Commit Graph
39 Commits
Author SHA1 Message Date
itsmattkc e0bb31b199 remove shifts, operations, and make caches node-agnostic 2022-05-16 19:53:38 -07:00
itsmattkc 9ba3f24f3b cache: fixed double write corruption 2022-05-16 16:10:54 -07:00
itsmattkc 23973699b4 render: improved overall flow 2022-05-16 15:41:50 -07:00
itsmattkc bfd6a407a5 Merge branch 'master' into hashremoval 2022-05-16 09:12:49 -07:00
itsmattkc 4b366a9b73 update copyright year to 2022 2022-05-10 10:38:56 -07:00
itsmattkc 4bc82ca6df removed hashes from previewautocacher 2022-05-09 19:40:58 -07:00
itsmattkc 2cf0b24e0f framehashcache: refactored to remove hashes 2022-05-09 19:04:59 -07:00
itsmattkc 006d42ecf1 framehashcache: use two relationally linked maps for optimization 2021-10-09 14:09:58 -07:00
itsmattkc 4bbc75392a nodes: rewrote and simplified length signaling system 2021-07-15 21:00:18 -07:00
itsmattkc c374a03e9d render: further simplified job times
Outside of silly mistakes, this should be significantly faster and more stable.
2021-07-12 15:14:37 -07:00
itsmattkc 310fd3541a various: greatly simplified job times
Hopefully this doesn't break everything
2021-07-11 23:33:21 -07:00
itsmattkc 8a58ee2bdb renderer: use iterator rather than creating vectors
Creating vectors from the FrameHashCache was one of the slowest functions in the program, this should be significantly faster.
2021-07-11 16:47:12 -07:00
itsmattkc 38c38daf8a framehashcache: use timestamp based vector internally
I think this will be significantly faster, but I guess we'll find out!
2021-07-10 15:37:06 -07:00
itsmattkc 33f71d7548 framehashcache: use static const string rather than function
Should be mildly faster
2021-06-16 14:29:43 -07:00
itsmattkc b07e132dbb framehashcache: allow access to frames being saved
Minor optimization and crash fix. If a frame is currently being
saved, it's still technically available in memory and can be
returned faster than loading from EXR. This also prevents
OpenEXR from trying to load a frame that we're midway through
saving.
2021-04-24 18:21:33 +10:00
itsmattkc 77d4b18a60 updated copyright year for 2021 [skip ci] 2021-04-17 18:56:50 +10:00
itsmattkc 6c30f8cf46 nodes: implemented infrastructure for caching any node
Not accessible through UI yet, but ported Footage node to it
(previously we used a hack to ensure Footage textures were
cached, now it's a formal infrastructure).
2021-04-11 18:56:39 +10:00
itsmattkc 75d4cd899b use hardcoded namespace
The macro defined namespaces confused the hell out of lupdate and more or less broke translations permanently. Looks like the only way we can do it is to have a hardcoded namespace, which goes against my instinct, but honestly how likely is it that we'll change the namespace anyway (I guess forks might want to do it, but that's their problem ;) )
2020-11-17 20:24:42 +11:00
itsmattkc d241090a9f finally updated the copyright year 2020-11-17 20:13:22 +11:00
itsmattkc 0d0766e4fb improved render channel count system 2020-11-15 22:16:40 +11:00
itsmattkc cabcd49b25 renderer: copy cache path before render to avoid race conditions 2020-09-05 18:48:54 +10:00
itsmattkc f6760386ce renderer: improved handling of pixel aspect ratio and footage cache 2020-08-21 14:43:13 +10:00
itsmattkc d85089f71a cache: improved setting manual cache path on a per project basis 2020-08-07 13:41:09 +10:00
itsmattkc df14c2aaa7 cache: better implemented changing disk cache folders
Not complete yet, but getting there.
2020-08-07 02:31:46 +10:00
itsmattkc fa02430d48 cache: reimplemented disk manager to enforce limit in preferences 2020-08-06 01:30:51 +10:00
itsmattkc c8def73b95 cache: mild auto-cache overhaul
Shifted from CacheTask to functionality built into RenderBackend. It was
a lot easier to control behavior this way without having to juggle a ton
of threads and race conditions.

Could likely be multithreaded further.
2020-08-05 02:19:04 +10:00
itsmattkc 36dd06d2dc viewer: set video and audio caches to children of the viewer
Ensures their thread gets changed along with the viewer. The project
gets loaded/created in a background thread so the GUI can remain
responsive, and is then moved to the main thread for intended event
handling. However if the caches aren't parented, the hierarchy breaks
and the caches remain in a thread whose event loop is quickly destroyed.
Now that they're parented, events can be properly queued on them once
again.
2020-07-10 18:34:16 +10:00
itsmattkc cbbf1c9149 framehashcache: use correct linesize when saving EXR
Fixes stride issues on non-standard frame sizes.
2020-06-16 15:38:23 +10:00
itsmattkc 2e3072623a renderer: use qvectors instead of qlists
Should be much more efficient for large-scale projects.
2020-06-16 02:45:16 +10:00
itsmattkc a9a13c4b0c framehashcache: remove code for saving integer-based frames to cache
Since we don't allow non-float cached frames now, there's no need to have
this code complicating the codebase.
2020-06-10 14:12:26 +10:00
itsmattkc 52ca18d5e1 viewer: moved cached frame loading code so that more classes have access to it 2020-06-09 02:35:13 +10:00
itsmattkc 430561e2ac sequence: made divider a sequence parameter rather than a viewer parameter
Doing this makes the preview resolution an explicit project setting rather
than a temporary UI setting. This will allow more consistency when caching.
2020-06-08 14:41:35 +10:00
itsmattkc 9ff9817b95 cache: use render job times to synchronize cache
Ensure old inaccurate jobs are never prioritized over current jobs.
2020-06-04 16:48:23 +10:00
itsmattkc e30d278327 playbackcache: made thread-safe
Since this class is regularly accessed by multiple threads, it makes the
most sense to make all of its functions thread-safe.
2020-06-03 03:38:28 +10:00
itsmattkc 11bfe61208 cache: began base implementation of "shifting" the cache
Should be a fairly important optimization to make.
2020-06-02 01:38:12 +10:00
itsmattkc 89ae6e026f cache: reimplemented auto-caching
Implemented a smarter auto-cache that's context sensitive. Caching will
automatically pause when playback begins or values are being changed and resume
when inactive.
2020-05-24 15:20:35 +10:00
itsmattkc 34fade9d81 cache: attempt to automatically re-match hashes on invalidate
This code is slow and unoptimized. Will require some more work.
2020-05-21 16:49:02 +10:00
itsmattkc 53a2956350 cache: added cache task that can be manually invoked 2020-05-15 18:36:02 +10:00
itsmattkc 863ae1b913 a first attempt at a new render system 2020-05-14 14:43:03 +10:00