Commit Graph
5214 Commits
Author SHA1 Message Date
vangel76 51bb54efde Merge branch 'plane-store-on-craft-exr' into decode-cap-on-craft-exr 2026-11-18 13:17:20 +01:00
vangel76 f6a867a951 Disk cache: an entry larger than the cache is skipped before it is sealed (#3750 review) 2026-11-18 13:17:19 +01:00
vangel76 3c48ddc72c OpenEXR limit: planes let go that are still in memory count toward the limit, so a request is refused for now rather than read beside them (#3750 review) 2026-11-18 13:07:17 +01:00
vangel76 0f872a46a6 Merge branch 'plane-store-on-craft-exr' into decode-cap-on-craft-exr 2026-11-18 13:03:28 +01:00
vangel76 c2c40200f6 Caching: the test of a still cached twice checks that its memory goes once nobody holds it (#3750 review) 2026-11-18 13:03:27 +01:00
vangel76 596a5ac55d Merge branch 'plane-store-on-craft-exr' into decode-cap-on-craft-exr 2026-11-18 13:03:07 +01:00
vangel76 9bbc0847ad Caching: a still cached both whole and as a frame can be let go, and a plane entry without memory to seal it is skipped (#3750 review) 2026-11-18 12:57:37 +01:00
vangel76 f0725e18bb OpenEXR limit: a decode thread that took over a decode lets the held planes go itself, and the region test reads the footage's size (#3750 review) 2026-11-17 18:12:03 +01:00
vangel76 6bee053990 Merge branch 'plane-store-on-craft-exr' into decode-cap-on-craft-exr
# Conflicts:
#	crates/media/src/exr_planes.rs
2026-11-17 17:40:27 +01:00
vangel76 3494ef8552 Caching: an oversized Cryptomatte plane stays beside held ones, disk copies are made a job at a time, and the review tests prove what they claim (#3750 review) 2026-11-17 17:40:16 +01:00
vangel76 e6a3c32357 OpenEXR limit: planes let go are gone before the request is read, a claimed decode without memory isn't read again, and the oversized-header test keeps the fixed limit (#3750 review) 2026-11-17 17:05:03 +01:00
vangel76 79fdc7b21b Merge branch 'plane-store-on-craft-exr' into decode-cap-on-craft-exr
# Conflicts:
#	crates/media/src/exr_planes.rs
2026-11-17 17:00:55 +01:00
vangel76 5d0bfe7869 Caching: memory is let go only where it frees memory, Purge keeps nothing decoding meanwhile, and regions read the rows they show (#3750 review)
- Cryptomatte planes: a plane a render holds is not offered for its gone channels, and the entry
  limit holds while renders hold every plane.
- Footage frames and whole stills held elsewhere are not offered to the memory accountant.
- Purge (clear frames) makes decodes in flight stale: their frames are not kept or joined.
- Plane store: a size refusal of a merged decode is not handed to a request that fits alone;
  slimming a shared set and copies for the disk cache take memory fallibly, and planes the disk
  cache can't take are not copied.
- Disk cache: the plane encoder's scratch is taken fallibly; no memory skips the entry.
- Region of interest: rows follow the projected transform of a 2D layer under a 3D parent, and
  a frame of another size than its footage is read whole.
- The box filter's work bound is checked without a wall clock.
2026-11-17 17:00:29 +01:00
vangel76 35bfef1cfb Merge the earlier fix/exr-decode-cap-3749 history; its changes are rebuilt above on craft-exr's seekable reader (#3749) 2026-11-17 11:10:55 +01:00
vangel76 c12320771c Merge branch 'plane-store-on-craft-exr' into decode-cap-on-craft-exr 2026-11-17 11:10:48 +01:00
vangel76 de69ca4866 OpenEXR limit: a limit set while the system's memory is read stays, and a reading without answer is asked for again (#3750 review)
The session's figure set during the reading was replaced by the older reading's
answer; it is now published only while nothing set a limit meanwhile. A reading
that gave no answer now makes earlier refusals for size out of date, so the
second reading is asked for by the next request.
2026-11-17 11:10:48 +01:00
vangel76 1613e8d24a OpenEXR: read through craft-exr's seekable reader as merged in craft-libs (#3301) 2026-11-17 11:10:35 +01:00
vangel76 fc4eddf2d5 Merge upstream/main into the plane store branch (#3301)
# Conflicts:
#	Cargo.toml
2026-11-17 11:10:07 +01:00
vangel76 c205773d92 Caching: memory is counted where it is held, and whole layers stay whole under a viewer region (#3301 review)
- The memory accountant is offered only layer buffers and preview frames nobody
  else holds, so what it counts as freed is freed; the frame cache's recycled
  buffers count and go first.
- A Cryptomatte plane over the matte cache's own limit is kept only where the
  common budget trims it.
- Disk cache entries read back take their buffers fallibly: memory the system
  can't give is a miss.
- The Layer panel, sampleImage and effect inputs read whole footage frames under
  a viewer region of interest.
- A 16K Cryptomatte source at half resolution is box-filtered, not looked up
  nearest-neighbour.
- The System Compatibility Report shows the cache budget in force.
- A Project panel thumbnail render never starts under a key held back from an
  earlier state.
- Convert Expression to Keyframes makes no keys on a property that can't be
  animated.
2026-11-17 11:09:13 +01:00
Dylan M. Taylor aeedb92ef2 Merge pull request #3801 from storytold/refactor/shared-svg-transform
svg: reuse shared permissive transform-list parsing
2026-11-17 03:16:41 -05:00
Dylan M. Taylor ce2df07d52 svg: integrate merged transform parser with current main 2026-11-17 02:16:35 -05:00
Dylan M. Taylor bb54263240 Merge remote-tracking branch 'origin/main' into refactor/shared-svg-transform 2026-11-17 02:16:32 -05:00
Dylan M. Taylor 7a34ceffe2 Merge pull request #3798 from storytold/refactor/shared-svg-colors
svg: use shared floating-point colour compatibility parser
2026-11-17 00:39:00 -05:00
Dylan M. Taylor ed7b40ab4d Merge shared CSS integration while retaining transform compatibility 2026-11-17 00:05:47 -05:00
Dylan M. Taylor 69b4115b36 Merge remote-tracking branch 'origin/main' into refactor/shared-svg-colors 2026-11-17 00:04:44 -05:00
Dylan M. Taylor c7068ae948 Merge pull request #3797 from storytold/refactor/shared-svg-stylesheet
svg: move stylesheet parsing and matching to craft-svg
2026-11-17 00:04:16 -05:00
Dylan M. Taylor 9953ca314d svg: consume shared compatibility transform-list parser 2026-11-16 23:22:52 -05:00
Dylan M. Taylor 76d98ecd4b svg: retain prefix-path API alongside color compatibility 2026-11-16 23:10:16 -05:00
Dylan M. Taylor f65ad2a42d Merge current main and retain CSS plus prefix-path shared APIs 2026-11-16 23:10:13 -05:00
Dylan M. Taylor d0cbc53494 Merge remote-tracking branch 'origin/main' into HEAD
# Conflicts:
#	Cargo.lock
#	Cargo.toml
2026-11-16 23:02:54 -05:00
Dylan M. Taylor 41da144555 Shape operators: cap each expanding operator's output (#2948) 2026-11-16 21:58:09 -05:00
Dylan M. Taylor defa16a26f svg: use shared bounded prefix path parsing (#3800) 2026-11-16 21:58:04 -05:00
Dylan M. Taylor 99a9af86b1 svg: pin color compatibility to merged shared library 2026-11-16 21:50:00 -05:00
Dylan M. Taylor d5755f5dd9 Merge remote-tracking branch 'origin/main' into refactor/shared-svg-colors 2026-11-16 21:49:57 -05:00
Dylan M. Taylor b29778b577 svg: pin stylesheet consumer to merged shared implementation 2026-11-16 21:10:12 -05:00
Dylan M. Taylor ca5ca106fa Merge remote-tracking branch 'origin/main' into refactor/shared-svg-stylesheet 2026-11-16 21:09:59 -05:00
Dylan M. Taylor 6d908d1301 Merge pull request #3215 from vangel76/cc-ball-action-params-3197
Ball Grid effect has all of After Effects' CC Ball Action controls, including flat balls without shading
2026-11-16 20:39:25 -05:00
Dylan M. Taylor 22dbd1612f Merge pull request #3246 from vangel76/waveform-reversed-3238
The Timeline waveform shows what is really heard for reversed layers and the Backwards effect
2026-11-16 20:37:23 -05:00
Dylan M. Taylor bf29ef51b0 Merge pull request #3794 from vangel76/fix/ocio-test-windows-path-3793
OCIO LUT test passes on Windows
2026-11-16 20:31:41 -05:00
Dylan M. Taylor a9d53efbb5 svg: route paint colours through shared floating-point compatibility parser 2026-11-16 19:51:35 -05:00
Dylan M. Taylor 3972e5181b svg: delegate stylesheet parsing and matching to craft-svg 2026-11-16 19:38:19 -05:00
vangel76 ba2cb5f759 OpenEXR limit: one decompressed block stays at 8 GiB whatever memory a read may use (#3749 review)
The read's limit (the floor or the memory budget) bounds the planes it
allocates; one block the decompressor allocates whole is capped apart, at
the fixed floor (8 GiB, 512 MiB in the browser), with craft-exr's
max_block_bytes. A larger block is a damaged or hostile header, refused
for its size before anything is read. That refusal doesn't depend on the
limit, so remembering it for requests asking at least as much stays
right; after a limit change it is only decided once more.
2026-11-16 23:38:19 +01:00
vangel76 c0c53353fe OpenEXR limit: memory the system refuses is neither remembered nor read again at once, and never handed to the whole file's decode (#3749 review)
craft-exr's OutOfMemory is now a refusal of its own (ReadError::OutOfMemory,
Refused::NoMemory) instead of a failed read: the store doesn't remember it,
doesn't retry it or read a copy, and a picture without memory for its
effects' channels is read alone, else refused with the reason, as for a
file too large; never the image crate's decode of the whole file. The
planes held, copied into a request's new set, are reserved fallibly too;
a copy for the disk cache that can't be had is not written. The tests
that set the process-wide disk cache take turns, and a test shows that
an unreadable request stays refused when the decode limit changes.
2026-11-16 23:35:44 +01:00
vangel76 f08a2290c2 OpenEXR limit: the planes a request ends with count together, held and from the disk cache too (#3749)
The limit applied to the planes read from the file just now, so planes held in
memory or handed back by the disk cache were never counted and repeated
requests could pass it. It now applies to what the request ends with, asked
before anything is read or allocated. When the held planes don't fit with the
missing ones they are let go and the request is read on its own; only a
request that doesn't fit alone is refused.
2026-11-16 23:11:26 +01:00
vangel76 a7e7cb6b6a OpenEXR limit: only a refusal for size depends on the limit's generation; sequentially consistent counters (#3749 review) 2026-11-16 23:10:29 +01:00
vangel76 bdf500f4ba OpenEXR limit: a refusal for size is decided anew when the limit changes; a failed reading is asked twice (#3749 review) 2026-11-16 23:09:59 +01:00
vangel76 b591f95ce4 OpenEXR: a file's channels may take what the machine's RAM allows, not a fixed 8 GiB (#3749)
The limit of one read is decided here (the 8 GiB floor, or the memory
budget when that is more) and handed to craft-exr as the read's max_bytes,
so the planes it reserves and the largest block it decompresses follow it.
craft-exr reserves the planes fallibly, so the reader's own fallible plane
allocation is not needed: memory the system can't give is a read that
failed just now (read_error), a plane over the limit a refusal for size.
2026-11-16 23:09:47 +01:00
vangel76 ea4ac2e5ed Refuse OpenEXR blocks larger than their chunk can hold
craft-exr now holds every block to what its chunk's bytes can
decompress to, using the ratio the index was built with: the index is
built with craft-exr's default ratio (no longer unbounded), a file with
such a block is refused when indexed and by safe_to_decode before the
image crate sees it, and each read checks it again. Memory the system
refuses during a read is a read that failed just now, not remembered,
instead of a file too large.
2026-11-16 23:07:23 +01:00
vangel76 c1aea6d65a Bound OpenEXR pictures in memory and keep effects' depth fill
The picture of an OpenEXR file (also the black one of a file without a
main picture) is checked before anything is read: no larger than a
still may decode to and than one read of the file may allocate, else
refused with the reason instead of allocated. Import refuses such a
picture as other stills. The fill of a picture channel in a part named
like depth is the background depth again, as effects read it; the
picture is 0 outside that part's data window when composed instead.
A read decompresses on at most four threads of its own, none for a
small file.
2026-11-16 22:59:38 +01:00
vangel76 d671f6e45b Show an OpenEXR file's main picture as After Effects does
The picture of an OpenEXR file is now craft-exr's main picture: the
unprefixed R, G, B (and A) of the first part that has all three, else
the first part's unprefixed Y (and A) as grey. A file of named layers
only (Blender's ViewLayer.Combined.R, a compositor's Image.R) has none
and shows black at its display window's size, opaque, as After Effects
shows it; it still imports and its layers stay readable by effects
(EXtractoR, Cryptomatte, the 3D Channel effects).

The own picture choice (the beauty pass by name, data passes, the first
channel as grey) is gone. The picture's channels keep a fill of 0
outside the data window even in a part named like depth, so the picture
of such a file stays what the image crate shows. The golden hashes of
the named-layer files are re-pinned (their channels are unchanged), and
the docs describe the new rule.
2026-11-16 22:30:46 +01:00