When a window with an OBSQTDisplay is closed, the surface can be
destroyed before the display is, opening a gap for OBS to attempt to
draw to the invalid surface.
This deletes the underlying OBSDisplay object when the actual surface
is destroyed, closing that gap.
Note: This appears to have been an issue previously with Wayland, as
hinted by the existing ifdefs.
obs-websocket's `GetInputDefaultSettings` request uses
obs_get_source_defaults() in order to fetch a source id's default
settings. This moves the default settings to the `.get_defaults`
callback so that our request does not return an empty object.
Does not affect existing functionality
Currently, ifdefs are used to determine if monitoring is supported.
This is difficult to maintain and restricts plugins from knowing if
monitoring is supported by OBS. This adds a runtime function to fix
that issue.
Previously, calling `obs_source_remove()` on an input would not fully
remove it if there was a scene item of the input on the current scene.
Now, when calling `obs_source_remove()` and there is a scene item of
the removed input on the current scene, the UI refreshes the scene
items, allowing the scene item to be released completely and the
input destroyed.
Add support for importing the Screen Capture source type from Streamlabs
scene collections. Since the Screen Capture source type is essentially
an abstraction layer for Display Capture, Game Capture, and Window
Capture, this adds some translation logic to translate Screen Capture
sources into one of those source types.
Adds support for using shared textures that were made with the
D3D11_RESOURCE_MISC_SHARED_NTHANDLE flag.
This increases the minimum required Windows version to Windows 8 or the
Platform Update for Windows 7. As official support is only for Win 8+
this does not change official support.
Move the python linkage after the obs target is defined so that we can
also add linker flags to preserve the unused python linkage. This is
required before GCC 11.
As Intel continues to add discrete graphics cards, maintaining a long
list of discrete device IDs to check against isn't wise. When i+I
platforms are detected, default to integrated graphics.
This fixes#2222 by ensuring python symbols are in the global symbol
table through a direct linkage to the executable. When python is
normally linked via frontend-tools.so its symbols are not global since
we dlopen frontend-tools.so. This causes issues for native python
modules which stopped directly linking python around 3.8 as the symbols
will not resolve.
This is pretty much a hack, but it minimizes the number of extra symbols
we add and hopefully isnt too terrible to maintain.
Instead of trying to open devices /dev/video<0-63> check /dev for
all /dev/videoX devices that are existent in the tree. This allows
finding devices with higher names than 63. E.g. /dev/video100 can
now be found and opened which previously failed.
_GNU_SOURCE #define is introduced for versionsort() which is a GNU
extension.
Fixes#4347.
Applies transformations correctly to copy/pasted sources when copy and
pasting multiple. Previously the transformation/crop state was stored in
a single variable on copy so it would only apply the transformations to
one of the selected sources on paste instead of what the state was when
copied.
Adds the following events:
- `OBS_FRONTEND_EVENT_SCENE_COLLECTION_CHANGING`
- `OBS_FRONTEND_EVENT_PROFILE_CHANGING`
Both emitted right before the change occurs. Can provide plugins with
better awareness of these critical operations happening, so that they
can react accordingly.
The commit 4a5ac3974 fixed the bug the first PTS is not zero. This
commit fixed the same issue in replay-buffer.
A variable `video_dts_offset` is renamed to `video_pts_offset` and the
variable is initialized by `pts` instead of `dts`.
Previously we would wait for pulse to attempt to read from the monitor
source and obs buffered at least 5ms of audio data before we tried to
fill the buffer. In some cases this resulted in consistently triggering
underruns in pulse.
Instead we try to fill the buffer immediately as obs outputs audio data
and while the pa buffer is not full. We also stop trying to grow the
buffer to prevent underruns after we reach 1s of latency.
The system tray assumed all outputs are inactive when it's initiated,
however outputs can be active if started via command line options. This
caused the wrong (only) text to be set if the output was in fact active.
The flatpak-builder tool now supports passing secrets options to the
build system. These options are not printed during the build, nor
added to the resolved manifest after build, so they don't leak env
vars from CI.
Make secret variables part of the Flatpak workflow environment, like
the main workflow. Pass the various services hashes and clientids to
the build system using the new "secret-opts" key.
When the filter view is created, the first item of the asyncFilters list
would be selected (if the source supports asyncFilters) and have its
properties shown.
This means that when no items were in the asyncFilters list, no
properties would be shown, even if an effect filter was present.
Adds a check to focus the effectFilters list in case there are only
effect filters, which makes the correct properties appear.
Previously glReadPixels was used to stage surfaces because Mac OS X at
that time supported some hardwares that didn't implement glGetTexImage.
However, OBS Studio now requires OpenGL 3.3 so that glGetTexImage on
macOS should work the same as on Linux.
The code using glReadPixels was introduced at 30af7dc. The minimum
requirements of the original OBS Studio for OSX was 10.8, which supports
hardwares with OpenGL 1.4. However, the function glGetTexImage requires
OpenGL >= 2.0.
Rodney
jpark37
tt2468
Ryan Foster
VodBox
gxalpha
Clayton Groeneveld
Kurt Kartaltepe
Guang Zhou
Christine Lin
jp9000
Florian Zwoch
Anthony Torres
Exeldro
David Plappert
Mike
lucha
Thomas Müller
Norihiro Kamae
zhaolong
Matt Gajownik
Georges Basile Stavracas Neto
Rafael Franco
Morten Bøgeskov