Mark Nauwelaerts [Thu, 6 Sep 2012 14:30:44 +0000 (16:30 +0200)]
deinterlace: do not leak getcaps result
Wim Taymans [Thu, 6 Sep 2012 14:23:28 +0000 (16:23 +0200)]
deinterlace: add support for bufferpool
Add bufferpool support to avoid a memcpy in the videosink when actively
interlacing.
Remove some commented obsolete code.
Wim Taymans [Thu, 6 Sep 2012 11:38:52 +0000 (13:38 +0200)]
deinterlace: proxy allocation query in passthrough
We can let the allocation query pass when we are operating in passthrough mode.
Wim Taymans [Thu, 6 Sep 2012 11:23:46 +0000 (13:23 +0200)]
deinterlace: use default event functions
instead of blindly forwarding unknown events.
Wim Taymans [Thu, 6 Sep 2012 11:23:30 +0000 (13:23 +0200)]
deinterlace: small cleanups
Wim Taymans [Thu, 6 Sep 2012 10:56:30 +0000 (12:56 +0200)]
deinterlace: call default query handlers
Call the default query handler instead of forwarding the query blindly. Fixes
issues of strides because of proxying the allocation query wrongly.
Wim Taymans [Thu, 6 Sep 2012 08:42:21 +0000 (10:42 +0200)]
v4l2: remove unused code.
Wim Taymans [Thu, 6 Sep 2012 08:42:06 +0000 (10:42 +0200)]
pulse: improve debug
Mark Nauwelaerts [Wed, 5 Sep 2012 09:50:05 +0000 (11:50 +0200)]
dvdemux: remove obsolete update newsegment handling code
Wim Taymans [Tue, 4 Sep 2012 10:35:53 +0000 (12:35 +0200)]
videobalance: avoid deadlock
_update_properties takes the object lock and should not be called when the
object lock is already taken.
Tim-Philipp Müller [Mon, 3 Sep 2012 11:46:03 +0000 (12:46 +0100)]
matroskamux: extract interlaced-ness of video track from interlace-mode field
instead of the old boolean "interlaced" field.
Tim-Philipp Müller [Mon, 3 Sep 2012 01:51:24 +0000 (02:51 +0100)]
video/x-xvid -> video/mpeg,mpegversion=4
Tim-Philipp Müller [Sun, 2 Sep 2012 01:50:50 +0000 (02:50 +0100)]
text/plain + text/x-pango-markup -> text/x-raw
Tim-Philipp Müller [Sun, 2 Sep 2012 00:31:53 +0000 (01:31 +0100)]
gst_message_new_duration -> gst_message_new_duration_changed
Wim Taymans [Thu, 30 Aug 2012 20:07:24 +0000 (22:07 +0200)]
session: also stop probatation on existing sources
Receiving an RTCP packet should also stop probation on sources we have seen
before.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=683065
Aleix Conchillo Flaque [Wed, 22 Aug 2012 23:36:21 +0000 (16:36 -0700)]
rtp: make rtp packet probation configurable (bug #682512)
Mark Nauwelaerts [Thu, 30 Aug 2012 10:21:01 +0000 (12:21 +0200)]
gdkpixbuf: adjust to modified video overlay composition API
Mark Nauwelaerts [Thu, 30 Aug 2012 09:30:01 +0000 (11:30 +0200)]
flacparse: fixup 0.11 port of suspect frame checking
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=682959
Mark Nauwelaerts [Tue, 28 Aug 2012 16:56:19 +0000 (18:56 +0200)]
avidemux: avoid invalid H264 bytestream codec_data
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=681369
Mark Nauwelaerts [Tue, 28 Aug 2012 17:00:44 +0000 (19:00 +0200)]
qtdemux: port segment event creation to 0.11
Mark Nauwelaerts [Tue, 28 Aug 2012 14:28:13 +0000 (16:28 +0200)]
qtdemux: release extra event ref when replacing pending newsegment event
David Corvoysier [Tue, 3 Jul 2012 15:50:24 +0000 (17:50 +0200)]
isomp4: add DASH tfdt box support
MPEG DASH has defined a set of new boxes to specify duration, indexes and
offsets of ISOBMFF fragments.
The Track Fragment Base Media Decode Time (tfdt) Box can in particular be
included inside a traf box to specify the absolute decode time, measured on the
media timeline, of the first sample in decode order in the track fragment.
This information can be used by the isomp4 demux to find out the current position of
an MP4 fragment in the timeline.
This patch adds code to isomp4 to:
- parse the tfdt box
- adjust the time/position member of the new segment sent when playback starts
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=677535
Tim-Philipp Müller [Sun, 26 Aug 2012 21:39:55 +0000 (22:39 +0100)]
docs: gst-launch -> gst-launch-1.0 and ffmpegcolorspace -> videoconvert
Tim-Philipp Müller [Sun, 26 Aug 2012 21:32:54 +0000 (22:32 +0100)]
docs: gst-launch-0.11 -> gst-launch-1.0
Tim-Philipp Müller [Sun, 26 Aug 2012 21:08:54 +0000 (22:08 +0100)]
deinterlace: the field in caps is "interlace-mode" not "interlace-method"
Fix deinterlace unit test. Need to set right field on output caps.
Also remove right field (not old 0.10 "interlaced" boolean field)
from caps in unit test before comparing old and new.
Tim-Philipp Müller [Sun, 26 Aug 2012 20:45:44 +0000 (21:45 +0100)]
tests: fix icydemux unit test
Was waiting for a tag message on the bus, which would never
come, because elements don't post those themselves any more
but let sinks post them from tag events. Only that there are
no sinks in this unit test.
Tim-Philipp Müller [Sun, 26 Aug 2012 20:27:00 +0000 (21:27 +0100)]
tests: fix videocrop crop_to_1x1 unit test for GRAY8 format
Update table with pixel values with the value actually produced
by videotestsrc.
Sjoerd Simons [Mon, 27 Aug 2012 07:00:45 +0000 (09:00 +0200)]
pulsesrc: Only print caps if they're provided
Michael Rubinstein [Fri, 24 Aug 2012 18:43:08 +0000 (19:43 +0100)]
videomixer: fix endianness check on systems where non-glib endianness defines are not set
On Windows LITTLE_ENDIAN without the G_ in was not defined, so the
test comes out wrong.
Wim Taymans [Wed, 22 Aug 2012 15:23:25 +0000 (17:23 +0200)]
udpsink: don't crash on NULL error
Check if there is an error before retrieving its message.
See https://bugzilla.gnome.org/show_bug.cgi?id=682481
Stefan Sauer [Wed, 22 Aug 2012 11:30:19 +0000 (13:30 +0200)]
Automatic update of common submodule
From 668acee to 4f962f7
Stefan Sauer [Wed, 22 Aug 2012 11:18:00 +0000 (13:18 +0200)]
configure: bump gtk-doc req to 1.12 (mar-2009)
This allows us to e.g. unconditionally use gtkdoc-rebase.
Martin Ertsaas [Wed, 22 Aug 2012 09:21:38 +0000 (11:21 +0200)]
osxvideosink: Make osxvideosink use the non-deprecated threading api from glib.
https://bugzilla.gnome.org/show_bug.cgi?id=682446
Arun Raghavan [Tue, 14 Aug 2012 10:10:31 +0000 (15:40 +0530)]
pulsesrc: Handle negotiation events
This makes sure that we:
a) Destroy an existing stream if a negotiate() request comes in: this is
required when receiving a downstream renegotiation request after a
stream has been created.
b) Create a new stream on prepare(): this is required since we do a
setcaps() in negotiate(), which causes the stream to be dropped by a
ringbuffer release() call (this does not happen during first negotiation
since the release is only done on a running ringbuffer). The subsequent
call to ringbuffer acquire() fails because the stream was lost on
release().
https://bugzilla.gnome.org/show_bug.cgi?id=681247
Arun Raghavan [Tue, 14 Aug 2012 10:08:27 +0000 (15:38 +0530)]
pulse: Clear unpositioned flag when setting positions
If converting a PA channel map to gst channel positions results in a
valid set of channel positions, we clear the unpositioned flag from the
ringbuffer spec.
Arun Raghavan [Tue, 14 Aug 2012 04:07:45 +0000 (09:37 +0530)]
pulsesrc: Remove redundant channel-mask setting for stereo case
The gstaudio helper libraries already take care of this case for us.
Arun Raghavan [Tue, 14 Aug 2012 04:06:30 +0000 (09:36 +0530)]
pulsesrc: Don't use memset to set invalid channel positions
This itereates over the GstAudioInfo to set invalid channel positions
rather than use memset() which works right now because it assumes that
GST_AUDIO_CHANNEL_POSITION_INVALID is -1.
Tim-Philipp Müller [Wed, 22 Aug 2012 09:30:04 +0000 (10:30 +0100)]
gdkpixbufsink: minor docs improvement
Tim-Philipp Müller [Wed, 22 Aug 2012 09:23:24 +0000 (10:23 +0100)]
gdkpixbuf: re-enable already-ported gdkpixbufsink
Tim-Philipp Müller [Wed, 22 Aug 2012 09:08:08 +0000 (10:08 +0100)]
gdkpixbuf: port gdkpixbufoverlay element to 0.11
Tim-Philipp Müller [Tue, 21 Aug 2012 23:00:46 +0000 (00:00 +0100)]
gdkpixbuf: re-enable already-ported gdkpixbuf element as gdkpixbufdec
Not sure why it as disabled exactly given that it had already
been ported (though without metas or baseclass).
Move plugin_init bits into separate source file, and rename
decoder element to gdkpixbufdec.
Tim-Philipp Müller [Tue, 21 Aug 2012 22:25:47 +0000 (23:25 +0100)]
gdkpixbuf: remove old and unused gst_loader source file
Once upon a time used to load GStreamer vids via GdkPixbuf API.
Aleix Conchillo Flaque [Thu, 16 Aug 2012 23:51:16 +0000 (16:51 -0700)]
rtspsrc: make jitterbuffer drop-on-latency available (fix #682055)
Conflicts:
gst/rtsp/gstrtspsrc.h
Huacai Chen [Tue, 21 Aug 2012 11:47:45 +0000 (19:47 +0800)]
v4l2: make gst_v4l2_fill_lists() adapt to kernel 3.3+
When do v4l2_ioctl() with VIDIOC_ENUMINPUT fails on some devices,
kernels before 3.3.0 return EINVAL, but newer kernels return ENOTTY.
This patch make those devices work well on kernel 3.3+.
Related kernel commit:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commit;h=
07d106d0a33d6063d2061305903deb02489eba20
Signed-off-by: Huacai Chen <chenhc@lemote.com>
Signed-off-by: Rui Wang <wangr@lemote.com>
Signed-off-by: Jie Chen <chenj@lemote.com>
Tim-Philipp Müller [Mon, 20 Aug 2012 22:30:38 +0000 (23:30 +0100)]
video/x-dvd-subpicture -> subpicture/x-dvd
Tim-Philipp Müller [Fri, 17 Aug 2012 19:52:42 +0000 (20:52 +0100)]
multifilesrc: fix example pipeline in docs
Stefan Sauer [Fri, 17 Aug 2012 12:59:57 +0000 (14:59 +0200)]
equalizer: enable presets for the n-band equalizer
Add a test for saving and restoring the preset.
Tim-Philipp Müller [Tue, 14 Aug 2012 00:20:19 +0000 (01:20 +0100)]
deinterlace: fix not-negotiated errors on variable or missing framerate in input caps
Remove some bogus code I added during porting that would error out
on missing or variable framerates in input caps. Handle this like
we do in 0.10
Fixes test_mode_disabled_passthrough unit test check.
Sjoerd Simons [Sun, 12 Aug 2012 11:16:32 +0000 (13:16 +0200)]
law: Filter layout caps field
The layout caps field shouldn't be passed through to the sink pad
of {mu,a}lawdec.
https://bugzilla.gnome.org/show_bug.cgi?id=681677
Anton Belka [Thu, 9 Aug 2012 16:41:34 +0000 (19:41 +0300)]
flacenc: allow a TOC with single alternative top-level entry
Allow a TOC that has a single alternative top-level entry
with multiple sequence sub-entries
https://bugzilla.gnome.org/show_bug.cgi?id=540891
Sebastian Dröge [Thu, 9 Aug 2012 08:31:39 +0000 (10:31 +0200)]
configure: And fix the GTK check to use the correct pkg-config package name
Sebastian Dröge [Thu, 9 Aug 2012 08:25:38 +0000 (10:25 +0200)]
configure: Fix GTK required version variable name
Matthias Clasen [Thu, 9 Aug 2012 07:35:23 +0000 (08:35 +0100)]
v4l2: fix build with recent kernels, the v4l2_buffer input field was removed
This was unused apparently and removed in the kernel in commit:
From
2b719d7baf490e24ce7d817c6337b7c87fda84c1 Mon Sep 17 00:00:00 2001
From: Sakari Ailus <sakari.ailus@iki.fi>
Date: Wed, 2 May 2012 09:40:03 -0300
Subject: [PATCH] [media] v4l: drop v4l2_buffer.input and V4L2_BUF_FLAG_INPUT
Remove input field in struct v4l2_buffer and flag V4L2_BUF_FLAG_INPUT which
tells the former is valid. The flag is used by no driver currently.
https://bugzilla.gnome.org/show_bug.cgi?id=681491
Conflicts:
sys/v4l2/gstv4l2bufferpool.c
Olivier Crête [Thu, 9 Aug 2012 00:25:36 +0000 (17:25 -0700)]
rtph264pay: Make it actually work after cleanups
Sebastian Dröge [Wed, 8 Aug 2012 15:39:07 +0000 (17:39 +0200)]
gst: Set alignment at the correct place of GstAllocationParams
Tim-Philipp Müller [Wed, 8 Aug 2012 15:25:58 +0000 (16:25 +0100)]
Back to development
Tim-Philipp Müller [Wed, 8 Aug 2012 14:22:04 +0000 (15:22 +0100)]
Release 0.11.93
Tim-Philipp Müller [Wed, 8 Aug 2012 14:17:22 +0000 (15:17 +0100)]
win32: add generated tuner-marshal/enumtypes files for v4l2src and update
And gst-indent the right rtp marshal files; add missing files to MANIFEST.
Tim-Philipp Müller [Wed, 8 Aug 2012 14:10:37 +0000 (15:10 +0100)]
gst: update disted orc files
Tim-Philipp Müller [Wed, 8 Aug 2012 10:31:59 +0000 (11:31 +0100)]
Silence some 'variable may be used uninitialized' compiler warnings
When compiling with -DG_DISABLE_ASSERT
Tim-Philipp Müller [Wed, 8 Aug 2012 09:56:51 +0000 (10:56 +0100)]
No code with side-effects inside g_assert() please
Olivier Crête [Tue, 7 Aug 2012 18:14:21 +0000 (11:14 -0700)]
multiudpsink: Return FLUSHING instead of ERROR on unlock
If the base class asks multiudpsink to unlock, then it should return
FLUSHING, not ERROR
Anton Belka [Thu, 26 Jul 2012 13:19:57 +0000 (16:19 +0300)]
flacenc: add TOC support
Add TOC as embedded cuesheets in flac files.
https://bugzilla.gnome.org/show_bug.cgi?id=54089
Mark Nauwelaerts [Tue, 7 Aug 2012 10:12:09 +0000 (12:12 +0200)]
flacparse: generate empty vorbiscomment for complete streamheaders if needed
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=681335
Olivier Crête [Tue, 7 Aug 2012 01:02:50 +0000 (18:02 -0700)]
rtpssrcdemux: Block pad while it is announced.
Block the RTP pad and associated RTCP pads while they are being
announced. This it to prevent a race where one is announced and
before the callback has connected it, the other one gets a buffer.
We can't use the "padlock" of ssrcdemux because it causes deadlocks.
Tim-Philipp Müller [Mon, 6 Aug 2012 14:00:57 +0000 (15:00 +0100)]
common: un-do accidental common update revert in commit
7b5925b5
Mark Nauwelaerts [Mon, 6 Aug 2012 12:50:53 +0000 (14:50 +0200)]
rtpmparobustdepay: set correct data_size for generated dummy frame
... which prevents getting stuck in a loop if such one is needed.
Mark Nauwelaerts [Mon, 6 Aug 2012 12:50:03 +0000 (14:50 +0200)]
rtpmparobustdepay: improve and fix debug statement
... so it really informs about next rather than past frame.
Mark Nauwelaerts [Mon, 6 Aug 2012 10:34:55 +0000 (12:34 +0200)]
rtpmparobustdepay: update available bytewriter space when repositioning
... and add some more assert to catch potential surprises early on.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=680558
Sebastian Dröge [Sat, 4 Aug 2012 10:47:44 +0000 (12:47 +0200)]
gst: Add stream-id to stream-start events
Sebastian Dröge [Sat, 4 Aug 2012 10:54:32 +0000 (12:54 +0200)]
matroskademux: Chain up to the parent class' query handler if no pad is provided
Andoni Morales Alastruey [Wed, 1 Aug 2012 23:48:29 +0000 (01:48 +0200)]
osxvideosink: add a better detection for the main run loop
Xavi Artigas [Fri, 27 Jul 2012 14:13:49 +0000 (16:13 +0200)]
directsoundsink: Do not overwrite the DS buffer when testing for AC3 support
https://bugzilla.gnome.org/show_bug.cgi?id=680706
Conflicts:
sys/directsound/gstdirectsoundsink.c
Tim-Philipp Müller [Sun, 5 Aug 2012 15:39:23 +0000 (16:39 +0100)]
Automatic update of common submodule
From 94ccf4c to 668acee
Olivier Crête [Fri, 3 Aug 2012 15:13:52 +0000 (16:13 +0100)]
rtpssrcdemux: Release lock before signalling new pad
This prevents a deadlock where something would try to push an event
through the SSRC demux from the callback, causing the pads to be iterated
and the lock taken.
Tim-Philipp Müller [Sat, 4 Aug 2012 15:10:16 +0000 (16:10 +0100)]
gst_tag_list_free -> gst_tag_list_unref
Mark Nauwelaerts [Fri, 3 Aug 2012 12:10:32 +0000 (14:10 +0200)]
rtspsrc: manage race between connection closing and flushing
... where the former can happen in task thread and the latter in mainloop
upon downward state change.
Mark Nauwelaerts [Fri, 3 Aug 2012 12:02:23 +0000 (14:02 +0200)]
flacdec: improve and relax audio frame parsing
... so as to properly recognize first audio frame.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=681077
Conflicts:
ext/flac/gstflacdec.c
René Stadler [Wed, 1 Aug 2012 10:16:41 +0000 (12:16 +0200)]
qtdemux: fix double unref of private tag buffer
Anton Belka [Mon, 30 Jul 2012 14:54:51 +0000 (17:54 +0300)]
wavparse: create TOC as needed
Avoid creating the toc if the wav has no or empty cue chunk.
Also a small code cleanup.
Tim-Philipp Müller [Sat, 28 Jul 2012 10:26:01 +0000 (11:26 +0100)]
wavparse: update for TOC API changes
Tim-Philipp Müller [Sat, 28 Jul 2012 10:22:43 +0000 (11:22 +0100)]
matroska: update for TOC API changes
Tim-Philipp Müller [Sat, 28 Jul 2012 10:20:08 +0000 (11:20 +0100)]
flacparse: update for TOC API changes
Sebastian Dröge [Fri, 27 Jul 2012 22:19:51 +0000 (00:19 +0200)]
tag: Update for taglist/tag event API changes
Mark Nauwelaerts [Fri, 27 Jul 2012 10:05:44 +0000 (12:05 +0200)]
qt(de)mux: pass private blob tags in a sample
... rather than a buffer, and the detailed info in the sample info
rather than caps.
Robert Swain [Fri, 27 Jul 2012 09:31:13 +0000 (11:31 +0200)]
videocrop: Don't return NULL from _transform_caps
If _transform_caps () returns NULL, the basetransform _transform_caps
tries to call gst_caps_is_subset () with a NULL subset which hits an
assertion.
Mark Nauwelaerts [Fri, 27 Jul 2012 09:26:18 +0000 (11:26 +0200)]
flacenc: obtain image type from the sample info
Mark Nauwelaerts [Fri, 27 Jul 2012 09:25:49 +0000 (11:25 +0200)]
flacenc: remove extraneous _unref
... since we did not obtain a buffer ref from the GstSample.
Robert Swain [Fri, 27 Jul 2012 08:14:23 +0000 (10:14 +0200)]
flacenc: Update to use GstSample tag setting API
Mark Nauwelaerts [Thu, 26 Jul 2012 14:34:21 +0000 (16:34 +0200)]
rtpmparobustdepay: modify buffer data rather than buffer itself
Mark Nauwelaerts [Thu, 26 Jul 2012 14:28:33 +0000 (16:28 +0200)]
rtpmparobustdepay: avoid leaking bytewriter instance
Robert Swain [Thu, 26 Jul 2012 14:04:23 +0000 (16:04 +0200)]
deinterlace: Fix timestamp adjustment and caps
Robert Swain [Thu, 26 Jul 2012 14:03:57 +0000 (16:03 +0200)]
deinterlace: Fix/simplify telecine state checks
Robert Swain [Thu, 26 Jul 2012 10:08:58 +0000 (12:08 +0200)]
deinterlace: Improve debug output
Robert Swain [Thu, 26 Jul 2012 10:08:36 +0000 (12:08 +0200)]
deinterlace: Fix low-latency pattern locking
Robert Swain [Tue, 24 Jul 2012 14:19:53 +0000 (16:19 +0200)]
deinterlace: RFF should be ignored in deinterlace
RFF only occurs on progressive frames in telecine sequences. For
deinterlace, we don't want these repeated fields as we will simply be
pushing the progressive frame and then moving on.
However, we need to consider RFF in order to correctly identify patterns
and adjust the timestamps.
Robert Swain [Tue, 24 Jul 2012 12:59:47 +0000 (14:59 +0200)]
deinterlace: Improve process logic
The logic now works better if we filter orphans, then progressive, then
telecine interlaced fields which need to be woven and fall through to
interlace. Telecine interlaced fields will be regularly deinterlaced if
there is no pattern lock for us to be sure that we have a telecine
pattern.
Telecine sequences that aren't 24fps progressive with RFF flags can't
really be tested until fieldanalysis is ported.
Mark Nauwelaerts [Wed, 25 Jul 2012 14:02:34 +0000 (16:02 +0200)]
flacenc: only set complete output caps once
... so as to avoid downstream complaints about missing streamheaders.
Mark Nauwelaerts [Wed, 25 Jul 2012 13:29:04 +0000 (15:29 +0200)]
flacdec: also support S24_32 output
Mark Nauwelaerts [Wed, 25 Jul 2012 13:28:14 +0000 (15:28 +0200)]
flacenc: pass correct parameters to encoder lib