Before this the the whole sequencer cache would be invalidated
when hanging sequence settings.
This was completely annoying because changing color balance
settings would re-load image file for which color balance
is happening on every change,
In fact it's still an issue if color balance is changing for
image strip itself, but if this strip has got effect and
color balance is changing for it file wouldn't be reloaded.
Replace pseudo-LRU approach of determining which buffer
to remove when running out of space allowed for cache
with approach which would remove the frame which is most
far away from newly added frame.
This is still a bit tricky because it's impossible to
distinguish which frame to delete in situation of:
CCCC...CC
^
it's either user wants to extend left segment of cached
frames and buffers from right segment should be removed
or he wants to join this two segments and in that case
buffers from right segment should be removed.
Would need a bit more investigation which situation
is more common in general usecase.
Additional changes:
- Cleanup some memutil files (which are familiar to cache limiter)
- Add option to make moviecache verbose. If DEBUG_MESSAGES is
defined in moviecache.c detailed logs would be printed to the
console.
- Movie caches are now named which helps reading debug messages.
Fixes Fix for [#25713] VSE shows and renders wrong straight alpha gradient even after convert to pr
(see revision: 34540, fix by Janne)
By not breaking the seqcache interface API.
Added comments to header file, so that it is easier to understand, how the
cache API is supposed to work.
- use NULL rather then 0 where possible (makes code & function calls more readable IMHO).
- set static variables and functions (exposed some unused vars/funcs).
- use func(void) rather then func() for definitions.
* Caching of the start and end stills were just referencing the original imbuf (which got premultiplied after the caching), so as a result most of the time the premul was applied twice.
* Now the start and end stills are stored in the cache as duplicates of the original (non modified) imbuf.
* documented and rewrote the render interface of the sequencer.
(now, the geometry / render_type / etc. settings are stored within a
seperate structure called SeqRenderData that is passed within the code.)
* that fixes
* cache problems, since the caching system didn't keep track of
proxy files vs. final renders.
* is a necessary step, to bring back frame blending in speed effect
(the SeqRenderData structure elements are already there)
* will make motion blur render options available within the sequencer!
* this patch also fixes:
* "easy retiming" using speed effects. (in Blender 2.49, you could
add a speed effect and resize the source track to retime it to that
length)
* adds labels for the Original dimensions for Image + Movie tracks
(worked in 2.49, too)
This patch cleans up the sequencer core by replacing the caching system
(TStripElems) with a hash based system, which is:
a) a lot faster
b) a lot more readable
c) a lot more memory conserving
The new caching system is also a good building ground for
a) sub frame precision rendering (even on scene strips)
b) multi core rendering (threaded rendering is still disabled, but can
be extended now to arbitrary core numbers)
I tested the code on an extensive editing session today and had no
crashes during 4 hours of editing. So I consider it very stable.
