#2410's dependency walk for a soloed 3D model was written before
reads_other_times took the effect count; it asks about all effects (None),
as the other walks do.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Carry property, Font and Media Replacement overrides into the collapsed nested EvalCtx so nested layers render beyond the precomp frame instead of falling back to a clipped, flattened image. Key the layer cache on instance overrides to keep distinct collapsed instances isolated.
The referenced model's solo render goes through the comp camera: its
depth of field (focus, aperture, iris, highlights) blurs it, none of
which is in the projection matrix — the evaluated Dof is hashed in now
(CodeRabbit, review of #2410).
A model with an animation clip selected is posed at its layer time
(Animation Speed and Looping are static and hashed): the layer time is
folded into the key, so every time is its own entry. A model without a
clip keeps one key across times, as before.
An Environment light's image lights Advanced 3D models, and its source
can be any layer's render (or any environment layer) — too open-ended
to hash: while one is on, the read is not cached. Each is covered by a
render test.
Conflicts were main's new code next to this branch's additions; both kept:
- crates/effects/src/lib.rs: main's Buf::pad_sides stays as main wrote it
(Transform's transform_grow bounds that padding itself); Buf::pad keeps
this branch's effective-pad clamp.
- crates/raster/src/lib.rs: main's quantize_px/round_half_up and this
branch's MAX_PADDED_SIDE (still equal to limits::MAX_SIZE on main).
CC Composite stays After Effects-pure: its Layer parameter is gone, so
parameter indices match AE again (scripts and expressions that address
parameters by index). The model reference lives in its own effect,
ec.3d.instance3d (3D Channel, CPU, one layer parameter), which replaces
the layer's pixels with the referenced layer's render — alpha included.
A referenced model's solo render is now hashed into the owner's cache key
instead of declining the entry (Bflat, review of #2410): the layer, its
animated values and transform, the mesh item, the comp camera and lights,
draft 3D, the renderer setting and the in/out gate. Covered by a render
test: the copy is cached, served on the second render, and re-rendered
when the camera pans or the model leaves its in/out range.
The mixdown evaluated levels and effect parameters once per block, and export mixes WAV/AIFF
in one-second blocks, so a level fade became a staircase that changed once a second. Animated
Audio Levels are now evaluated on a 1 ms grid of comp samples and interpolated in dB between,
and animated per-sample effects (Stereo Mixer, Backwards) run in 1 ms pieces with their
parameters at each piece. The grid is aligned to comp sample indices, so the result still does
not depend on the block size.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Resolve the Interpret Footage form against main's split into
interpret_footage_form and project_settings_form, keeping the Cineon
Settings fields and the pointUnits readback. Restore the Arc, rayon
prelude and cineon_frame imports that main's explicit use lists dropped.
- Default disk cache size 20 GB (was 100 GB); settings version 3 migrates the stored old
default and keeps a size the user chose.
- The cache keeps max(2 GB, 5% of the disk) free: sysinfo::disk_space (df -Pk / PowerShell
DriveInfo) is read in the background from memory_tick and caps the limit
(DiskCache::set_disk_space); cache.diskStats reports limitBytes.
- Eviction goes down to 90% of the limit, so the LRU sort runs once per tenth of the limit
written instead of on every write at the limit.
- Failed writes are counted (cache.diskStats failed) and the first is logged.
Also fixes#2063.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Accelerator::comp_frame_yuv420 (default None) renders a top-level comp frame and
converts it over the background to the H.264 export's 8-bit BT.709 limited-range
4:2:0, every byte within one level of the CPU conversion (the GPU kernel follows
it operation for operation, including NaN and half-away-from-zero ties, but a
driver may contract its arithmetic). A 4K frame then reads back 12 MB of planes
instead of 133 MB of RGBA f32.
The H.264 export asks the accelerator for planes when a frame needs nothing after
the render (progressive, project depth, no crop/resize, rendered at the output
size) and otherwise renders and converts on the CPU as before. Deferred
readbacks (a browser worker) keep the pass-based path. Sizes the kernel can't
take (width not a multiple of 8, odd height) are declined before any GPU work.
Like Cx::frames, h264_frames caps GPU frames in flight only while the
accelerator is up (map_frames_while with the same liveness check): once the GPU
stops for good, the remaining frames render on the CPU side by side instead of
one at a time.
rgba_to_yuv420 moves to effectcraft_raster::yuv so the export and the GPU tests
share one reference. The GPU tests skip without an adapter (tests::gpu()), as
the crate's others do.
Segmentations and rendered frames are keyed by their strokes, which now
segment differently where strokes overlap; bumping both versions keeps old
entries from showing the old result. Frozen mattes are stored in the project
and are untouched.
Much of a layer buffer is often transparent (a stroked ring, a glow-padded
text layer, a precomp canvas), yet every destination pixel sampled it
bilinearly, and 8/16 bpc comps then re-quantised the layer's whole rotated
bounding box.
- Bilinear warps skip 16x16 source tiles whose pixels (and a one-pixel
apron) all have zero alpha: their samples are transparent and were never
drawn.
- The fast affine warp reports the rectangle it changed and, given the
comp's levels, quantises each pixel as it blends it, so draw_comp has
nothing left to re-quantise for that layer. quantize_px moves to
effectcraft-raster for this.
- quantize_rect skips runs of four all-zero (+0.0) pixels, which quantise
to themselves.
- to_rgba8 / to_rgba8_over convert row by row instead of per 4-byte item.
Every pixel stays bit-identical: tests compare the warp against the
reference path, fused quantising against quantising after blending, and
the zero-run skip against per-pixel quantising.
With the previous commit, demo comp to H.264 (`cargo xtask bench --scenario
demo --format h264 --frames 300`, 11 interleaved runs on a 16-thread Ryzen):
wall 9.26 s -> 6.97 s median (-24.8%), 8.55 s -> 6.55 s min (-23.3%); CPU
time -23%. Output byte-identical.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Conflicts came from upstream splitting crates/project/src/lib.rs and moving the
frames tests. Re-applied the branch's change in the new places:
- Layer::same_pixels_with now in crates/project/src/layer.rs
- Comp::same_pixels_with, same_settings(o, with_markers), pixel_form_with and the
timing-marker helpers in crates/project/src/comp.rs (same_pixels_ignoring_markers
stays removed)
- Project::reads_markers in crates/project/src/item.rs
- crates/ui-egui/src/frames.rs uses reads_markers/same_pixels_with; the branch's
two marker tests replace the old ones in crates/ui-egui/src/frames/tests.rs
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
text/layout: upstream split paragraph() into Para/place_line helpers and moved the tests to
layout/tests.rs. The branch's changes are re-applied there: the resolved paragraph direction
is passed to margins_for/avail_for/align_for (RTL point text with Justify sits flush right),
first_ascent/plausible_first and ParaLine.first/style_first (box first baseline is the d
height, vertical text keeps the ascent), and the Justify All last-line cost in
optimal_breaks. The branch's seven new layout tests moved to layout/tests.rs.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
bflatastic