Commit graph

19575 commits

Author SHA1 Message Date
Matthew Waters
557ca6fda5 glimagesink: request a resize on caps/3d mode changes
Fixes incorrect aspect ratio on OSX when changing caps or the 3d
output mode.

https://bugzilla.gnome.org/show_bug.cgi?id=755111
2015-09-17 20:26:23 +10:00
Matthew Waters
3addb6bbf4 glwindow: add API to request a resize event on the next draw
- glimagesink needs to be able to resize the viewport on aspect ratio
  changes resulting from either caps changes or 3d output mode changes.
- Performing a glViewport outside the GstGLWindow::resize callback
  will not have the winsys' stack of viewports required to correctly
  place the output frame.

Provide a function to request a resize on the next draw event from the
winsys.

Also track size changes inside the base GstGLWindow class rather
than in each subclass.

https://bugzilla.gnome.org/show_bug.cgi?id=755111
2015-09-17 20:26:23 +10:00
Matthew Waters
d088a5294b gl/uploadelement: fail earlier if we could not upload the buffer 2015-09-17 20:21:43 +10:00
Matthew Waters
1625f74632 gl/x11: store the correct dimension from the resize events
small typo s/width/height/
2015-09-17 20:21:43 +10:00
Vineeth T M
e98addcc31 avdtputil: fix GError memory leak during failures
https://bugzilla.gnome.org/show_bug.cgi?id=755142
2015-09-17 11:55:49 +02:00
Vineeth T M
00438fdff0 gl: Fix GError leaks during failures
https://bugzilla.gnome.org/show_bug.cgi?id=755140
2015-09-17 11:52:55 +02:00
Tim-Philipp Müller
0cb160a685 intervideosink: render preroll frames as well
Derive from GstVideoSink so that preroll frames will automatically
get rendered too, unless the show-preroll-frame property is set to
FALSE. Fixes intervideosrc only picking up frames if intervideosink
is in PLAYING state.

https://bugzilla.gnome.org/show_bug.cgi?id=755049
2015-09-16 17:20:52 +01:00
Jan Schmidt
86a8ff50c0 dvbsuboverlay: Check if downstream supports GstVideoOverlayComposition.
Fix the negotiation of GstVideoOverlayComposition by checking
intersection with the peer caps, rather than just accept-caps,
which might only check the pad template.

https://bugzilla.gnome.org/show_bug.cgi?id=755113
2015-09-17 00:49:58 +10:00
Vincent Penquerc'h
50400fa2a6 mpdparser: support for negative repeat count in segments
Implements negative repeat segment fields, defined in 5.3.9.6.1.
2015-09-16 09:44:20 +01:00
Vincent Penquerc'h
69bfa8222f mpdparser: properly read signed r values for S elements
The spec defines these as signed in 5.3.9.6.1.
Since we don't support this behavior, warn and default to 0
(non repeating), which is the spec's default when the value
is not present.

https://bugzilla.gnome.org/show_bug.cgi?id=752480
2015-09-16 09:44:19 +01:00
Vincent Penquerc'h
23ea8ccb43 mdpparser: catch failures to parse
https://bugzilla.gnome.org/show_bug.cgi?id=752480
2015-09-16 09:44:19 +01:00
Matthew Waters
7ae0dd2361 glimagesink: avoid updating the viewport in the draw loop
Fixes flashes/incorrect aspect ratio when resizing glimagesink on OSX.
2015-09-16 18:04:51 +10:00
Sebastian Dröge
5674dd75d9 adaptivedemux: Fix playback of live streams
dashdemux seeks each live stream to its current fragment in the beginning, but
the base class does not know about this. Update the demuxer segment with this
seek so we generate the correct SEGMENT event and can actually play the
stream.

This needs some refactoring at some point.

https://bugzilla.gnome.org/show_bug.cgi?id=755047
2015-09-15 22:05:19 +02:00
Thiago Santos
7e834a5d85 opusdec: remove check for number of channels
opus decoder can convert from different number of channels, no
need to check, just let it negotiate and create a new decoder if
needed.

https://bugzilla.gnome.org/show_bug.cgi?id=755059
2015-09-15 16:00:43 -03:00
Thiago Santos
0ed501ab03 adaptivedemux: Only overwrite last return if it was OK
Prevents overwriting other conditions that would be more important,
such as EOS.

https://bugzilla.gnome.org/show_bug.cgi?id=755042
2015-09-15 12:46:31 -03:00
Thiago Santos
bd710e5e1b dashdemux: fix return when trying to advance in subfragmented stream
Even if it doesn't actually advance the subfragment in the default way
for streams that have subfragments, it can help the base class to return
EOS when there is no more fragments instead of signaling it that it should
continue downloading.

https://bugzilla.gnome.org/show_bug.cgi?id=755042
2015-09-15 12:46:31 -03:00
Tim-Philipp Müller
fe8de1857a gl: bufferpool take into account video stride alignment requirements
when allocating memory. Fixes crashes with avdec_h265 in the AVX2
code path which requires 32-byte stride alignment, but the
GstAllocationParams only specified a 16-byte alignment.

https://bugzilla.gnome.org/show_bug.cgi?id=754120
2015-09-15 16:11:23 +01:00
Thibault Saunier
b6ad1b0c22 pitch: Set seqnum on newly created segment event
https://bugzilla.gnome.org/show_bug.cgi?id=755012
2015-09-15 13:41:18 +02:00
Sebastian Dröge
637106e287 audioaggregator: Fix mixup of running times and segment positions
We have to queue buffers based on their running time, not based on
the segment position.

Also return running time from GstAggregator::get_next_time() instead of
a segment position, as required by the API.

Also only update the segment position after we pushed a buffer, otherwise
we're going to push down a segment event with the next position already.

https://bugzilla.gnome.org/show_bug.cgi?id=753196
2015-09-14 19:57:00 +02:00
Sebastian Dröge
97fe89f351 audioaggregator: Use stream time in the position query instead of segment position
https://bugzilla.gnome.org/show_bug.cgi?id=753196
2015-09-14 19:56:51 +02:00
Sebastian Dröge
eb97bb0adb videoaggregator: Fix mixup of running times and segment positions
We have to queue buffers based on their running time, not based on
the segment position.

Also return running time from GstAggregator::get_next_time() instead of
a segment position, as required by the API.

Also only update the segment position after we pushed a buffer, otherwise
we're going to push down a segment event with the next position already.

https://bugzilla.gnome.org/show_bug.cgi?id=753196
2015-09-14 19:56:43 +02:00
Sebastian Dröge
336ca3207a aggregator: Document that get_next_time() should return running time
https://bugzilla.gnome.org/show_bug.cgi?id=753196
2015-09-14 19:56:33 +02:00
Sebastian Dröge
c9f60db2d4 mpdparser: Don't consider period start times in periods with segment lists either
https://bugzilla.gnome.org/show_bug.cgi?id=754222
2015-09-14 19:53:18 +02:00
Sebastian Dröge
d9c45e918f mpdparser: Fix unit test that assumed that fragment timestamps should include the period start timestamp
https://bugzilla.gnome.org/show_bug.cgi?id=754222
2015-09-14 19:53:18 +02:00
Sebastian Dröge
0d0455e346 dashdemux: Export the period start time to the base class
https://bugzilla.gnome.org/show_bug.cgi?id=754222
2015-09-14 19:53:03 +02:00
Sebastian Dröge
d1c5669a38 adaptivedemux: Properly implement timestamping of multi-period streams
Each period will start again with pts 0 + period presentation offset, which is
also going to be the presentation time inside the container stream if any.
However all periods together should form a continuous timeline, with regard to
stream time and running time.

For making this possible we keep track of the "user requested segment", i.e.
the seek events, inside the demuxer without adjusting anything and taking this
demuxer segment only as orientation for modified segments per stream.

This per stream segments will have their segment.start at pts that would be
produced for this stream in this period, and the segment.base/time adjusted so
that this pts maps to the running and stream time this period should have in
the context of all other periods.

https://bugzilla.gnome.org/show_bug.cgi?id=754222
2015-09-14 19:53:03 +02:00
Sebastian Dröge
b5697b8ced Revert "dashdemux: Subtract the period start time from the presentation offset"
This reverts commit 626a8f0a74.

This allows us to get the plain presentation offset and the period start time
separately. We have to adjust the timestamp by the presentation offset, but
the period start time should only adjust the stream time and running time.

https://bugzilla.gnome.org/show_bug.cgi?id=752409
2015-09-14 19:53:03 +02:00
Sebastian Dröge
bb05bbafd5 Revert "dashdemux: Include the period start in the fragment timestamps in all cases"
This reverts commit e671ad25a9.

The timestamps should restart at 0 again for each period, but we have to
adjust the segment to map those timestamps to the actual stream time and
running time of that period.

Otherwise we would have timestamps that conflict with the ones from the tfdt
inside the MP4 container, which are restarting at 0 for each period.

https://bugzilla.gnome.org/show_bug.cgi?id=752409
2015-09-14 19:53:03 +02:00
Sebastian Dröge
cfe2871a5e mpdparser: Consider the repeat count when checking if a segment is the last one
Otherwise we play only the first repetition of the last segment and then EOS.

https://bugzilla.gnome.org/show_bug.cgi?id=752480
2015-09-14 19:49:14 +02:00
Thiago Santos
cc599bb5b5 dashdemux: also check for subfragments on has_next_fragment
In dash isombff profile the fragment is split into subframents where
bitrate switching is possible. Also take that into consideration
when checking if a stream has next fragments.
2015-09-14 13:56:10 -03:00
Miguel París Díaz
25f0787c1c opusenc: improve deprecated properties docs
https://bugzilla.gnome.org/show_bug.cgi?id=754819
2015-09-11 23:33:19 +02:00
Miguel París Díaz
f000cd4963 opusenc: do not throw g_warning when getting deprecated properties
https://bugzilla.gnome.org/show_bug.cgi?id=754819
2015-09-11 23:33:19 +02:00
Thiago Santos
ba97ec74be videoaggregator: fix caps query to properly handle alpha formats
Only accept alpha if downstream has alpha as well. It could
theoretically accept alpha unconditionally if blending is
properly implemented for handle it but at the moment this
is a missing feature.

Improves the caps query by also comparing with the template
caps to filter by what the subclass supports.

https://bugzilla.gnome.org/show_bug.cgi?id=754465
2015-09-11 18:03:26 -03:00
Jan Schmidt
9f890c5935 vtdec: Use R/RG textures for NV12 when the context provides them
Fixes playback to GL memory on iOS, where the colours are messed
up by passing Luminance/LuminanceAlpha textures where
color convert expects R/RG textures.

https://bugzilla.gnome.org/show_bug.cgi?id=754504
2015-09-12 04:33:34 +10:00
Jan Schmidt
0d7d7f6c92 vtdec: Prefer GL output caps.
Prefer GL caps by putting them first in the pad template and
intersecting using GST_CAPS_INTERSECT_FIRST.
2015-09-12 04:33:34 +10:00
Sreerenj Balachandran
fcd42b7d77 codecparsers: h265: Fix the selection of Active Ref Pic Set
If short_term_ref_pic_set_sps_flag is FALSE, the ShortTermRefPicSet
structure is supposed to derive from slice header. Which means,
CurrRpsIdx is equal to num_short_term_ref_pic_sets. But the number
of refpicsets communicated via sps header is only num_short_term_ref_pic_sets - 1.
And we are using slice_header structure to reference the last entry, which is
ShortTermRefPicSet[num_short_term_ref_pic_sets].

https://bugzilla.gnome.org/show_bug.cgi?id=754834
2015-09-10 21:12:30 +02:00
Sreerenj Balachandran
81435c8891 codecparsers: h265 : Fix default scaling list values
Fix the default_scaling_list values based on Table 7-6

https://bugzilla.gnome.org/show_bug.cgi?id=754834
2015-09-10 21:10:17 +02:00
Sebastian Dröge
b2149e7bb4 decklinkvideosink: less-than-zero comparison of an unsigned value is never true
Use the correct type, GstClockTimeDiff, instead.

CID 1323742
2015-09-10 14:11:58 +03:00
Matthew Waters
38d360675c gtk, qt: more specifically define the compile time requirements
Otherwise we could include headers/configurations that will
never been installed.

https://bugzilla.gnome.org/show_bug.cgi?id=754732
2015-09-10 13:11:06 +10:00
Matthew Waters
85745d2f63 build/qt: indent configure file
don't mix tabs and spaces

https://bugzilla.gnome.org/show_bug.cgi?id=754466
2015-09-10 00:34:06 +10:00
Matthew Waters
bca1fd1040 qt: use our function table instead of directly calling gl functions
Otherwise when building with --as-needed we would need to link to
a GL or GLES library.

https://bugzilla.gnome.org/show_bug.cgi?id=754732
2015-09-10 00:33:40 +10:00
Matthew Waters
591494ff54 gl/api: use public rather than private define
HAVE_IOS is only defined for the build of this module so
attempting to use gstgl in iOS would result in incorrect GL
includes.
Use GST_GL_HAVE_PLATFORM_EAGL instead for choosing the iOS GL
header.
2015-09-10 00:33:19 +10:00
Sebastian Dröge
b6498c32f3 decklinkvideosink: Handle pipelines where the running time does not start around 0 properly
We were converting all times to our internal running times, that is the time
the sink itself spent in PLAYING already. But forgot to do that for the
running time calculated from the buffer timestamps. As such, all buffers were
scheduled much later if the pipeline's running time did not start at 0.

This happens for example if a base time is explicitly set on the pipeline.

https://bugzilla.gnome.org/show_bug.cgi?id=754528
2015-09-09 16:59:02 +03:00
Sebastian Dröge
a6744a915b decklinkvideosink: Consider pipeline latency, render delay and ts offset when scheduling frames
Without this, we will schedule all frames too late in live pipelines.

https://bugzilla.gnome.org/show_bug.cgi?id=754666
2015-09-09 16:58:45 +03:00
Kouhei Sutou
21bfa428d1 directsoundsrc: fix build error for 64bit Windows build by MinGW
Casting to UINT from HMIXER generates the following warning with
64bit Windows target MinGW:

    gstdirectsoundsrc.c: In function 'gst_directsound_src_mixer_find':
    gstdirectsoundsrc.c:733:30: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
         mmres = mixerGetDevCaps ((UINT) dsoundsrc->mixer,
                                  ^
    cc1: all warnings being treated as errors

We can use portable GPOINTER_TO_UINT() macro for this propose.

https://bugzilla.gnome.org/show_bug.cgi?id=754756
2015-09-09 16:57:08 +03:00
Sebastian Dröge
bab55828ae gl/eagl: Also unbind renderbuffer after setting up framebuffer
https://bugzilla.gnome.org/show_bug.cgi?id=754757
2015-09-09 10:10:47 +03:00
Sebastian Dröge
c97144b23b gl/eagl: Don't make context uncurrent just to make it current again in the next line
Also binding the framebuffer again is unnecessary then as it was just bound a
few lines before while the context was current.

https://bugzilla.gnome.org/show_bug.cgi?id=754757
2015-09-09 10:10:46 +03:00
Vincent Penquerc'h
6d1eda9391 mpdparser: fix period start time scaling
This GstStreamPeriod start value is expressed in nanoseconds,
and the glib time addition function expects microseconds.
There seems to have been a confusion with GstPeriodNode's start
field, which is expressed in milliseconds.

Additionally, add a warning if the timestamp modification did
not succeed, and NULL was returned.
2015-09-07 12:52:03 +01:00
Jan Schmidt
779ad060a4 faac: Add bitrate info to the tags output.
Makes it possible for muxers to know the target bitrate as soon
as encoding starts, which flvmux now uses.
2015-09-05 23:42:50 +10:00
Nirbheek Chauhan
cf786a9399 compositor: Ensure all arguments to CLAMP are signed int
If any of the arguments to CLAMP are unsigned integers, the comparison causes
an automatic conversion of the signed int to unsigned, which causes -1 to become
UINT_MAX and get clamped to the high value of the CLAMP instead of 0.

See 716 at http://c0x.coding-guidelines.com/6.3.1.8.html

Also add a test for this.

https://bugzilla.gnome.org/show_bug.cgi?id=754576
2015-09-04 23:32:36 +02:00