Similar to what qtdemux does in `gst_qtdemux_stream_update_segment()`,
in which case we want to resend tags according to our position.
By setting the changed flag for all stream tags we ensure they will all
be (re-)send as we progrees.
This notably ensures that rotation tags are not lost when video players
finish playback and jump back to the beginning of a video.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8429>
Colorimetry bt2100-pq is a variant of colorspace V4L2_COLORSPACE_BT2020
where the transfer function is SMPTE 2084 which is used by HDR content.
Colorimetry 1:4:5:3 is a full-range variant of colorspace
V4L2_COLORSPACE_470_SYSTEM_BG
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9166>
When there are gaps between segments in the DASH manifest (e.g. it's a
recording of a live stream during which the camera was turned on and off
a couple times), seeking to a timestamp that belongs into a gap leads to
repeat_index calculation yielding a nonsensical negative number (because
ts < segment->start).
In such event set the playback to continue at the first repetition of
the segment that follows after the gap.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8620>
When producing the profile_teir_level() struct, account for emulation
prevention bytes (0x0, 0x0, 0x3) in the source SPS. Also copy from the correct
starting point in the source SPS and don't put the NAL header into codec_data.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9205>
When creating a faststart file, qtmux has to take into account that
any 32 bit stco atom may need to be converted into its 64 bit version
due to the offset that is added before the mdat data by the moov atom,
the extra atoms and the mdat header. If a stco atom is converted into
co64, the chunk offset of all stco and co64 atoms has to be increased,
which in turn may also cause that other stco atoms need to be converted
into co64, modifying the chunk offsets again.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9102>
Certain encoder user controls as well as codec controls can be
changed while the pipeline is in playing state.
e.g. Codec controls like video_bitrate, frame_ltr_index,
use_ltr_frames, etc. and user controls like horizontal_flip,
vertical_flip can be changed while the pipeline is running.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9168>
Use gst_pad_push_event() instead of manually getting peer pad and calling
gst_pad_send_event() otherwise the pad probe on the sinkpad wouldn't get the
events.
Also ensure navigation events are posted to the bus even when gst_pad_push_event()
fails in the render function, so navigation remains functional regardless
of rendering issues.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9130>
If the driver has size alignment requirement, suggest the
difference between aligned size required by diver and the actual
size of the frame as padding requirement to the upstream element.
This will ensure that the buffer size allocation is as per the
driver requirement.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/8971>
At the moment we are always accepting the input events to forward into
GStreamer infrastructure. This works but we might have other uses for
such events elsewhere in the QtQuick scene so allow opting out to this
behaviour.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9085>
There is a small window during stopping when tasks have been add
to the transfer_requests queue, but never transfered to the
active_transfers array and causes the fallowing error:
GTask 0x7f2c4400e930 (source object: (nil), source tag: (nil)) finalized
without ever returning (using g_task_return_*()).
This potentially indicates a bug in the program.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9060>
In the rare case when we get EOS on all pads before we get any buffer we
would finish the file without writing headers, i.e. the file would be
corrupt. This patch makes sure that a header is always written.
Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/9061>
