platform/upstream/gstreamer.git
14 years agortspreal: don't construct config header with uninitialised bytes
Tim-Philipp Müller [Sun, 31 Jan 2010 13:03:33 +0000 (13:03 +0000)]
rtspreal: don't construct config header with uninitialised bytes

Turns out 4 + 4 + 2 + (4 * 2) is actually 18 and not 22. This avoids
a presumably unintentional padding of uninitialised bytes at the end
of the CONT tags chunk, which should be harmless but causes warnings
in valgrind (see #608533 for a test URL).

14 years agortspreal: add finalize function so we can free streams and rulebook
Tim-Philipp Müller [Sat, 30 Jan 2010 19:12:24 +0000 (19:12 +0000)]
rtspreal: add finalize function so we can free streams and rulebook

Fix memory leak in Real RTSP component (#608533).

14 years agortspreal: fix minor memory leak
Tim-Philipp Müller [Sat, 30 Jan 2010 19:10:45 +0000 (19:10 +0000)]
rtspreal: fix minor memory leak

Caps take their own reference when a buffer is added to them, so
unref buffer after adding it to caps (#608533).

14 years agordtdepay: unref input buffer when done
Tim-Philipp Müller [Sat, 30 Jan 2010 19:06:34 +0000 (19:06 +0000)]
rdtdepay: unref input buffer when done

Fixes memory leak, see #608533.

14 years agoAutomatic update of common submodule
Tim-Philipp Müller [Sat, 30 Jan 2010 15:19:56 +0000 (15:19 +0000)]
Automatic update of common submodule

From 15d47a6 to 96dc793

14 years agox264enc: make use of Access Unit delimiters configurable
Mark Nauwelaerts [Fri, 29 Jan 2010 18:47:03 +0000 (19:47 +0100)]
x264enc: make use of Access Unit delimiters configurable

API: GstX264Enc:aud

14 years agoasfdemux: Do not subtract padding twice
Thiago Santos [Fri, 22 Jan 2010 18:40:28 +0000 (15:40 -0300)]
asfdemux: Do not subtract padding twice

Only subtract implicit padding if an explicit one isn't
provided. Avoids subtracting it twice and causing
parsing errors.

Fixes #607698

14 years agoassert: g_assert_not_reached() cannot replace return statement
Stefan Kost [Fri, 22 Jan 2010 14:55:14 +0000 (16:55 +0200)]
assert: g_assert_not_reached() cannot replace return statement

Fix build with assert being turned off.

14 years agoAutomatic update of common submodule
Tim-Philipp Müller [Wed, 20 Jan 2010 00:55:12 +0000 (00:55 +0000)]
Automatic update of common submodule

From 14cec89 to 15d47a6

14 years agoasfdemux: Don't forget to update flow variable
Edward Hervey [Mon, 18 Jan 2010 17:01:55 +0000 (18:01 +0100)]
asfdemux: Don't forget to update flow variable

Forgot to update the return value in the loop.

14 years agoasfdemux: Check flow return on every push
Edward Hervey [Mon, 18 Jan 2010 16:49:06 +0000 (17:49 +0100)]
asfdemux: Check flow return on every push

We previously only aggregated flow returns after the while(push) loop,
which meant that in some cases we would end-up not properly aggregating
the flow returns.

This is based on the same flow aggregation algorithm as oggdemux.

14 years agox264enc: Add stream-format to output caps
Thiago Santos [Mon, 11 Jan 2010 19:14:44 +0000 (16:14 -0300)]
x264enc: Add stream-format to output caps

Adds stream-format to output caps of x264enc that
informs if the stream is in bytestream of avc format.

Fixes #606662

14 years agoxingmux: Fix unaligned memory access
Arnaud Patard [Mon, 11 Jan 2010 11:23:35 +0000 (12:23 +0100)]
xingmux: Fix unaligned memory access

ARM/SPARC need 32bit alignment but xingmux accesses possibly
unaligned memory, which leads to SIGBUS.

Fixes bug #586464.

14 years agoasfdemux: Use GST_STR_NULL in a couple of places.
Michael Smith [Thu, 7 Jan 2010 22:36:47 +0000 (14:36 -0800)]
asfdemux: Use GST_STR_NULL in a couple of places.

Fixes crashing on some of the log statements on win32.

14 years agormdemux: Parse and post bitrate for streams
Thiago Santos [Thu, 7 Jan 2010 19:36:08 +0000 (16:36 -0300)]
rmdemux: Parse and post bitrate for streams

Parse the bitrate of the streams and post their tags.

Fixes #599299

14 years agoasfdemux: Post bitrate tag
Thiago Santos [Thu, 7 Jan 2010 16:54:21 +0000 (13:54 -0300)]
asfdemux: Post bitrate tag

If stream bitrate object is available, post the bitrate
tags.

Fixes #599297

14 years agomp3parse: minor validation check of (Xing, VBRI) metadata
Mark Nauwelaerts [Mon, 4 Jan 2010 14:19:25 +0000 (15:19 +0100)]
mp3parse: minor validation check of (Xing, VBRI) metadata

... to detect e.g. a truncated file, rendering some of the metadata invalid.

14 years agomp3parse: use proper total_time and total_bytes in various cases
Mark Nauwelaerts [Mon, 4 Jan 2010 13:59:06 +0000 (14:59 +0100)]
mp3parse: use proper total_time and total_bytes in various cases

The correct basis for (Xing, VBRI) seek table calculations is the
byte size and duration provided by that metadata, rather than some
other (possibly even estimated) one.  This also prevents an infinite
conversion loop in (unlikely) case where a TOC is provided without
such corresponding (duration) metdata.

14 years agomp3parse: conserve stop time for non-accurate seek
Thiago Santos [Tue, 8 Dec 2009 22:55:04 +0000 (19:55 -0300)]
mp3parse: conserve stop time for non-accurate seek

Use the same strategy as accurate seeks to store
pending non-accurate seeks to avoid overwriting non-definite
stop times. When doing non-accurate seeks our position
reporting might drift off by some secs and the stream can
end up before it should.

Fixes #603695

14 years agoAutomatic update of common submodule
Mark Nauwelaerts [Mon, 21 Dec 2009 18:12:59 +0000 (19:12 +0100)]
Automatic update of common submodule

From 47cb23a to 14cec89

14 years agoamrwbdec: give decoder a rank so decodebin/playbin will use it
Tim-Philipp Müller [Fri, 18 Dec 2009 16:04:12 +0000 (16:04 +0000)]
amrwbdec: give decoder a rank so decodebin/playbin will use it

14 years agomp3parse: return false when we can't seek
Thiago Santos [Tue, 8 Dec 2009 22:01:50 +0000 (19:01 -0300)]
mp3parse: return false when we can't seek

When upstream can't seek, we return false as well

14 years agolame: Avoid crash when seeking before negotiating
Thiago Santos [Wed, 2 Dec 2009 14:21:22 +0000 (11:21 -0300)]
lame: Avoid crash when seeking before negotiating

lame's 'lgv' variable is only initialized when the caps
is negotiated, whenever a seek happens before that, it would
attempt to call a function on an empty pointer, causing the crash.

Fixes #603515

14 years agoAutomatic update of common submodule
Tim-Philipp Müller [Tue, 1 Dec 2009 15:08:07 +0000 (15:08 +0000)]
Automatic update of common submodule

From 87bf428 to 47cb23a

14 years agoAutomatic update of common submodule
Sebastian Dröge [Tue, 1 Dec 2009 13:17:50 +0000 (14:17 +0100)]
Automatic update of common submodule

From da4c75c to 87bf428

14 years agoAutomatic update of common submodule
Edward Hervey [Fri, 27 Nov 2009 17:56:13 +0000 (18:56 +0100)]
Automatic update of common submodule

From 53a2485 to da4c75c

14 years agomp3parse: fix non-flushing seek
Mark Nauwelaerts [Thu, 26 Nov 2009 14:54:45 +0000 (15:54 +0100)]
mp3parse: fix non-flushing seek

Specifically, in addition to clearing lots of variables/offsets
when receiving newsegment, also clear leftover data to match.

14 years agox264enc: Make upstream GstForceKeyUnit thread-safe
Olivier Crête [Sat, 21 Nov 2009 02:32:31 +0000 (21:32 -0500)]
x264enc: Make upstream GstForceKeyUnit thread-safe

Also send the GstForceKeyUnit event downstream when an upstream on is received,
allowing muxers or payloaders to take appropriate actions.

https://bugzilla.gnome.org/show_bug.cgi?id=602556

14 years agoAutomatic update of common submodule
Tim-Philipp Müller [Thu, 19 Nov 2009 10:31:25 +0000 (10:31 +0000)]
Automatic update of common submodule

From a3e3ce4 to 53a2485

14 years agoasfdemux: Don't call strlen() on NULL pointers
Benjamin Gaignard [Wed, 18 Nov 2009 08:58:39 +0000 (09:58 +0100)]
asfdemux: Don't call strlen() on NULL pointers

Fixes bug #602280.

14 years agoasfdemux: Remove old pads when new ones are added
Thiago Santos [Mon, 9 Nov 2009 18:02:05 +0000 (15:02 -0300)]
asfdemux: Remove old pads when new ones are added

The old pads were being removed before adding the new ones,
we should add the new ones first.

Fixes #599718

14 years agoasfdemux: Handle chained asfs on pull mode
Thiago Santos [Mon, 9 Nov 2009 16:30:45 +0000 (13:30 -0300)]
asfdemux: Handle chained asfs on pull mode

Adds chained asfs handling to pull mode. It now checks if
there is a new asf header after the last packet (when it
is possible to know how many packets are) or it tries
checking if a processed packet that fails is an header
object.

Fixes #599718

14 years agoasfdemux: properly do chained asfs on push mode
Thiago Santos [Mon, 9 Nov 2009 13:24:46 +0000 (10:24 -0300)]
asfdemux: properly do chained asfs on push mode

To properly do chained asfs work with playbin2, we need to
push eos on the old pads before removing them.

Fixes #599718

14 years agoasfdemux: add support for chained asfs (push mode)
Thiago Santos [Tue, 27 Oct 2009 20:48:03 +0000 (17:48 -0300)]
asfdemux: add support for chained asfs (push mode)

Adds support for detecting and playing chained asfs
in push mode. asfdemux tries to detect a new asf start
by identifying the header object guid in a input buffer.
When it finds it, it resets its state, removing its pads
and creates new ones for the new file.

14 years agoasfdemux: fix two small leaks
Tim-Philipp Müller [Thu, 5 Nov 2009 18:33:09 +0000 (18:33 +0000)]
asfdemux: fix two small leaks

14 years agoasfdemux: prefer WM/TrackNumber over WM/Track, it's more reliable
Tim-Philipp Müller [Thu, 5 Nov 2009 18:19:58 +0000 (18:19 +0000)]
asfdemux: prefer WM/TrackNumber over WM/Track, it's more reliable

WM/Track has a 0 base but is often wrongly written as starting from 1,
so not as reliable as WM/TrackNumber which always starts from 1.

14 years agoasfdemux: WM/Track starts counting from 0, adjust to start from 1
Tim-Philipp Müller [Thu, 5 Nov 2009 18:11:55 +0000 (18:11 +0000)]
asfdemux: WM/Track starts counting from 0, adjust to start from 1

14 years agoasfdemux: map WM/TrackNumber to GST_TAG_TRACK_NUMBER as well
Tim-Philipp Müller [Thu, 5 Nov 2009 18:11:14 +0000 (18:11 +0000)]
asfdemux: map WM/TrackNumber to GST_TAG_TRACK_NUMBER as well

There's both WM/Track and WM/TrackNumber.

14 years agoa52dec: Fix debug statement: Pass pos variable instead of time() function
Jan Schmidt [Wed, 4 Nov 2009 15:52:09 +0000 (15:52 +0000)]
a52dec: Fix debug statement: Pass pos variable instead of time() function

14 years agodvdsubdec: Fix printf format string warning
Jan Schmidt [Wed, 4 Nov 2009 15:50:17 +0000 (15:50 +0000)]
dvdsubdec: Fix printf format string warning

14 years agoasfdemux: Fix bogus variable used uninitialised warnings
Jan Schmidt [Wed, 4 Nov 2009 15:46:04 +0000 (15:46 +0000)]
asfdemux: Fix bogus variable used uninitialised warnings

14 years agopo: Don't create backup .po files
Jan Schmidt [Wed, 4 Nov 2009 13:55:39 +0000 (13:55 +0000)]
po: Don't create backup .po files

As well as preventing creation of useless backup files, it works
around a bug in gettext 0.17 on OS/X

14 years agoasfdemux: fix c99-style comments.
Michael Smith [Thu, 29 Oct 2009 18:39:13 +0000 (11:39 -0700)]
asfdemux: fix c99-style comments.

14 years agoasfdemux: accept fragments in a continued packet where the subsequent fragments
Michael Smith [Thu, 29 Oct 2009 17:34:17 +0000 (10:34 -0700)]
asfdemux: accept fragments in a continued packet where the subsequent fragments
declare a size of 0. Fixes bug 600037.

14 years agormutils: fix byteswapping
Wim Taymans [Tue, 27 Oct 2009 11:33:24 +0000 (12:33 +0100)]
rmutils: fix byteswapping

fix the byteswapping code that was wrong because of the side effects of the
READ/WRITE macros.

Fixes #599676

14 years agox264enc: Adapt to slightly modified x264 API
Robert Swain [Tue, 20 Oct 2009 19:00:44 +0000 (20:00 +0100)]
x264enc: Adapt to slightly modified x264 API

Fixes #599095.

14 years agoasfdemux: careful to avoid crash on bogus data
Thiago Santos [Mon, 26 Oct 2009 20:31:19 +0000 (17:31 -0300)]
asfdemux: careful to avoid crash on bogus data

When receiving bogus data, we have to avoid subtracting a value
larger than 'size' from 'size' variable, resulting in a wrap
that would make 'size' a really large bogus value.

Fixes #599333

14 years agowin32: Commit bumped version number
Jan Schmidt [Mon, 26 Oct 2009 00:56:37 +0000 (00:56 +0000)]
win32: Commit bumped version number

14 years agompegaudioparse: Don't use expensive glib ways to get an enum nick.
Edward Hervey [Sat, 17 Oct 2009 11:46:13 +0000 (13:46 +0200)]
mpegaudioparse: Don't use expensive glib ways to get an enum nick.

Fixes #598761

This removes a good 50% of processing time for parsing a buffer.

We do this by simply... getting the nicks that we already have handy
instead of going through the expensive glib system.

14 years agoconfigure.ac: And back to development we go...
Edward Hervey [Sat, 24 Oct 2009 18:36:41 +0000 (20:36 +0200)]
configure.ac: And back to development we go...

14 years agoRelease 0.10.13
Jan Schmidt [Wed, 21 Oct 2009 16:16:00 +0000 (17:16 +0100)]
Release 0.10.13

14 years agoUpdate .po files
Jan Schmidt [Wed, 21 Oct 2009 16:02:42 +0000 (17:02 +0100)]
Update .po files

14 years agoAutomatic update of common submodule
Stefan Kost [Fri, 16 Oct 2009 07:16:55 +0000 (10:16 +0300)]
Automatic update of common submodule

From 85d1530 to 0702fe1

14 years ago0.10.12.3 pre-release
Jan Schmidt [Thu, 15 Oct 2009 22:56:55 +0000 (23:56 +0100)]
0.10.12.3 pre-release

14 years agoAutomatic update of common submodule
Edward Hervey [Wed, 14 Oct 2009 08:41:48 +0000 (10:41 +0200)]
Automatic update of common submodule

From a3e3ce4 to 85d1530

14 years agoa52dec: Fix reference leak on all input buffers in 'dvd mode'
Jan Schmidt [Tue, 13 Oct 2009 12:05:32 +0000 (13:05 +0100)]
a52dec: Fix reference leak on all input buffers in 'dvd mode'

https://bugzilla.gnome.org/show_bug.cgi?id=598272

14 years ago0.10.12.2 pre-release
Jan Schmidt [Mon, 12 Oct 2009 12:54:27 +0000 (13:54 +0100)]
0.10.12.2 pre-release

14 years agodocs: Remove files accidentally added
Jan Schmidt [Mon, 12 Oct 2009 10:31:22 +0000 (11:31 +0100)]
docs: Remove files accidentally added

Remove the gst-plugins-ugly-plugins.args.new and
gst-plugins-ugly-plugins.signals.new files, that appear
to have been accidentally added in commit
f5d046a0b034f9bd7274291f7131ee5db1a30052

14 years agoa52dec: Add some debugging for new segment events
Jan Schmidt [Tue, 15 Sep 2009 23:00:28 +0000 (00:00 +0100)]
a52dec: Add some debugging for new segment events

14 years agompeg2dec: fix warning in macosx snow leopard
Josep Torra [Sun, 11 Oct 2009 14:18:37 +0000 (16:18 +0200)]
mpeg2dec: fix warning in macosx snow leopard

14 years agompegstream: fix warning in macosx snow leopard
Josep Torra [Sun, 11 Oct 2009 14:16:09 +0000 (16:16 +0200)]
mpegstream: fix warning in macosx snow leopard

14 years agompegaudioparse: fix warning in macosx snow leopard
Josep Torra [Sun, 11 Oct 2009 14:14:08 +0000 (16:14 +0200)]
mpegaudioparse: fix warning in macosx snow leopard

14 years agodvdsubdec: fix warning on macosx snow leopard
Josep Torra [Sun, 11 Oct 2009 14:09:11 +0000 (16:09 +0200)]
dvdsubdec: fix warning on macosx snow leopard

14 years agoasfdemux: fix warning in macosx snow leopard
Josep Torra [Sun, 11 Oct 2009 14:06:25 +0000 (16:06 +0200)]
asfdemux: fix warning in macosx snow leopard

14 years agomp3parse: don't fail SEEKING query when upstream query fails for TIME format
René Stadler [Thu, 8 Oct 2009 16:58:25 +0000 (19:58 +0300)]
mp3parse: don't fail SEEKING query when upstream query fails for TIME format

14 years agoAutomatic update of common submodule
Jan Schmidt [Thu, 8 Oct 2009 10:01:03 +0000 (11:01 +0100)]
Automatic update of common submodule

From 19fa4f3 to a3e3ce4

14 years agobuild: fprintf, sprintf, sscanf need stdio.h
Stefan Kost [Wed, 7 Oct 2009 11:22:09 +0000 (14:22 +0300)]
build: fprintf, sprintf, sscanf need stdio.h

14 years agomad: add missing include to fix the build
Stefan Kost [Fri, 2 Oct 2009 15:27:11 +0000 (18:27 +0300)]
mad: add missing include to fix the build

14 years agodvdlpcm: whitespace fixes
Wim Taymans [Mon, 5 Oct 2009 10:13:51 +0000 (12:13 +0200)]
dvdlpcm: whitespace fixes

14 years agoupdate spec file
Christian F.K. Schaller [Sun, 4 Oct 2009 21:51:41 +0000 (22:51 +0100)]
update spec file

14 years agox264enc: adjust configure checking to changed upstream
Mark Nauwelaerts [Fri, 2 Oct 2009 13:22:38 +0000 (15:22 +0200)]
x264enc: adjust configure checking to changed upstream

14 years agox264enc: conditionally adjust to new upstream API
Mark Nauwelaerts [Thu, 1 Oct 2009 16:58:42 +0000 (18:58 +0200)]
x264enc: conditionally adjust to new upstream API

Fixes #596517.

14 years agompegaudioparse: prevent infinite (re)syncing
Mark Nauwelaerts [Thu, 24 Sep 2009 15:49:52 +0000 (17:49 +0200)]
mpegaudioparse: prevent infinite (re)syncing

Conflicts:

gst/mpegaudioparse/gstmpegaudioparse.c

14 years agomp3parse: Refactor checking for sync. Make resyncing more reliable.
Michael Smith [Tue, 22 Sep 2009 19:13:38 +0000 (12:13 -0700)]
mp3parse: Refactor checking for sync. Make resyncing more reliable.

Previously, we could get false sync relatively easily - it sometimes happened
on real files. This cleans the code up a fair bit, and makes it require more
confirmation that we've found valid sync before continuing.

14 years agompegaudioparse: ensure 2 valid headers in a row when resyncing
Mark Nauwelaerts [Thu, 17 Sep 2009 14:12:29 +0000 (16:12 +0200)]
mpegaudioparse: ensure 2 valid headers in a row when resyncing

14 years agodvddemux: remove bogus ifndef
Tim-Philipp Müller [Fri, 11 Sep 2009 09:05:02 +0000 (10:05 +0100)]
dvddemux: remove bogus ifndef

14 years agoAutomatic update of common submodule
Sebastian Dröge [Sat, 5 Sep 2009 08:23:49 +0000 (10:23 +0200)]
Automatic update of common submodule

From 94f95e3 to 19fa4f3

14 years agompeg2dec: Fix GstIndex handling, incl. refcounting and NULL indizes
Sebastian Dröge [Tue, 1 Sep 2009 10:22:39 +0000 (12:22 +0200)]
mpeg2dec: Fix GstIndex handling, incl. refcounting and NULL indizes

14 years agodvdsubparse: GstAdapter is not a GstObject and should be freed with g_object_unref
Tim-Philipp Müller [Mon, 31 Aug 2009 12:42:27 +0000 (13:42 +0100)]
dvdsubparse: GstAdapter is not a GstObject and should be freed with g_object_unref

14 years agoRemove Ronald Bultje from Authors field
David Schleef [Mon, 24 Aug 2009 20:59:05 +0000 (13:59 -0700)]
Remove Ronald Bultje from Authors field

Replaced with "GStreamer maintainers
<gstreamer-devel@lists.sourceforge.net>" or just removed,
depending on the number of other authors.

14 years agoasfdemux: Remove old non-built asfmux code
David Schleef [Mon, 24 Aug 2009 20:57:14 +0000 (13:57 -0700)]
asfdemux: Remove old non-built asfmux code

Remove so people don't confuse it with the new asfmux code
in -bad.

14 years agoa52dec: Only add the MM_ACCEL_DJBFFT flag if it's defined
Sebastian Dröge [Sun, 23 Aug 2009 11:35:46 +0000 (13:35 +0200)]
a52dec: Only add the MM_ACCEL_DJBFFT flag if it's defined

It's not defined for older liba52 versions.

14 years agoa52dec: Allow liba52 to use djbfft based IMDCT transform
Mart Raudsepp [Sun, 23 Aug 2009 11:34:32 +0000 (13:34 +0200)]
a52dec: Allow liba52 to use djbfft based IMDCT transform

liba52 in a52dec-0.7.4 does not have any MMX, MMXEXT or 3DNOW based
IMDCT transform acceleration. It does however have a software based
acceleration using the djbfft library (D.J. Bernstein's library for
fourier transforms - Extremely fast library for floating-point
convolution). So allow liba52 to use it through the MM_ACCEL_DJBFFT
flag.
The liba52 copy in MPlayer sources does have SSE, 3dnowext, 3dnow
and AltiVec implementations, but those are checked for first, and
djbfft is chosen only if none of those is available - good in the
case of some distributions including a port of the MPlayer changes
in their system a52dec library.

The down and upmix code in liba52 doesn't seem to be disturbed by
this additional MM_ACCEL flag and will still use MMX, SSE or 3DNOW
versions if passed from oil_cpu_get_flags (SSE currently is not).

Fixes bug #592787.

14 years agoamr: Use opencore-amr pkg-config files if possible
Sebastian Dröge [Mon, 17 Aug 2009 15:31:10 +0000 (17:31 +0200)]
amr: Use opencore-amr pkg-config files if possible

This makes it possible to build the plugins when the libraries
are installed at non-standard locations.

Fixes bug #591348.

14 years agompegaudioparse: use metadata (xing, vbri) provided bytesize for conversions
Mark Nauwelaerts [Fri, 14 Aug 2009 10:07:40 +0000 (12:07 +0200)]
mpegaudioparse: use metadata (xing, vbri) provided bytesize for conversions

Metadata provided seek tables are consistent with metadata's view of
total size, which typically matches real size, but need not do so
(e.g. a truncated file).  Fixes seeking and position reporting
in such truncated files (although duration based on metadata may then
still be incorrect).

14 years agoconfigure: bump core/base requirements to released versions
Tim-Philipp Müller [Mon, 10 Aug 2009 21:22:39 +0000 (22:22 +0100)]
configure: bump core/base requirements to released versions

to avoid confusion.

14 years agoconfigure: Remove all mentions to the id3tag plugin which was removed
Sebastian Dröge [Tue, 4 Aug 2009 08:44:43 +0000 (10:44 +0200)]
configure: Remove all mentions to the id3tag plugin which was removed

14 years agomad: make check for mad independent of the check for id3tag
Marvin Schmidt [Sat, 1 Aug 2009 06:43:21 +0000 (08:43 +0200)]
mad: make check for mad independent of the check for id3tag

Fixes bug #590432.

15 years agortpasfdepay: set padding size to the correct value
Thiago Santos [Fri, 31 Jul 2009 03:25:43 +0000 (00:25 -0300)]
rtpasfdepay: set padding size to the correct value

asf packets in rtp packets should come with their padding fields
set to 0 and the depayload must update them to the correct
value before pushing downstream

15 years agoamr: Add AMR-WB decoder and AMR-NB encoder and decoder
Iago Toral [Mon, 27 Jul 2009 18:12:20 +0000 (20:12 +0200)]
amr: Add AMR-WB decoder and AMR-NB encoder and decoder

These are based on the OpenCore codecs.

Fixes bug #584890.

15 years agoAutomatic update of common submodule
Stefan Kost [Thu, 23 Jul 2009 21:43:07 +0000 (00:43 +0300)]
Automatic update of common submodule

From fedaaee to 94f95e3

15 years agodvdnav: remove dvdnav plugin
Tim-Philipp Müller [Wed, 22 Jul 2009 10:10:38 +0000 (11:10 +0100)]
dvdnav: remove dvdnav plugin

Remove dvdnav plugin, it has no future. resindvd is the plugin to
use for dvd playback. Note that dvdnav was never really enabled
in -ugly, you had to edit configure.ac to get it built.

15 years agompeg2dec: Fix unused variables compiler warnings in unit test
Sebastian Dröge [Sat, 20 Jun 2009 12:31:06 +0000 (14:31 +0200)]
mpeg2dec: Fix unused variables compiler warnings in unit test

15 years agoautogen.sh: Use printf instead of 'echo -n'. Check for automake-1.1[01]
Jan Schmidt [Fri, 26 Jun 2009 14:21:12 +0000 (15:21 +0100)]
autogen.sh: Use printf instead of 'echo -n'. Check for automake-1.1[01]

Check for more automake command variants. Use printf instead of 'echo -n'
for portability

15 years agoAutomatic update of common submodule
Olivier Crête [Mon, 13 Jul 2009 16:24:14 +0000 (12:24 -0400)]
Automatic update of common submodule

From 5845b63 to fedaaee

15 years agoMake git ignore more files
Tim-Philipp Muller [Tue, 10 Feb 2009 13:25:54 +0000 (13:25 +0000)]
Make git ignore more files

15 years agoasfdemux: Refactor multiple packet pull.
Edward Hervey [Mon, 29 Jun 2009 09:10:42 +0000 (11:10 +0200)]
asfdemux: Refactor multiple packet pull.

This also fixes a bug by which the first buffer (in a multi-packet mode)
passed to asf_demux_parse_packet() would have a GST_BUFFER_SIZE of the
full incoming buffer and not just of the single asf packet.

Fixes corrupted frames introduced by latest commit.

15 years agoasfdemux: use the right accurate field
Wim Taymans [Mon, 29 Jun 2009 08:58:49 +0000 (10:58 +0200)]
asfdemux: use the right accurate field

Remove accurate variable and its faulty use because the real variable is an
instance variable.

15 years agoasfdemux: Sprinkle branch prediction macros accross the code
Edward Hervey [Sun, 28 Jun 2009 15:48:11 +0000 (17:48 +0200)]
asfdemux: Sprinkle branch prediction macros accross the code

15 years agoasfdemux: Delay newsegment handling until we have a keyframe.
Edward Hervey [Sun, 28 Jun 2009 15:43:12 +0000 (17:43 +0200)]
asfdemux: Delay newsegment handling until we have a keyframe.

We now have a chance for packets to be collected before we send out the
newsegment. If we're not in accurate seeking (keyunit) it will set
the segment start/time to the keyframe's timestamp.