 255 from column 4310). Add the measured ImageArea so a D90 NEF opens on the camera's frame instead of the uncropped plane, and make a measured profile leave a file that carries its own colour calibration (a DNG) untouched. * Codecs: open Cineon (.cin) film frames read-only Cineon is the 10-bit logarithmic film-scan format DPX grew from. Read it from the same Kodak header family: the generic header (big-endian, orientation at 192, channel count at 193, the channel-1 bits/dimensions), one or three channels, three 10-bit samples packed into each 32-bit word on a word boundary. The 10-bit code values open as U16 (round(c x 65535 / 1023)); the file's log black/white points are a display choice and are not applied. A depth other than 10 bits or a channel count other than 1/3 is Unsupported; truncation errors rather than panicking. Read-only (no Cineon writer), so .cin joins the open filters and Cineon read closes part of #1049's G10 gap. * Vector: bound flattening memory, keep over-budget dashes, cheaper edge build (#5195) - flatten: a path's flattened points are capped at 1 << 20; past it every curve keeps a proportional share of its pieces (at least one), so hostile handles from files (vmsk, SVG, effect outlines) cannot exhaust memory. fill_rasterizer flattens the whole path under one budget. - stroke: past the dash step budget (or below coordinate precision), dash_polyline keeps the dashes made so far and draws the rest solid instead of dropping the stroke. Open polylines are no longer copied. - raster: coordinates are clamped to +-1e15 and edges with a non-finite slope (vanishingly short) are skipped, so huge finite coordinates no longer produce NaN coverage. fill_rasterizer sorts edges once instead of after every component (stable sort, identical order). - Cleanups: dead miter `half`, dead loop-length check in trace_mask, shared mask render body, map_knots helper, shared straight-segment test. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * PSD: cap pattern memory across a parse, reject mixed-depth pattern channels (#5198) - Patterns: one PatternBudget (decoded bytes + zero padding) is shared by every pattern of a block, a .pat file and the pattern blocks of one document; .abr patt sections are charged to the samples' budget. - Patterns: a channel whose depth differs from the first is rejected (the parsed pattern could not be written back). - Slices: descriptor numbers no longer wrap; out-of-range ids are no id, coordinates clamp to i32, and a v6 trailing entry without a usable id attaches its outsets to no slice. - LayerInfo::unpadded_len is computed from field sizes instead of serializing every record into a scratch buffer. - abr desc presets are moved out instead of cloned; UnFl and shmd counts use the bounded preallocation; ImageData::decode_channel (decoded every plane to return one, unused outside its test) is removed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * Vector: bound stroke outline points (#5211) - stroke_polygons dashes all lines first, then counts round caps and joins. When their full-resolution tessellation would exceed 1 << 20 outline points, every round cap and join gets the same smaller segment count (down to 4 per circle, 1 per join; area-exact radius kept). Strokes under the budget are unchanged; no piece is dropped. - The dash step budget (100000) is now shared by all lines of a stroke instead of per polyline, so many subpaths cannot multiply the dash (and cap) count. Past it the rest is drawn solid, as before. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * PSD: write tiff blocks without copies, inflate merged ZIP once - ImageSourceData::to_bytes writes the global mask and global blocks straight from the model with 4-byte alignment instead of cloning each block (data included) to set its padding. Same bytes. - The tiff transcoder's nested regions and blocks write into the parent's buffer (taken and handed back, error or not) instead of building their own output and copying it up per nesting level. Bodies that fail or are not transcoded are truncated before being copied as stored or dropped. Same bytes. - PsdFile::from_bytes_and_merged decodes the merged image with the parse when asked: a ZIP stream is inflated once (the decode's stream check is the parse's); the parse-only check runs only when the decode fails, to keep from_bytes' errors. PSD import uses it with merged_needed. - PsdBuilder::composite documents that Gray/CMYK composites stay straight. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * Content-Aware Fill: measure a thin hole by its thickness, not its bounds (#1366) Removing long, thin objects (coiled guitar strings, wires) left their outline as darker and lighter blotches and copied pieces of other strings into the fill. The pyramid shrank until the hole's bounding box was four patches across: for a string across the image that is a level where the hole has swallowed its surroundings, so the matches found there copy whatever looks like the blur, and the finer levels (refine-only for wide bounds) keep them. A hole at most a quarter as thick as its bounds are long now counts as thin: the pyramid stops while it is still about two patches thick, and that level starts from a membrane across it (over a few pixels, the surroundings themselves) instead of the onion peel, whose exhaustive search is what made a shallow pyramid unaffordable. Compact holes are unchanged bit for bit. Refs #1366 * perf(compose,gpu): effect caches reserve tile addresses instead of pinning layers Effect map cache entries kept a clone of their layer (a deep clone for groups) so the tile addresses the cache key hashes couldn't be reused. That kept an edited or deleted layer's pixels alive, uncounted, until the entry was evicted. Keep a Weak to each hashed tile instead: it reserves the address without keeping the pixels. The GPU compositor's styled-group pins and shape-source tile snapshots do the same. Part of #4530. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(history): count snapshot-only tiles against the history budget History panel snapshots outlived every budget: a Merged Layers snapshot is a full flattened copy that neither the history budget, the GPU budget's document bytes nor Purge accounted for. Trim now counts the tiles only snapshots hold as fixed cost, the GPU budget includes them, and Purge > Histories no longer reports tiles a kept snapshot still holds as freed. Part of #4530. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * Content-Aware Fill: every part of a thin selection keeps its surroundings (#1366) Two more causes of ghosts when removing strings and the small knobs selected with them: * Auto sampling summarises the selection's surroundings by a few clusters. Around a long selection those come from the strings' surroundings, so a small part elsewhere (a tuning knob in a differently coloured corner) lost its own surroundings and was filled from far away in another colour. Around a thin selection the Auto area now always reaches as far as the selection is thick. * The membrane that starts a thin hole now starts a patch radius out, past the anti-aliased fringe or halo that a selection drawn close to the object leaves outside it. Refs #1366 * Content-Aware Fill: peel a thin hole where that stays affordable (#1366) Starting every thin hole from a membrane smoothed Spot Healing on film grain (spot_healing_keeps_film_grain_through_overlapping_strokes: grain 9.0 against 12.1 around). The onion peel copies the grain; it was only given up because its exhaustive search is too slow on the big level a long thin hole stops at. It now starts the thin hole whenever hole cells times sources stay within 2^28 comparisons, and the membrane only beyond. Refs #1366 * Select Subject: keep subject parts coloured like the backdrop, leave enclosed gaps open Users on #2788 reported Select Subject dropping clothes that are close in colour to a plain wall and bridging the gap between an arm and the torso. Both come from the same place: the saliency pass labels low-contrast parts of the subject as background, the background colour model then learns them and explains them away, and the final blind hole fill closes every enclosed gap up to 5 % of the mask. Two changes in crates/algo: - A second GrabCut pass whose background model only knows the backdrop. A colour model fitted to the image frame (minus a margin around the salient blob, in case the subject runs out of the picture) marks every pixel it explains as probable background; the blob and whatever the backdrop cannot account for start as probable foreground. Only components that touch the first pass's result survive. - Hole filling that consults the colour models: specks (<= 0.2 % of the mask) always fill, larger holes up to 5 % fill only when their pixels fit the foreground model better than the background's. Tests: two regressions (a cream-shouldered figure on a near-white wall keeps its shoulders; a ring keeps its backdrop-coloured hole open), plus ignored timing and picture-dump helpers. The 6 MP first run goes from about 100 ms to about 200 ms on one core; the P29 repeat path is untouched. * Keyboard shortcut sets: Set menu, New Set, Save, Delete, .kys Export and Summarize (K2) Edit › Keyboard Shortcuts gets Photoshop's set management. A set is a saved copy of the bindings (Preferences.shortcutSets, the one in use in shortcutSet; "" is PhotoCraft Defaults), and the Set menu shows "(modified)" when the bindings differ from it. New Set…, Save and Delete wait for OK like the rest of the dialog. Export… writes the dialog's bindings as a .kys file (commands then tools, Photoshop's macOS key spelling), each row carrying our id as pc-id so PhotoCraft imports its own sets back exactly, whatever the label; Summarize… writes Photoshop's HTML list of every bound command by menu. edit.keyboardShortcuts takes useSet (before the call's edits), saveSet, deleteSet, sets and activeSet (after them), exportKys and summarize; every reply reports the sets, the set in use and whether it is modified. Bad names, keys or counts are errors that change nothing. * .kys import: tool keys and sets saved on Windows (K3) Photoshop's <tool> rows now import: each matches our tool by name (our tool names are Photoshop's) or by pc-id in sets PhotoCraft exported, and a letter other than the tool's own becomes its override, while its own letter keeps the tool cycling with the others (#5036). A set saved on Windows spells the primary modifier Ctrl where a Mac set writes Cmd and keeps Control for the Control key. A set with no Cmd or Opt anywhere but some Ctrl is read as a Windows set, its Ctrl as our Cmd: Ctrl+L is the Cmd+L default, not a Control-key override that would also mean Control on a Mac. The import reports `windows`. * cargo fmt the extension lists Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Gert Pellin <gert@pellin.be> Co-authored-by: Vikram Bhaskaran <vikram.bhaskaran@outlook.com> Co-authored-by: vangel76 <vangel1976@googlemail.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: Finn Newick <finnnk@gmail.com> Co-authored-by: Brandon Thomas <echelon@gmail.com> Co-authored-by: Diego Chagas <drochachagas@gmail.com> Co-authored-by: Tanish Manku <tmanku.dev@gmail.com>
photocraft-raw
A clean-room, pure-Rust camera raw decoder and developer. The crate is standalone (no workspace
dependencies; its decoding dependencies are flate2 for Deflate DNG and jxl-oxide, a pure-Rust
JPEG XL decoder, for DNG 1.7, native builds only), has no unsafe, does no I/O (&[u8] in), builds for wasm32-unknown-unknown
(sequential there, rayon-parallel on native) and never panics on hostile input: every offset is
bounds-checked and sizes are checked against Limits before allocating.
use photocraft_raw::{develop, DevelopOptions, Demosaic};
let dev = develop(&bytes, &DevelopOptions { demosaic: Demosaic::Ahd, ..Default::default() })?;
// dev.rgb: interleaved 16-bit RGB in ProPhoto RGB (ROMM primaries, D50, gamma 1.8)
// dev.warnings: anything approximated or not applied
photocraft-io uses it so opening a raw file yields a normal 16-bit RGB document tagged with the
built-in ProPhoto-compatible profile.
Sources (clean-room)
Implemented only from public specifications, papers and observation of files:
- TIFF 6.0, TIFF/EP (ISO 12234-2) and the Adobe DNG Specification 1.7.
- ITU-T T.81 (ISO 10918-1) Annex H: lossless JPEG, process 14 ("LJ92").
- The published description of Canon's CR2 container (header, raw IFD, slice tag 0xC640).
- Publicly documented maker-note / private tags (ExifTool's tag tables): Canon ModelID (0x0010), SensorInfo (0x00E0) and ColorData (0x4001), Nikon WB_RBLevels (0x000C) and BlackLevel (0x003D), Sony BlackLevel (0x7310), WB_RGGBLevels (0x7313), SonyRawFileType (0x7000) and SonyToneCurve (0x7010), the PanasonicRaw IFD0 tags, Olympus ImageProcessing (0x2040) and CameraSettings (0x2020) preview tags, the Fujifilm RAF header and metadata records (RawImageCropTopLeft / RawImageCroppedSize, FujiLayout, XTransLayout, WB_GRGBLevels, RawExposureBias) and FujiIFD (RawImageFullWidth / Height, BitsPerSample, StripOffsets / ByteCounts, BlackLevel, WB_GRBLevels).
- Nikon compressed NEF: Laurent Clévy's NEF structure notes and Bill Claff's "NEF Compression"
(prose: the
0x0096table layout, a lossy curve followed by a fixed-table Huffman stage), ExifTool's Nikon tag-name documentation and ITU-T T.81 Annex F / H. The Huffman tables are not stored in the files; they were recovered by black-box analysis of CC0 raw.pixls.us samples (the same scenes shot compressed and uncompressed) for LightCraft (storytold/lightcraft#86) and verified there on ~45 files from ~35 bodies. - Sony cRAW (ARW 2): H. Dietz, "Sony ARW2 Compression: Artifacts And Credible Repair" (Electronic Imaging 2016) and the RawDigger / diglloyd write-ups of the 11 + 7-bit scheme; the exact bit layout and tone-curve scale were established by observation of sample files.
- Panasonic RW2 RawFormat 5, uncompressed Olympus ORF and uncompressed Fujifilm RAF: established by observation of sample files (bit packing, page layout, sample justification; for RAF the three sample storages and the reversed order of the X-Trans / Bayer layout records, each verified by developing with every candidate phase against the camera's preview).
- Panasonic RW2 RawFormat 3/4 (vendor-compressed): the fixed-rate 16 bytes per 14 pixels, the predictor groups with per-parity predictors, the 2/4/8-bit field widths, the page-permuted downward bit order and the black level sitting 15 counts above the recorded tags were all established by observation of sample files.
- Demosaicing: Malvar, He & Cutler (ICASSP 2004); Hirakawa & Parks, "Adaptive homogeneity-directed demosaicing" (IEEE TIP 2005); for X-Trans and other non-Bayer CFAs, gradient-weighted green interpolation (Lu & Tan, "Color filter array demosaicking: new method and performance measures", IEEE TIP 2003) with edge-aware colour differences.
- McCamy's CCT approximation (1992); the Bradford chromatic adaptation transform.
- Per-file colour from the camera's embedded JPEG (
fit_look,src/look.rs): the method follows the prose description in LightCraft'sdocs/camera-preview-colour.md(storytold/lightcraft#165, Apache-2.0, same organisation); no LightCraft code was used. Pool-adjacent-violators isotonic regression (Ayer et al., 1955), Tukey's biweight (Beaton & Tukey, 1974), ridge regression (Hoerl & Kennard, 1970) and the CIE 1976 L*a*b* colour difference.
No code from dcraw, LibRaw, rawspeed, rawler, rawloader or darktable was read or used, and no camera colour tables were copied.
Support matrix
| Format | Status |
|---|---|
| DNG | Uncompressed (8–16 bit, packed or not), Deflate (compression 8, predictors 1 and 2), lossless JPEG, lossy JPEG (DNG 1.4 compression 34892: 8-bit baseline JPEG tiles or strips, gray for CFA, YCbCr-coded RGB for LinearRaw; decoded by zune-jpeg, the stored 8-bit values then go through the same LinearizationTable / BlackLevel / WhiteLevel as every other depth) and JPEG XL (DNG 1.7, e.g. Samsung Expert RAW; decoded by jxl-oxide; not in wasm32 builds, where it is Unsupported); strips and tiles; CFA (Bayer) and LinearRaw (three-sample colour, or one-sample monochrome such as Leica Monochrom and film-scanner DNGs, developed as neutral grey with no white balance or colour matrix); LinearizationTable, BlackLevel (+ repeat, DeltaH/V), WhiteLevel, ActiveArea, DefaultCrop, ColorMatrix1/2, CameraCalibration, ForwardMatrix, AnalogBalance, AsShotNeutral / AsShotWhiteXY, BaselineExposure, Orientation, OpcodeList2 GainMap (lens shading) |
| DNG (lossy JPEG, floating point; opcodes other than GainMap; ProfileGainTableMap) | Unsupported / not applied (reported) |
| CR2 | Lossless JPEG with slices, borders and as-shot white balance from the maker note, black measured on the masked border. CR2 has no CFA tag and the row phase varies by model, so it is measured from the data (the green diagonal), with a Canon model-ID table as the fallback (see src/cr2.rs) |
| CR2 sRAW / mRAW | Unsupported |
| NEF / NRW, ARW, PEF and other TIFF/EP raws | Uncompressed and lossless-JPEG (incl. Sony lossless ARW) CFA data |
| Nikon compressed NEF (compression 34713): lossless and lossy type 1 / 2, 12 and 14 bit | Decoded: fixed Huffman tables + maker-note 0x0096 seeds and curve (see src/nefc.rs). BlackLevel 0x003d is read in 14-bit units. "Lossy after split" files are unsupported (preview fallback) |
| Sony compressed ARW ("cRAW", SonyRawFileType 2) | Decoded: 11-bit min/max + 7-bit delta blocks, SonyToneCurve to 14 bits, the curve's 512 black level when no BlackLevel tag is written (the first-generation bodies, e.g. the ILCE-7) |
| Panasonic / Leica RW2, RawFormat 5 (12- and 14-bit packed) | Decoded, with PanasonicRaw black / white / WB / sensor borders |
| Panasonic / Leica RW2, RawFormat 3 and 4 (vendor-compressed, 12-bit) | Decoded; the black level is read 15 counts above the recorded tag; PanasonicRaw white / WB / sensor borders |
| Olympus ORF, uncompressed 16-bit (E-1, E-400…) | Decoded, with ImageProcessing black / WB / ValidBits / crop |
| Nikon "lossy after split" NEF, Sony "Compressed RAW 2", Pentax compressed PEF, RW2 RawFormat 6 and newer, Olympus compressed ORF | Unsupported: no public description of these codes was found apart from GPL decoder source, which this crate may not use (clean-room). photocraft-io opens the embedded JPEG preview instead |
| Fujifilm RAF, uncompressed (X-Trans I–V, GFX and Bayer bodies since about 2010) | Decoded: 16-bit containers (either byte order), 12-bit LSB-first packing, 14-bit packing in 32-bit words; X-Trans 6×6 or Bayer 2×2 layout, black, as-shot WB, crop, RawExposureBias, orientation from the preview's EXIF |
| Fujifilm compressed RAF, early FinePix / SuperCCD RAF (no CFA TIFF) | Unsupported (no public description of the compression); preview fallback |
| CR3 | Native CRX unsupported; bounded JPEG-track/PRVW preview fallback with CMT1 orientation. I/O labels the result as an 8-bit camera preview |
| X-Trans and other periodic non-Bayer CFAs (up to 16×16) | Demosaiced by one pattern-agnostic edge-directed method, whatever Demosaic is chosen |
Nikon maker-note compression modes 6/10 select checked little-endian packed 12/14-bit
rows even when TIFF Compression is 34713. The modern lossy 0x44 0x40 curve uses
its smaller encoded domain while preserving sensor ADUs. Valid Nikon CropArea
coordinates take precedence over model fallbacks. HE/HE* and lossy-after-split
remain unsupported. Four pinned CC0 Z6 files cover both bit depths and both paths.
Development pipeline
- Linearization table, per-position black level, scale to the white level.
- White balance (as shot; or grey-world when the file has none; or explicit multipliers), normalized so the smallest multiplier is 1, then clip to 1 so blown highlights stay white.
- Demosaic:
Bilinear,Mhc(Malvar–He–Cutler) orAhd(default) for Bayer; non-Bayer CFAs (X-Trans) always use the edge-directed generic method (demosaic_cfa). - Camera → XYZ (D50) per the DNG specification (ColorMatrix interpolated by the white's correlated colour temperature, or ForwardMatrix), or a camera look fitted to the embedded JPEG, → linear ProPhoto; exposure (BaselineExposure + user EV); the look's tone curve; gamma 1.8; 16 bits.
- Orientation.
Development validates public sensor/crop/CFA metadata before allocation. The working-memory limit checks peak live development buffers instead of summing disjoint phases. AHD limits simultaneous scratch bands only when required; normal load balancing and pixels are preserved. The default 2 GiB limit is unchanged and excludes caller-owned sensor/document storage.
Files without colour calibration (CR2, NEF, ARW, RW2, ORF, RAF…) have no camera matrix. Every one of
them carries a JPEG the camera rendered from the same sensor data, so fit_look estimates the
camera's rendering per file: both the sensor data (white-balanced as develop does) and the
decoded JPEG are reduced to a common grid of at most 256 cells on the long edge, and a 3×3
camera → ProPhoto matrix plus one monotone tone curve (applied per channel after exposure) are fitted
so that the developed raw matches the JPEG, by alternating isotonic regression for the curve and
weighted, ridge-regularised least squares for the matrix (residuals scaled to approximate L*,
Tukey biweights against misaligned edges and local tone mapping; clipped and near-black cells left
out). One cell in three is held out: the look is used only when it shows the same picture in colour,
the matrix is plausible and the held-out mean CIE76 ΔE drops to at most 85 % of the fallback's.
Pass it as DevelopOptions::camera_look; develop then notes "colour fitted to the camera's
embedded JPEG (mean ΔE …)". The crate cannot decode JPEG itself: photocraft-io decodes the
smallest preview at least 512 pixels long, fits and develops in one go.
Camera Raw's tuned import fits at the default exposure and as-shot white balance first,
then applies the user's Exposure, Temperature and Tint to that same rendering.
The D5100's measured image area is 4928 × 3264 at (8, 8) in a 4992 × 3280
sensor plane; the D90's is 4288 × 2848 at (10, 10) in a 4352 × 2868 plane; each
removes the border without changing levels, WB or calibration.
Measured image areas preserve an explicit file crop.
For a slightly tighter camera JPEG, fitting can align the reference proxy using a bounded 90–100% centred search on training cells. It requires a strong correlation improvement and retains the held-out colour gates; the developed crop and camera calibration are unchanged.
Without a usable fit (no preview, an unrelated, monochrome or differently shaped one, or a fit that
does not clearly help) the documented neutral fallback stays: the white-balanced camera channels are
treated as linear sRGB primaries, with no tone curve (plausible but flat, greyish colours; converting
to DNG gives calibrated colour), and develop says so.
Measured on 8 Nikon D3200 NEFs (12-bit lossy compressed, indoor scenes) against each file's embedded JPEG (mean CIE76 ΔE, 300-pixel grid, full develop output in sRGB): fallback 11.6–26.0, fitted 1.0–7.5; all 8 fits were accepted. The highest remaining errors are shots where the camera brightened shadows locally (Active D-Lighting), which one global curve cannot follow. The fit (preview decode included) takes about 50–70 ms on a 24 MP file. It is an estimate of the camera's look, not a measured calibration: picture styles and hue-dependent rendering are approximated.
Cameras with a profile measured against Photoshop's Camera Raw (src/cameras.rs: the Nikon D4) carry
a calibration, so no look is fitted for them. On 6 D4 NEFs against Photoshop 25.4's default Camera
Raw rendering (quarter size, medians over the frames) the measured profile is 0.048 stop from its
luminance with a mean CIE76 ΔE of 1.8; the look fitted to the embedded JPEG would be 0.30 stop and
ΔE 6.4, as it follows the camera's picture style instead. The open-time re-develop
(photocraft-io's import_raw_tuned) and the tone curve its previews use (tone_curve) take the
same look as the default open, so changing a slider does not change the colour rendering.
Tools
cargo run --release -p photocraft-raw --example rawinfo -- [--dump] [--demosaic ahd] [--png DIR] FILE...
prints what was decoded, times decode and develop, and can write sRGB PNG previews and the
embedded JPEG previews.
cargo run --release -p photocraft-io --example rawlook -- [--png DIR] FILE... times the
camera-JPEG colour fit, reports ΔE against the embedded JPEG with the neutral fallback and with the
fit, and can write 1200-pixel sRGB PNGs of both and of the JPEG; --synthetic times it on a
generated 24 MP file.