Split out of #1286 (Dehaze V3) so it can be reviewed and land on its own.
`min_filter` / `max_filter` are separable van Herk / Gil-Werman filters:
three comparisons per pixel per pass whatever the radius, edges clamped,
NaNs skipped like f32::min, a short buffer or empty plane returned
unchanged. `Image::transpose` (row-parallel) does the column pass.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Grid thumbnails looked much more denoised than the loupe. `nr_params` sized
the radii from the output's scale relative to the source buffer it was given,
not the original: a 256-px cell drawn from the 512-px thumbnail level used
half the full-size radius on pixels that each average ~23 sensor pixels, and
below 0.4x (0.35x for colour) the radius stopped shrinking at all. The
guided filter's eps (the variance treated as noise) also stayed at its
full-size value although downscaling divides the noise variance by 1/scale²,
so texture in small previews was flattened as noise.
- `SourceInfo::native_long` (filled by the engine from the catalog's photo
size, kept when the decoder supplies its own facts, carried to web
workers) and `Plan::detail_scale` = output pixels per original pixel
(from `px_per_long`, so a crop no longer changes it).
- `nr_params(s, scale)`: radii x scale with no floor, eps x scale². Below
1 px the smallest blur the box Gaussian makes is blended in by the ratio
of variances (`blur::box_sigma`); a fade under 1% skips the filter.
At scale 1 the parameters are unchanged: 1:1 views and full-size exports
render as before. The GPU takes the same parameters.
- RENDER_CACHE_VERSION 63: cached thumbnails are redrawn.
Measured on a 1536x1024 noisy synthetic scene (tests_nr_scale.rs), error
of the preview against the full-size render scaled down, as a fraction of
leaving NR off, luminance NR 50 at preview scales 0.5 / 0.33 / 0.25 /
0.167 / 0.1 / 0.05: before 0.53 / 0.70 / 1.04 / 2.26 / 3.87 / 6.07, after
0.28 / 0.37 / 0.66 / 0.90 / 0.99 / 1.00. A 256-px render from the 512-px
level now matches one from the original (mean 0.07 LSB vs 0.71 before).
On merge: the HDR histogram is applied in develop_reporting (#712 split develop), is skipped for
draft renders (a second render per drag tick), and HdrBins::bin clamps a hand-built split so it
can't underflow (test hdr_bins_with_any_split_stay_in_range).
An HDR edit's preview shows its SDR rendition, so its histogram said nothing
about the highlights above SDR white. Now it shows the HDR rendition:
`pipeline::hdr_histogram` renders the edit small (384 px, float HDR, same
framing) and bins it with SDR on the left half and the HDR range on the right
(`Histogram::of_hdr`, `HdrBins`), on a fixed scale of Hdr::MAX_EV_LIMIT (5)
stops above SDR white so it doesn't rescale when the headroom limit changes.
The engine's `develop()` (CPU and GPU renders) and the web worker swap it in
for HDR edits; high-bit-depth, windowed and proofed requests keep the image's
histogram, and a display profile's histogram doesn't replace it. Web workers
carry the HDR bins across their wire (`Histogram::to_words` / `from_words`).
The panel draws it in Lightroom's layout: a solid line at SDR white, dashed
lines between the stops, the headroom limit as an orange line with the range
past it dimmed, a bar under the plot (white; one band per stop in Visualize
HDR range's colours while that is on), SDR / HDR labels and an info tooltip.
Follow-up to #646, ported from #466.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- color: RgbSpace::to_xyz/from_xyz and bradford() no longer expect a
matrix inverse. Degenerate (collinear) primaries, which readers reject
before building a space from a file, leave the primaries unscaled /
give the identity; the constant Bradford and OKLab matrices fall back
to the identity, which can't happen (round trips are tested).
- pipeline geometry: sample() matches (mode, warp) instead of
expect("warp"); a warp plan without a warp (sample_plan never makes
one) samples through the equivalent affine.
- color, geom, raster, develop, scenes, preview and pipeline now #![deny]
unwrap/expect/panic/unimplemented/todo/unreachable; clippy.toml as in
the steering PR.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Rust 1.99 (stable since 2026-09-28) adds clippy::chunks_exact_to_as_chunks and widens
clippy::question_mark, which failed `cargo xtask ci` on main. Same behaviour: the
remainder that chunks_exact skipped is the remainder as_chunks returns separately.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- memory::budget (default min(25% RAM, 1.5 GiB); LIGHTCRAFT_MEMORY_MB, app.memoryBudget):
half for engine caches (sources + rendered previews, LRU across them via a global
use tick; the photo on screen stays), a quarter for decodes in flight (WorkGate:
thumbnails, prefetch and import probes wait; loupe/export loads never wait), an
eighth for the GPU free pool.
- GPU: best-fit buffer reuse (<= 25% slack), buffers released outside a render are
pooled at once (no stranded retired lists), pool trimmed after 3 s idle.
- Decodes drop file bytes and raw samples early, colour/crop/orient in place.
- macOS: apps install malloc_zone_pressure_relief as the release hook (after each
decode, import and idle trim): freed large blocks no longer stay resident.
14 raws + loupe + 12 steps (M4 Pro): peak RSS 1.9-2.3 GB -> 0.83-0.99 GB, footprint
2.5-2.6 GB -> 0.91-0.96 GB, heap peak 930 -> 656 MB; stepping median ~13 ms.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New L3 crate `lightcraft-gpu`: a wgpu device (Metal/Vulkan/DX12; none on wasm32),
WGSL kernels compiled once, buffer/dispatch/readback helpers, and a renderer that
runs the per-pixel stage (tone map, highlights/shadows, clarity/texture/sharpen,
dehaze, local mask terms, OkLCh colour ops, grading, vignette, gamut map, sRGB
encode, curves, grain) on the device. Earlier stages still run on the CPU and are
uploaded (per-stage hybrid). Stage results are cached per view next to the CPU
StageCache (new `StageCache::extension`).
The CPU pipeline stays the oracle and shares its resolved parameters with the
kernels: `finish::FinishParams` / `mask_terms`, `pipeline::plan` (stage-cache keys),
`local::plane_sigmas` / `nr_params` / `wb_matrix_for`, exact LUTs and OkLab
matrices exported for the shader prelude. Equivalence tests (skip without an
adapter): mean |d| < 0.5 LSB, max <= 3 LSB; measured max 1 LSB on Apple M4 Pro.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The guided base, clarity band, texture band and dark channel (and the paired blurs inside the
guided filter) are independent; each alone scales poorly (subsampled grids, short rows), so they now
run under raster::par_join. Same arithmetic, identical pixels. render_bench gains a per-feature
breakdown of the per-pixel stage (ONLY=finish).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- finish: sRGB encode through an interpolated 4096-entry table (< 2e-5 from the exact curve,
tested), exp2 instead of powf(2, x).
- guided / guided_fast: coefficient and output passes run row-parallel (were serial loops);
Image::zip_map.
- blur_bench reports process CPU time next to wall-clock (meaningful on a loaded machine).
- RENDER_CACHE_VERSION 3 (exposure-after-filters and the table change pixels by <= 2 LSB).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
3–10x faster blurs on a 3000x2400 plane (min-of-9; machine under load, see blur_bench example).
LIGHTCRAFT_PROFILE=1 now also times log-lum, base, clarity, texture, dehaze, masks and NR.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
