Similarly as with other properties, we should aim to make this into user config option.
For the time being remove, as our outdated default user-agent seems to be breaking
few HLS streaming services.
Do not assume that HW can do every format that GLUpload/GLColorConvert can.
Check what video formats GL context actually supports and advertise that,
so a format that HW cannot do will never be picked up.
Create a GL context handler subclass of recently added context handler base class and
rework code to use it. This simplifies a lot of things, like: switching of importers
at runtime, handling context queries, sharing data between importers, etc.
In order to bridge work with GL between GTK and GStreamer reliably, we need to create and
check context capabilities ahead of time. We also need to be able to query context at any
time during playback. Also we should avoid changing importers on context queries, this can
lead to all sort of problems.
To be able to achieve all of above, introduce new context handler base class that works
separately of importers.
When we prefer GLES, make sure its no lower then 3.1 version, so it can color convert 10bit videos,
otherwise core GL 3.2 that GTK4 normally defaults to takes precedence.
With auto behavior we should select the best one, both performance and support wise.
Otherwise user can request one or another with GST_GL_API env that works best
with his specific HW.
Unlike legacy sink that used GLArea, GdkPaintable snapshots without pre-multiplying
sizes with scale factor. Remove scaling multiply of our widget dimensions,
so both video and widget width/height will be calculated correctly.
This fixes both image and black borders displaying issues on displays with
non-100% scaling applied.
GStreamer does not take pixel aspect ratio into account for overlay rectangles.
If we do so, we end up with subtitles that are displayed off-center.
Behave like any other GStreamer sink on this matter.
Instead figuring out what kind of pool we should make just use the upstream provided
functionality that comes as part of GLUpload API.
Fixes uploads with GstVideoGLTextureUploadMeta, which legacy VAAPI plugin uses on GLX.
This tries to avoid "scrambled" image on Linux with Intel GPUs that are mostly used
together with x86 CPUs at the expense of using slightly slower non-direct DMABuf import,
by not going with GLES by default on EGL.
The "GST_GL_API" env can still be used to force one or another context type.
We need for GStreamer to be aware of currently used DRM modifier first. On the other
hand forcing GL everywhere would break most embedded systems that only work with GLES,
so do it for x86 processors only... and pray for it to work.
See: https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/1236
There were various problems with importer loader code. One defect was that
it kept managing a single global list of available importers and marking
which one is currently used on it. This made it not work correctly for
multiple sink instances in single process and was not thread safe.
This commit changes importer loader code into a GstObject subclass, which
keeps its own list of importers per instance and unlike before makes it
possible to free this data from memory when destroyed.
Now only open modules are kept always in memory until process finishes
since we cannot unload them once loaded anyway.
We do not support zero width/height (described in our caps). Set initial video
dimensions to 1x1. They are going to be updated on first buffer after video info is set.
This avoids running into https://gitlab.freedesktop.org/gstreamer/gstreamer/-/issues/1139
and makes sure we do not do any divisions by zero here either.
Add support for premultiplied alpha flag detection when creating GdkMemoryTextures.
Also check if overlays have premultiplied alpha and if they do set this flag for
their video info, so it will be detected when creating a GdkMemoryTexture out of them.
Fixes performance issues with karaoke subtitles, as now GTK does not have to perform
unnecessary multiplication.
Make GL base importer class and all its subclasses that access parent props,
use a different mutex lock then the main importer class.
This way main importer class can continue processing buffers without waiting for
GL-based subclasses, as most GL operations (e.g. upload) are time consuming.
We cannot compile app as a whole on MS Windows yet, but with those changes it is now possible to compile and run our new video sink alone on Windows OS
We do not want to build GL-based importers when GTK GL windowing support is missing,
which might happen because of missing deps or on unsupported platform
Too many problems with VAAPI decodebin, meanwhile VA works fine and there is already
a pending MR on GStreamer to enable it by default, so we do that a little early.
We also do promise working HW accel out of box, so enable all common NVDEC elements too.
They can always be disabled back in player preferences window if any problems do happen.
We can now remove the alternations of v4l2codecs as those are PRIMARY+1 by default and we
now require GStreamer 1.20+ so no need for version checking there too.
Fixes#236
Install plugin lib stuff into a dir named "clapper-1.0" instead of "clapper-0.0". Even through our API is still very unstable, we cannot go down with versioning. The sink importers being installed there are not part of public API anyway.
No need to auto build rawimporter if we are building gluploader as it
will be always loaded on startup, but never used.
Skip the build of it when its meson option is set to "auto".
Add "clappergluploader" module which uses "glupload" and "glcolorconvert" internally,
allowing either uploading RAW system memory into GPU with GL colorspace conversion or
importing DMABufs into GL memory and making a GdkTexture out of them at one go.
Add new GStreamer plugin with custom video sink for Clapper video player.
The main difference is that unlike the old one, this does not operate using `GtkGLArea`
anymore, but processes and displays `GdkTextures` directly through `GtkPicture` widget.
Also this one is installed like any other GStreamer plugin, thus can be used via
`gst-launch-1.0` binary or even by other GTK4 apps if they wish to integrate it.
In order to not depend on GL stuff at build time, this plugin uses seperate GModules
called "importers" in order to import different kind of memories into GdkTexture. This
allows expanding its capabilities further then we were able to do before.
Some mobile distros seem to be under assumption that playbin3 is required to
get HW accel and setting this env by default is good idea. Both are wrong.
Check this env presence to avoid problems as factory will always return
playbin3 if set (even if playbin2 was requested).
Someone might be holding a copied media info in which case tag
list ref count is not 1. Use gst_tag_list_make_writable to
assure that ref count is 1, so we can insert more tags into the list.
There is no guarantee that received later tags also contain values from
earlier ones, as they might come from different element.
Combine them instead while replacing old values with newer ones.
The video/audio decoder changed signal was not working correctly in case of
multiple streams with multiple decoders in single file.
We need to listen to the current-(video/audio) signal, when it changes find
corresponding "input-selector", get stream ID from its active pad and then
find the decoder in the pipeline that handles this stream ID. Similarly for
playbin3, but use stream ID from the "streams-selected" signal.
We change few default plugin ranks during init to whitelist them, but we do not update
their values from GST_PLUGIN_FEATURE_RANK env afterwards, so do that.
The ENV should be preferred over default config.
