If an item's show/hide transition was active the underlying source
would not previously not be enumerated, resulting in reference leaks by
the UI or other components waiting for the source to be deactivated to
release it.
Removes the internal helper `remove_without_release()` and branches
`obs_sceneitem_remove()` into public and internal versions.
The `obs_sceneitem_set_transition()` calls were missing from the
`remove_without_release()` function, so this resolves that issue
implicitly. Code cleanup mentioned above is done to hopefully avoid
these mistakes again.
Before the commit 763dddbbaf, hotkeys to show and hide scene items are
distinguished by source name instead of ID. When migrating from the
legacy data structure, the pointer to the data was not released.
An already-removed item has a NULL `item->parent`, meaning that calling
`full_lock(scene)` results in undefined behavior. This makes the method
return earlier if the specified item is removed instead of attempting
to lock the scene.
No thread safety is changed, because it wasn't thread-safe to begin
with.
This fixes flipping behaviour when positional alignement inside the
bounding box is other than centered, bringing it in line with the
behaviour when centered, i.e. flip the item in-place in the bounding
box.
To avoid passing `struct darray *` type, which cannot hold the type
information, this commit defines array types and uses that types on the
function parameters.
When creating a group, a newly allocated group `sub_scene` is released
just after calling `obs_scene_add_internal`.
If another thread released the scene-item, which is the sub_scene,
use-after-free might happen.
When the `pos` variable is non-zero, audio does not get mixed correctly.
This is due to the fact that the `pos` variable was erroneously being
applied to the input rather than the output.
When a variable `pos` became larger than `AUDIO_OUTPUT_FRAMES`, `count`
will get overflowed number. To avoid the overflow, continue the loop
when `AUDIO_OUTPUT_FRAMES - pos` is not positive.
Previously, `custom_size` was checked at the end of the `scene_load`
function. If the scene contained no "items" array, the `custom_scene`
loading code would never be run.
This moves the `custom_size` code above the return statement.
Fixes studio mode preview on SDR monitor for HDR canvas rendering SDR
source into SDR swap chain. Needs to render SDR source into HDR render
target, and then tonemap into SDR swap chain for preview.
Due to the way the frontend works, the logic for inserting the row for
the group in the UI was dropped since the new item_add message will
insert one automatically since it forces a refresh of the scene items.
The reliance on that implicit row insertion when grouping items is a
bit of a code smell, although the alternative would be to add logic to
disable handling the signal for that instant, which is probably a
worse choice.
Rec. 2020 is really an SDR spec, but I think HDR10 made it okay to slap
PQ on it, call it an HDR spec. Rec. 2100 came along after and formally
allowed the use of PQ/HLG, so we should use 2100 instead.
This provides the framework for automatically compositing SDR and HDR
sources together. Source will need to leverage the new
video_get_color_space to opt into HDR support.
