platform/upstream/gst-plugins-bad.git
10 years agocodecparsers_h264: initialize some fields of pic_timing structure
Aurélien Zanelli [Fri, 1 Aug 2014 14:40:49 +0000 (16:40 +0200)]
codecparsers_h264: initialize some fields of pic_timing structure

Otherwise pic timing structure can have invalid cpb_removal_delay,
dpb_output_delay or pic_struct_present_flag which are blindly retrieved
in h264parse.

https://bugzilla.gnome.org/show_bug.cgi?id=734124

10 years agodecklinksink: Initialize pixel format to bmdFormat8BitYUV
Sebastian Dröge [Fri, 1 Aug 2014 16:02:10 +0000 (18:02 +0200)]
decklinksink: Initialize pixel format to bmdFormat8BitYUV

https://bugzilla.gnome.org/show_bug.cgi?id=725871

10 years agodecklinksink: Catch errors during data flow and error out properly
Sebastian Dröge [Fri, 1 Aug 2014 14:58:23 +0000 (16:58 +0200)]
decklinksink: Catch errors during data flow and error out properly

Otherwise we will just deadlock the thread if e.g. scheduling a video
frame fails for whatever reason.

10 years agompegtspacketizer: Store PCR time/byte offsets in 64 bit integers
Sebastian Dröge [Tue, 5 Aug 2014 09:45:55 +0000 (11:45 +0200)]
mpegtspacketizer: Store PCR time/byte offsets in 64 bit integers

32 bit integers are going to overflow, especially the PCR offset to
the first PCR will overflow after about 159 seconds. This makes playback
of streams stop at 159 seconds as suddenly the timestamps are starting
again from 0. Now we have a few more years time until it happens again
and 64 bits are too small.

10 years agodashdemux: support downloading segments in reverse order
George Kiagiadakis [Wed, 6 Aug 2014 15:11:20 +0000 (18:11 +0300)]
dashdemux: support downloading segments in reverse order

When a seek with a negative rate is requested, find the target
segment where gstsegment.stop belongs in and then download from
this segment backwards until the first segment.

This allows proper reverse playback.

10 years agogl/docs: remove superflous 'the'
Matthew Waters [Fri, 1 Aug 2014 07:51:08 +0000 (17:51 +1000)]
gl/docs: remove superflous 'the'

10 years agoglupload: don't determine if frame needs upload by pointer compare
Wang Xin-yu (王昕宇) [Wed, 6 Aug 2014 12:15:38 +0000 (08:15 -0400)]
glupload: don't determine if frame needs upload by pointer compare

https://bugzilla.gnome.org/show_bug.cgi?id=734269

10 years agodashdemux: return the error to the source immediately to stop faster
Thiago Santos [Thu, 31 Jul 2014 23:52:39 +0000 (20:52 -0300)]
dashdemux: return the error to the source immediately to stop faster

When flushing, this will prevent dashdemux from trying to download more
fragments or more chunks of the same fragment before stopping.

Also improves the error handling to not transform everything non-ok into
an error.

https://bugzilla.gnome.org/show_bug.cgi?id=734014

10 years agompegtspacketizer: avoid timestamp overflows
Thiago Santos [Tue, 29 Jul 2014 05:11:54 +0000 (02:11 -0300)]
mpegtspacketizer: avoid timestamp overflows

Cause timing to break in the pipeline that can lead to a stall

https://bugzilla.gnome.org/show_bug.cgi?id=733837

10 years agogl: document GST_GL_* environment variables
Matthew Waters [Fri, 1 Aug 2014 06:41:13 +0000 (16:41 +1000)]
gl: document GST_GL_* environment variables

https://bugzilla.gnome.org/show_bug.cgi?id=733245

10 years agoglimagesink: keep the uploaded buffer around on successful redisplay
Matthew Waters [Tue, 29 Jul 2014 03:25:22 +0000 (13:25 +1000)]
glimagesink: keep the uploaded buffer around on successful redisplay

We might need it later to perform a redisplay.  GstGLUpload will take
of releasing the previous buffer when it receives a new buffer.

https://bugzilla.gnome.org/show_bug.cgi?id=733726

10 years agoglmemory: reenable the texture_rg support for !eagl
Matthew Waters [Tue, 29 Jul 2014 01:47:55 +0000 (11:47 +1000)]
glmemory: reenable the texture_rg support for !eagl

The GST_GL_HAVE_PLATFORM_EAGL is always defined we need to compare
against the value instead.

10 years agoopencv templatematch: Set caps to BGR order
David Rothlisberger [Fri, 15 Jun 2012 12:19:06 +0000 (13:19 +0100)]
opencv templatematch: Set caps to BGR order

templatematch operates on BGR data. In fact, OpenCV's IplImage always
stores color image data in BGR order -- this isn't documented at all in
the OpenCV source code, but there are hints around the web (see for
example
http://www.cs.iit.edu/~agam/cs512/lect-notes/opencv-intro/opencv-intro.html#SECTION00041000000000000000
and http://www.comp.leeds.ac.uk/vision/opencv/iplimage.html ).

gst_templatematch_load_template loads the template (the image to find)
from disk using OpenCV's cvLoadImage, so it is stored in an IplImage in
BGR order. But in gst_templatematch_chain, no OpenCV conversion
functions are used: the imageData pointer of the IplImage for the video
frame (the image to search in) is just set to point to the raw buffer
data. Without this fix, that raw data is in RGB order, so the call to
cvMatchTemplate ends up comparing the template's Blue channel against
the frame's Red channel, producing very poor results.

10 years agotemplatematch: mark pads as proxy caps
Thiago Santos [Mon, 28 Jul 2014 16:10:35 +0000 (13:10 -0300)]
templatematch: mark pads as proxy caps

Allows negotiation to happen properly

10 years agogl: fix multi gl object leaks
Wang Xin-yu (王昕宇) [Wed, 23 Jul 2014 02:25:31 +0000 (10:25 +0800)]
gl: fix multi gl object leaks

1. fix FBO leaks in decide_allocation
2. fix texture leaks in decide_allocation and reset
3. fix texture leaks in FBO incomplete error path

10 years agotsdemux: Do not scan for keyframe when in push mode.
Mathieu Duponchelle [Mon, 21 Jul 2014 16:22:18 +0000 (18:22 +0200)]
tsdemux: Do not scan for keyframe when in push mode.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=733490

10 years agotsdemux: Do not scan for a keyframe in non-accurate mode.
Mathieu Duponchelle [Mon, 21 Jul 2014 16:21:41 +0000 (18:21 +0200)]
tsdemux: Do not scan for a keyframe in non-accurate mode.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=733490

10 years agoopenjpeg: Don't remember -lopenjp2 as libs when falling back to openjpeg1
Sebastian Dröge [Mon, 21 Jul 2014 15:34:52 +0000 (17:34 +0200)]
openjpeg: Don't remember -lopenjp2 as libs when falling back to openjpeg1

10 years agogl: Reorder CFLAGS to include in-source dirs first in examples
Руслан Ижбулатов [Mon, 21 Jul 2014 11:07:28 +0000 (11:07 +0000)]
gl: Reorder CFLAGS to include in-source dirs first in examples

https://bugzilla.gnome.org/show_bug.cgi?id=733426

10 years agoopenjpeg: Win32-compatible OpenJPEG-2.0 detection
Руслан Ижбулатов [Tue, 6 May 2014 09:28:14 +0000 (09:28 +0000)]
openjpeg: Win32-compatible OpenJPEG-2.0 detection

OpenJPEG 2.0 API uses stdcall on W32 by default. This prevents normal
autoconf library macros from finding its functions.
A more compatible check is to acutally link a program that includes a
real header.

https://bugzilla.gnome.org/show_bug.cgi?id=733487

10 years agoopenni2src: Fix unitialized variable compiler warning
Sebastian Dröge [Mon, 21 Jul 2014 08:14:17 +0000 (10:14 +0200)]
openni2src: Fix unitialized variable compiler warning

gstopenni2src.cpp:721:14: error: variable 'oni_ts' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized]
  } else if (src->color->isValid () && src->sourcetype == SOURCETYPE_COLOR) {
             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
gstopenni2src.cpp:752:26: note: uninitialized use occurs here
  GST_BUFFER_PTS (buf) = oni_ts - src->oni_start_ts;
                         ^~~~~~
gstopenni2src.cpp:721:10: note: remove the 'if' if its condition is always true
  } else if (src->color->isValid () && src->sourcetype == SOURCETYPE_COLOR) {
         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
gstopenni2src.cpp:721:14: error: variable 'oni_ts' is used uninitialized whenever '&&' condition is false [-Werror,-Wsometimes-uninitialized]
  } else if (src->color->isValid () && src->sourcetype == SOURCETYPE_COLOR) {
             ^~~~~~~~~~~~~~~~~~~~~~
gstopenni2src.cpp:752:26: note: uninitialized use occurs here
  GST_BUFFER_PTS (buf) = oni_ts - src->oni_start_ts;
                         ^~~~~~
gstopenni2src.cpp:721:14: note: remove the '&&' if its condition is always true
  } else if (src->color->isValid () && src->sourcetype == SOURCETYPE_COLOR) {
             ^~~~~~~~~~~~~~~~~~~~~~~~~
gstopenni2src.cpp:642:18: note: initialize the variable 'oni_ts' to silence this warning
  uint64_t oni_ts;
                 ^
                  = 0

10 years agofacedetect: improve enum GstFaceDetectUpdates descriptions
Nicola Murino [Mon, 21 Jul 2014 19:03:26 +0000 (21:03 +0200)]
facedetect: improve enum GstFaceDetectUpdates descriptions

https://bugzilla.gnome.org/show_bug.cgi?id=730509

10 years agofacedetect: make updates on_change works as expected
Nicola Murino [Wed, 21 May 2014 10:59:57 +0000 (12:59 +0200)]
facedetect: make updates on_change works as expected

https://bugzilla.gnome.org/show_bug.cgi?id=730509

10 years agoRelease 1.4.0 1.4.0
Sebastian Dröge [Sat, 19 Jul 2014 15:27:11 +0000 (17:27 +0200)]
Release 1.4.0

10 years agoUpdate .po files
Sebastian Dröge [Sat, 19 Jul 2014 14:45:25 +0000 (16:45 +0200)]
Update .po files

10 years agopo: Update translations
Sebastian Dröge [Sat, 19 Jul 2014 10:47:43 +0000 (12:47 +0200)]
po: Update translations

10 years agowrappercamerabinsrc: set src to ready when there are no pending buffers
Thiago Santos [Wed, 16 Jul 2014 21:27:15 +0000 (18:27 -0300)]
wrappercamerabinsrc: set src to ready when there are no pending buffers

Setting to ready will block waiting for buffers to be reclaimed, so flush
before setting to null to make sure no buffers are pending

https://bugzilla.gnome.org/show_bug.cgi?id=733072

10 years agoexamples: fix compilation of gl fxtest example on win32 with Gtk+ 3.0
Tim-Philipp Müller [Thu, 17 Jul 2014 09:05:47 +0000 (10:05 +0100)]
examples: fix compilation of gl fxtest example on win32 with Gtk+ 3.0

https://bugzilla.gnome.org/show_bug.cgi?id=733063

10 years agoopenslesringbuffer: Provide the size of our array to GetDestinationOutputDeviceIDs
Sebastian Dröge [Wed, 16 Jul 2014 22:11:48 +0000 (00:11 +0200)]
openslesringbuffer: Provide the size of our array to GetDestinationOutputDeviceIDs

Otherwise it does not and just fails. It needs to know the size of the
array to not write too much to it.

10 years agoopenslessink: Silence some error debug output to log output
Sebastian Dröge [Wed, 16 Jul 2014 22:11:21 +0000 (00:11 +0200)]
openslessink: Silence some error debug output to log output

These are not really errors, just unsupported features we don't
necessarily need.

10 years agoaggregator: Reset flow_return *after* stopping the srcpad task.
Mathieu Duponchelle [Wed, 16 Jul 2014 14:57:35 +0000 (16:57 +0200)]
aggregator: Reset flow_return *after* stopping the srcpad task.

Otherwise it might be set in an already running aggregate function.

10 years agoaggregator: Flush sinkpads when stopping
Thibault Saunier [Thu, 10 Jul 2014 11:18:21 +0000 (13:18 +0200)]
aggregator: Flush sinkpads when stopping

All values are meaningless in that case, so we should make sure that
we clean everything

10 years agoaggregator: Do not forget to reset the flow return when stoping
Thibault Saunier [Thu, 10 Jul 2014 11:15:55 +0000 (13:15 +0200)]
aggregator: Do not forget to reset the flow return when stoping

Setting it to FLUSHING when the element is not started, and to OK
when it starts.

10 years agoaggregator: Handle event seqnum
Thibault Saunier [Tue, 8 Jul 2014 14:48:08 +0000 (16:48 +0200)]
aggregator: Handle event seqnum

10 years agomxfmux: Avoid taking unnecessary ref
Sebastian Rasmussen [Sun, 13 Jul 2014 14:07:02 +0000 (16:07 +0200)]
mxfmux: Avoid taking unnecessary ref

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=733123

10 years agogl: Include GL_CFLAGS while checking for GL typedefs
Mart Raudsepp [Wed, 16 Jul 2014 13:31:27 +0000 (16:31 +0300)]
gl: Include GL_CFLAGS while checking for GL typedefs

Otherwise those checks may fail at configure time if they contain extra
include paths, while at build time they are included, potentially causing
incompatible typedefs between system GL headers and gstreamer compatibility
prototypes.

https://bugzilla.gnome.org/show_bug.cgi?id=733248

10 years agogl: Always include EGL_CFLAGS in GL_CFLAGS when EGL_LIBS is added to GL_LIBS
Mart Raudsepp [Wed, 16 Jul 2014 13:29:10 +0000 (16:29 +0300)]
gl: Always include EGL_CFLAGS in GL_CFLAGS when EGL_LIBS is added to GL_LIBS

They should be handled in tandem, in case any EGL provider could require some
CFLAGS and set them (possibly once moved to prefer pkg-config files),
such as for a custom header location.

10 years agopo: update POTFILES
Piotr Drąg [Mon, 17 Mar 2014 15:09:46 +0000 (16:09 +0100)]
po: update POTFILES

https://bugzilla.gnome.org/show_bug.cgi?id=726555

10 years agod3dvideosink: Open Direct3D devices in a threadsafe way
Sebastian Dröge [Tue, 15 Jul 2014 11:29:24 +0000 (13:29 +0200)]
d3dvideosink: Open Direct3D devices in a threadsafe way

Otherwise we'll get crashes when using the device from multiple
threads, e.g. when using multiple sinks at once.

https://bugzilla.gnome.org/show_bug.cgi?id=707523

10 years agocurlsshsink: Include gst.h and other stuff before anything else
Sebastian Dröge [Fri, 11 Jul 2014 19:15:59 +0000 (21:15 +0200)]
curlsshsink: Include gst.h and other stuff before anything else

10 years agowrappercamerabinsrc: Fix caps filter caps proxying
Nicolas Dufresne [Fri, 11 Jul 2014 14:40:42 +0000 (10:40 -0400)]
wrappercamerabinsrc: Fix caps filter caps proxying

The notify signal is triggered when caps is changed. But instead of
proxying the fixed caps, we query for the caps. Hence, when we go to
READY state, we endup setting template caps on the proxied caps
filter instead of NULL, which leads to negoitation failure. Correctly
proxy NULL caps if this is the new caps. Fixes not negotiated error
when running in cheese. Also fix a leak of caps string in one of the
trace.

https://bugzilla.gnome.org/show_bug.cgi?id=732741

10 years agogl: Always build gstglmixer, not only when full OpenGL support is enabled
Sebastian Dröge [Fri, 11 Jul 2014 11:58:55 +0000 (13:58 +0200)]
gl: Always build gstglmixer, not only when full OpenGL support is enabled

10 years agocurl: Use winsock2 instead and the GLib macro for checking if we're on Windows
Sebastian Dröge [Fri, 11 Jul 2014 11:25:37 +0000 (13:25 +0200)]
curl: Use winsock2 instead and the GLib macro for checking if we're on Windows

10 years agomotioncells: Use no network API if not required
Sebastian Dröge [Fri, 11 Jul 2014 11:25:18 +0000 (13:25 +0200)]
motioncells: Use no network API if not required

10 years agoext: Include winsock.h on Windows when required
Alexey Pavlov [Fri, 11 Jul 2014 11:21:21 +0000 (13:21 +0200)]
ext: Include winsock.h on Windows when required

https://bugzilla.gnome.org/show_bug.cgi?id=733052

10 years agoRelease 1.3.91
Sebastian Dröge [Fri, 11 Jul 2014 10:13:22 +0000 (12:13 +0200)]
Release 1.3.91

10 years agogl: Link to all required libraries but not more
Sebastian Dröge [Fri, 11 Jul 2014 10:04:07 +0000 (12:04 +0200)]
gl: Link to all required libraries but not more

10 years agovideo: Properly include headers in Makefile.am
Sebastian Dröge [Fri, 11 Jul 2014 10:01:12 +0000 (12:01 +0200)]
video: Properly include headers in Makefile.am

and don't set noinst_HEADERS twice.

10 years agoUpdate .po files
Sebastian Dröge [Fri, 11 Jul 2014 09:05:59 +0000 (11:05 +0200)]
Update .po files

10 years agoatdec: Change rank to MARGINAL until it can properly handle multichannel audio
Sebastian Dröge [Fri, 11 Jul 2014 07:43:44 +0000 (09:43 +0200)]
atdec: Change rank to MARGINAL until it can properly handle multichannel audio

https://bugzilla.gnome.org/show_bug.cgi?id=727757

10 years agogl: Move GstGLMixer to the plugin for now
Sebastian Dröge [Fri, 11 Jul 2014 07:41:05 +0000 (09:41 +0200)]
gl: Move GstGLMixer to the plugin for now

It depends on GstAggregator and we don't want to install headers
for that yet.

https://bugzilla.gnome.org/show_bug.cgi?id=732207

10 years agolibs: Don't install headers and pc files for libgstwayland/badvideo/badbase
Sebastian Dröge [Fri, 11 Jul 2014 07:33:57 +0000 (09:33 +0200)]
libs: Don't install headers and pc files for libgstwayland/badvideo/badbase

These will disappear after 1.4.0 and it would be rather annoying if
people started depending on them.

https://bugzilla.gnome.org/show_bug.cgi?id=732207

10 years agoopus: Fix a double-unref in the Opus header code
Philip Withnall [Thu, 10 Jul 2014 14:52:46 +0000 (15:52 +0100)]
opus: Fix a double-unref in the Opus header code

The headers were never getting reffed when being added to the headers
list, which is later unreffed-and-freed by the caller (e.g.
gst_opus_parse_parse_frame()).

https://bugzilla.gnome.org/show_bug.cgi?id=733013

10 years agopo: Update translations
Sebastian Dröge [Fri, 11 Jul 2014 06:54:30 +0000 (08:54 +0200)]
po: Update translations

10 years agoexamples: Add CAT section and CA descriptor support
Edward Hervey [Thu, 10 Jul 2014 09:29:41 +0000 (11:29 +0200)]
examples: Add CAT section and CA descriptor support

https://bugzilla.gnome.org/show_bug.cgi?id=732986

10 years agompegts: Add parsing for CA descriptor (0x05)
Edward Hervey [Thu, 10 Jul 2014 09:24:12 +0000 (11:24 +0200)]
mpegts: Add parsing for CA descriptor (0x05)

https://bugzilla.gnome.org/show_bug.cgi?id=732986

10 years agocamerabin2: Flush downstream after setting src to READY
Nicolas Dufresne [Mon, 7 Jul 2014 15:52:18 +0000 (11:52 -0400)]
camerabin2: Flush downstream after setting src to READY

v4l2src requires all buffers to come back in order to cleanly stop
streaming. Flushing the pipline should force all buffers to come
back.

https://bugzilla.gnome.org/show_bug.cgi?id=732741

10 years agotests/icles/dccp: Release reference to parent
Sebastian Rasmussen [Sun, 6 Jul 2014 09:28:34 +0000 (11:28 +0200)]
tests/icles/dccp: Release reference to parent

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=732977

10 years agoglimagesink: Keep aspect ratio by default
Nicolas Dufresne [Wed, 9 Jul 2014 19:03:42 +0000 (15:03 -0400)]
glimagesink: Keep aspect ratio by default

The expected default behaviour for video sink is to maintain the
aspect ratio. Fix the default value to reflect this. The property
default was already TRUE, but the value was not initially TRUE.

10 years agompegts: No need to check for NULL before calling g_free()
Sebastian Rasmussen [Sat, 5 Jul 2014 23:55:50 +0000 (01:55 +0200)]
mpegts: No need to check for NULL before calling g_free()

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=732789

10 years agompegts: Don't confuse slice allocator with regular one
Sebastian Rasmussen [Sat, 5 Jul 2014 23:55:16 +0000 (01:55 +0200)]
mpegts: Don't confuse slice allocator with regular one

Previously selector_bytes and private_data_bytes were sometimes allocated and
free using the normal allocator and sometimes using the slice allocator.
Additionally prefer g_strdup() to g_memdup() for strings.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=732789

10 years agompegts: Fix dvb linkage leaks
Edward Hervey [Wed, 9 Jul 2014 05:50:13 +0000 (07:50 +0200)]
mpegts: Fix dvb linkage leaks

CID #1224130

10 years agompegts: use getter for egde linkage descriptor type
Stefan Ringel [Fri, 4 Jul 2014 12:04:47 +0000 (14:04 +0200)]
mpegts: use getter for egde linkage descriptor type

https://bugzilla.gnome.org/show_bug.cgi?id=730914

10 years agompegts: docs: add missed *_free methods
Stefan Ringel [Fri, 4 Jul 2014 12:02:22 +0000 (14:02 +0200)]
mpegts: docs: add missed *_free methods

https://bugzilla.gnome.org/show_bug.cgi?id=730914

10 years agompegts: fix annotation
Stefan Ringel [Tue, 24 Jun 2014 17:18:56 +0000 (19:18 +0200)]
mpegts: fix annotation

https://bugzilla.gnome.org/show_bug.cgi?id=730914

10 years agompegts: Add padding to public structures
Edward Hervey [Wed, 9 Jul 2014 05:38:38 +0000 (07:38 +0200)]
mpegts: Add padding to public structures

Allows use to add API in the future without breaking ABI. We broke the API/ABI
once between 1.2 and 1.4, let's try to avoid this in the future even if this
is an unstable library.

https://bugzilla.gnome.org/show_bug.cgi?id=730914

10 years agompegts: atsc: fix leak of short name
Thiago Santos [Tue, 8 Jul 2014 18:43:32 +0000 (15:43 -0300)]
mpegts: atsc: fix leak of short name

10 years agompegts: fix leak of language code
Thiago Santos [Tue, 8 Jul 2014 17:56:13 +0000 (14:56 -0300)]
mpegts: fix leak of language code

10 years agovideoaggregator: Fix some more the locking logic in update_src_caps
Thibault Saunier [Sun, 6 Jul 2014 21:30:53 +0000 (23:30 +0200)]
videoaggregator: Fix some more the locking logic in update_src_caps

We need the GST_OBJECT_LOCK only to iterate the sinkpads, nothing else.

https://bugzilla.gnome.org/show_bug.cgi?id=732750

10 years agovideoaggregator: fix broken locking in update_src_caps function
Tim-Philipp Müller [Sun, 6 Jul 2014 21:16:48 +0000 (22:16 +0100)]
videoaggregator: fix broken locking in update_src_caps function

We would unlock an already-unlocked mutex that we never re-locked.

https://bugzilla.gnome.org/show_bug.cgi?id=732750

10 years agoaggregator: fix locking
Tim-Philipp Müller [Sun, 6 Jul 2014 15:17:06 +0000 (16:17 +0100)]
aggregator: fix locking

We would unlock a mutex we never locked on SEGMENT
events.

10 years agofacedetect: Make cascades dir a little more portable
Nicolas Dufresne [Sat, 5 Jul 2014 01:00:38 +0000 (21:00 -0400)]
facedetect: Make cascades dir a little more portable

10 years agohanddetect: Use already mapped image
Nicolas Dufresne [Sat, 5 Jul 2014 00:55:11 +0000 (20:55 -0400)]
handdetect: Use already mapped image

No need to map again the image, it's already handled by the base class.

10 years agohanddetect: Move size warning in set_caps so it's called once
Nicolas Dufresne [Sat, 5 Jul 2014 00:39:49 +0000 (20:39 -0400)]
handdetect: Move size warning in set_caps so it's called once

10 years agofaceblur: Port to OpencvVideoFilter base class
Nicolas Dufresne [Sat, 5 Jul 2014 00:33:41 +0000 (20:33 -0400)]
faceblur: Port to OpencvVideoFilter base class

This fixes issue whit black frames when special memory, like GlMemory is
in used.

https://bugzilla.gnome.org/show_bug.cgi?id=732756

10 years agocvvideofilter: Don't try to make buffer writable
Nicolas Dufresne [Sat, 5 Jul 2014 00:31:50 +0000 (20:31 -0400)]
cvvideofilter: Don't try to make buffer writable

First this is handle by base transform, hence this is a no-op, and if it wasn't it
would lead to a buffer copy being leaked, and then an unreffed buffer being
pushed downstream.

https://bugzilla.gnome.org/show_bug.cgi?id=732756

10 years agocvvideofilter: Check buffer_map return value
Nicolas Dufresne [Sat, 5 Jul 2014 00:30:20 +0000 (20:30 -0400)]
cvvideofilter: Check buffer_map return value

Check the resturn value and cleanly fail if we could not mapped the buffers.

https://bugzilla.gnome.org/show_bug.cgi?id=732756

10 years agomimdec: Print invalid fourcc in error message in hex
Sebastian Rasmussen [Fri, 4 Jul 2014 01:19:42 +0000 (03:19 +0200)]
mimdec: Print invalid fourcc in error message in hex

Previously this was printed as characters which caused later processing
of the error message to sometimes warn about non-UTF-8 characters.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=732715

10 years agoaiffparse: Print invalid fourcc in error message in hex
Sebastian Rasmussen [Fri, 4 Jul 2014 01:20:22 +0000 (03:20 +0200)]
aiffparse: Print invalid fourcc in error message in hex

Previously this was printed as characters which caused later processing
of the error message to sometimes warn about non-UTF-8 characters.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=732715

10 years agoconfigure: fix uninitialized variables with --disable-external
Matthieu Bouron [Fri, 4 Jul 2014 15:51:58 +0000 (17:51 +0200)]
configure: fix uninitialized variables with --disable-external

https://bugzilla.gnome.org/show_bug.cgi?id=732744

10 years agoopenni2src: Add proper clean up of OpenNI2 objects
Arun Raghavan [Tue, 1 Jul 2014 07:04:44 +0000 (12:34 +0530)]
openni2src: Add proper clean up of OpenNI2 objects

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Don't embed C++ objects in our GObject
Arun Raghavan [Tue, 1 Jul 2014 06:48:45 +0000 (12:18 +0530)]
openni2src: Don't embed C++ objects in our GObject

Since C++ objects shoudl be properly constructed, we keep only pointers
to them and manually construct them on the heap.

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Close device when stopping the stream
Arun Raghavan [Tue, 1 Jul 2014 06:48:19 +0000 (12:18 +0530)]
openni2src: Close device when stopping the stream

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Fix timestamping
Arun Raghavan [Tue, 1 Jul 2014 06:28:36 +0000 (11:58 +0530)]
openni2src: Fix timestamping

OpenNI2 makes no guarantees of timestamp starting from zero, just that
it will be a millisecond timestamp. Make timestamps start from zero
manually so things work correctly.

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Make the location property not be mandatory
Arun Raghavan [Tue, 1 Jul 2014 06:26:59 +0000 (11:56 +0530)]
openni2src: Make the location property not be mandatory

Our calls to device open already handle the unset location case (by
opening any available device).

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Open device on NULL->READY
Arun Raghavan [Tue, 1 Jul 2014 06:26:15 +0000 (11:56 +0530)]
openni2src: Open device on NULL->READY

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Fix deadlock when _get_caps() is called before READY
Arun Raghavan [Tue, 1 Jul 2014 06:23:46 +0000 (11:53 +0530)]
openni2src: Fix deadlock when _get_caps() is called before READY

The object lock was not being dropped in the empty case. Restructured
the code a bit to make this sort of error less likely.

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agoopenni2src: Mark element as a live source
Arun Raghavan [Tue, 1 Jul 2014 06:21:21 +0000 (11:51 +0530)]
openni2src: Mark element as a live source

https://bugzilla.gnome.org/show_bug.cgi?id=732535

10 years agod3dvideosink: Always lock the D3D surfaces in write mode
Sebastian Dröge [Thu, 3 Jul 2014 17:10:26 +0000 (19:10 +0200)]
d3dvideosink: Always lock the D3D surfaces in write mode

Locking them in readonly mode can give different stride to mapping
in write mode, which then causes rendering to be broken.

Happened on all (many?) NVIDIA GPUs.

Thanks to voskater15@gmail.com for hinting at the problem.

https://bugzilla.gnome.org/show_bug.cgi?id=712809

10 years agod3dhelpers: Swap UV planes properly for YV12 as compared to I420
Sebastian Dröge [Thu, 3 Jul 2014 17:05:22 +0000 (19:05 +0200)]
d3dhelpers: Swap UV planes properly for YV12 as compared to I420

If we only do it in one place colors will look funny.

10 years agoamc: Properly integrate COLOR_QCOM_FormatYVU420SemiPlanar32m again
Sebastian Dröge [Thu, 3 Jul 2014 15:53:33 +0000 (17:53 +0200)]
amc: Properly integrate COLOR_QCOM_FormatYVU420SemiPlanar32m again

https://bugzilla.gnome.org/show_bug.cgi?id=732683

10 years agoamcaudiodec: audio/mpeg is always MPEG 1 layer 3
Sebastian Dröge [Thu, 3 Jul 2014 13:07:23 +0000 (15:07 +0200)]
amcaudiodec: audio/mpeg is always MPEG 1 layer 3

Configuring e.g. layer 2 on such a decoder usually fails unless
it additionally also accepts audio/mpeg-L2.

10 years agompegtspacketizer: Don't dereference NULL if we have no PCR yet
Sebastian Dröge [Thu, 3 Jul 2014 12:15:36 +0000 (14:15 +0200)]
mpegtspacketizer: Don't dereference NULL if we have no PCR yet

Can happen sometimes if the duration is requested before we
received enough data with a PCR.

10 years agotsdemux: Add locking around mpegtspacketiser groups.
Jan Schmidt [Thu, 3 Jul 2014 11:39:19 +0000 (21:39 +1000)]
tsdemux: Add locking around mpegtspacketiser groups.

Otherwise occasionally DURATION queries can crash, if they happen at
the wrong moment.

10 years agoglcocoa: initalize NSApp asap when using gst-launch
Julien Isorce [Thu, 3 Jul 2014 08:00:32 +0000 (09:00 +0100)]
glcocoa: initalize NSApp asap when using gst-launch

See https://bugzilla.gnome.org/show_bug.cgi?id=732661

10 years agoapplemedia: enable vtdec and vtenc only on >= 10.8 where VideoToolbox is public
Alessandro Decina [Wed, 2 Jul 2014 10:54:45 +0000 (12:54 +0200)]
applemedia: enable vtdec and vtenc only on >= 10.8 where VideoToolbox is public

Fixes: https://bugzilla.gnome.org/show_bug.cgi?id=723670

10 years agovtdec: avoid segfault when output isn't configured
Thiago Santos [Wed, 2 Jul 2014 07:54:29 +0000 (04:54 -0300)]
vtdec: avoid segfault when output isn't configured

This can help when the callback is called during/after videodecoder
base class shut down

https://bugzilla.gnome.org/show_bug.cgi?id=728435

10 years agovtdec: ignore the dropped flag if buffer was received
Thiago Santos [Wed, 2 Jul 2014 03:27:12 +0000 (00:27 -0300)]
vtdec: ignore the dropped flag if buffer was received

Apparently there is some issue with VT that makes it mark the dropped
flag even though the buffer was decoded.

https://bugzilla.gnome.org/show_bug.cgi?id=728435

10 years agod3dvideosink: Don't leak all surfaces
Sebastian Dröge [Wed, 2 Jul 2014 08:31:49 +0000 (10:31 +0200)]
d3dvideosink: Don't leak all surfaces

This was broken when disabling the buffer pool exporting.

Also disable buffer pool a bit more efficient...

10 years agod3dvideosink: PostMessage() takes integers as last parameters, not pointers
Sebastian Dröge [Wed, 2 Jul 2014 08:01:34 +0000 (10:01 +0200)]
d3dvideosink: PostMessage() takes integers as last parameters, not pointers