platform/upstream/gst-plugins-good.git
14 years agoBump gst-plugins-base requirement to 0.10.25.1
Jan Schmidt [Wed, 18 Nov 2009 02:14:46 +0000 (02:14 +0000)]
Bump gst-plugins-base requirement to 0.10.25.1

14 years agoqtdemux: identify IMA adpcm in qt properly.
Michael Smith [Wed, 18 Nov 2009 01:59:13 +0000 (17:59 -0800)]
qtdemux: identify IMA adpcm in qt properly.

14 years agoBack to development -> 0.10.17.1
Jan Schmidt [Wed, 18 Nov 2009 01:27:37 +0000 (01:27 +0000)]
Back to development -> 0.10.17.1

14 years agoAdd release 0.10.17 to the doap file
Jan Schmidt [Tue, 17 Nov 2009 01:53:08 +0000 (01:53 +0000)]
Add release 0.10.17 to the doap file

14 years agoRelease 0.10.17 RELEASE-0.10.17
Jan Schmidt [Tue, 17 Nov 2009 01:25:30 +0000 (01:25 +0000)]
Release 0.10.17

14 years agoUpdate .po files
Jan Schmidt [Tue, 17 Nov 2009 00:18:22 +0000 (00:18 +0000)]
Update .po files

14 years ago0.10.16.3 pre-release
Jan Schmidt [Fri, 13 Nov 2009 02:07:25 +0000 (02:07 +0000)]
0.10.16.3 pre-release

14 years agov4l2: Make sure to initialize variables before using them
Sebastian Dröge [Tue, 10 Nov 2009 10:52:24 +0000 (11:52 +0100)]
v4l2: Make sure to initialize variables before using them

14 years ago0.10.16.2 pre-release
Jan Schmidt [Mon, 9 Nov 2009 20:06:03 +0000 (20:06 +0000)]
0.10.16.2 pre-release

14 years agojpegdec: free temporary buffer when changing state to NULL
Tim-Philipp Müller [Mon, 9 Nov 2009 15:20:00 +0000 (15:20 +0000)]
jpegdec: free temporary buffer when changing state to NULL

Free temporary allocations in the state change function and not
only when the object is finalised.

14 years agojpegdec: only allocate as much temporary memory as needed for indirect decoding
Tim-Philipp Müller [Mon, 9 Nov 2009 11:40:25 +0000 (11:40 +0000)]
jpegdec: only allocate as much temporary memory as needed for indirect decoding

When we can't decode directly into the output buffer, make our temp buffers
only as big as needed instead of allocating for the worst case scenario (well,
we still alloc more than strictly needed for some cases, but significantly
less than before).

14 years agopulsesink: printf format fix
Tim-Philipp Müller [Thu, 5 Nov 2009 23:46:58 +0000 (23:46 +0000)]
pulsesink: printf format fix

14 years agoraw1394: printf format fixes
Tim-Philipp Müller [Thu, 5 Nov 2009 23:44:27 +0000 (23:44 +0000)]
raw1394: printf format fixes

14 years agoequalizer: printf format fix
Tim-Philipp Müller [Thu, 5 Nov 2009 23:40:15 +0000 (23:40 +0000)]
equalizer: printf format fix

14 years agoavimux: do not write empty INFO list
Thiago Santos [Thu, 5 Nov 2009 15:13:44 +0000 (12:13 -0300)]
avimux: do not write empty INFO list

avoid writing an empty INFO list chunk, both because
it is useless and because vlc refuses to play the
resulting file.

14 years agoequalizer: Notify about band property changes caused by changing number of bands
Sebastian Dröge [Thu, 5 Nov 2009 09:54:12 +0000 (10:54 +0100)]
equalizer: Notify about band property changes caused by changing number of bands

14 years agoequalizer: Make changes to band properties and the number of bands threadsafe
Sebastian Dröge [Thu, 5 Nov 2009 09:45:59 +0000 (10:45 +0100)]
equalizer: Make changes to band properties and the number of bands threadsafe

14 years agoequalizer: Fix stupid off by two bug
Sebastian Dröge [Thu, 5 Nov 2009 09:30:46 +0000 (10:30 +0100)]
equalizer: Fix stupid off by two bug

14 years agoequalizer: Add band property to select the band filter type
Sebastian Dröge [Thu, 5 Nov 2009 07:18:05 +0000 (08:18 +0100)]
equalizer: Add band property to select the band filter type

This allows per band configuration of a peak, low shelf or
high shelf filter, which can be very useful if the band frequencies
and widths are manually configured.

14 years agoequalizer: Fix code style
Sebastian Dröge [Thu, 5 Nov 2009 07:17:53 +0000 (08:17 +0100)]
equalizer: Fix code style

14 years agoequalizer: Some cleanup
Sebastian Dröge [Thu, 5 Nov 2009 07:03:13 +0000 (08:03 +0100)]
equalizer: Some cleanup

14 years agoceltpay/depay : change GST_DEBUG_OBJECT to GST_LOG_OBJECT in pay_handle_buffer and...
Gabriel Millaire [Wed, 7 Oct 2009 13:31:19 +0000 (09:31 -0400)]
celtpay/depay : change GST_DEBUG_OBJECT to GST_LOG_OBJECT in pay_handle_buffer and depay_process

14 years agoceltpay/depay: Negotiate parameters through caps
Gabriel Millaire [Fri, 2 Oct 2009 21:04:43 +0000 (17:04 -0400)]
celtpay/depay: Negotiate parameters through caps

celtdepay : added default framesize(480) channels(1) and clockrate(32000)
            depay_setcaps : now gets channels and framesize from string with default value
            depay_process : now adds timestamp to outbuf
            Added frame_size to GstRtpCeltDepay
            Changed some GST_DEBUG to GST_DEBUG_OBJECT or GST_LOG_OBJECT
celtpay : getcaps : gets channel and framesize and sets caps
          Added frame-size to static caps for audio/x-celt

14 years agodeinterlace: Pull in CFLAGS and LIBS flags from -base before core before system.
Jan Schmidt [Wed, 4 Nov 2009 15:58:34 +0000 (15:58 +0000)]
deinterlace: Pull in CFLAGS and LIBS flags from -base before core before system.

14 years agopo: Don't create backup .po files
Jan Schmidt [Thu, 15 Oct 2009 15:33:24 +0000 (16:33 +0100)]
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 agoqtdemux: init variables to make compiler on osx build bot happy
Edward Hervey [Wed, 4 Nov 2009 15:47:42 +0000 (16:47 +0100)]
qtdemux: init variables to make compiler on osx build bot happy

14 years agoqtdemux: init variables to make compiler on osx build bot happy
Tim-Philipp Müller [Tue, 3 Nov 2009 16:04:37 +0000 (16:04 +0000)]
qtdemux: init variables to make compiler on osx build bot happy

14 years agojpegdec: don't allocate big arrays on the stack
Stefan Kost [Tue, 3 Nov 2009 15:35:15 +0000 (17:35 +0200)]
jpegdec: don't allocate big arrays on the stack

Add the arrays to the instance data and allocate on first use.

14 years agodeinterlace: remove pointless call to gst_element_no_more_pads()
Tim-Philipp Müller [Sun, 1 Nov 2009 15:57:44 +0000 (15:57 +0000)]
deinterlace: remove pointless call to gst_element_no_more_pads()

14 years agolevel: fix decay to be smooth
Stefan Kost [Sat, 31 Oct 2009 22:29:57 +0000 (00:29 +0200)]
level: fix decay to be smooth

The length not having any fractional part as it was promoted to gdouble after
dividing two guint64.

14 years agolevel: calculate the message-intervall when it changes
Stefan Kost [Sat, 31 Oct 2009 22:29:24 +0000 (00:29 +0200)]
level: calculate the message-intervall when it changes

14 years agolevel: clocktime is a guint64, use right macro to init fields
Stefan Kost [Sat, 31 Oct 2009 22:14:08 +0000 (00:14 +0200)]
level: clocktime is a guint64, use right macro to init fields

14 years agolevel: use more g-style types
Stefan Kost [Sat, 31 Oct 2009 22:10:01 +0000 (00:10 +0200)]
level: use more g-style types

14 years agopulsesink: Only set the volume on stream connection if pulse >= 0.9.20 is available
Sebastian Dröge [Fri, 30 Oct 2009 08:27:59 +0000 (09:27 +0100)]
pulsesink: Only set the volume on stream connection if pulse >= 0.9.20 is available

In older versions the volume set during stream connection had
no defined sematic and usually it was a relative volume. What
was needed for our use case is an absolute volume though, otherwise
the volume will be always decreased on stream connection if it's
less than 100%.
Since pulse 0.9.20 that volume is always an absolute volume if
flat volumes are used and relative otherwise, which is the same
as for pa_context_set_sink_input_volume().

Relevant pulse changesets:
http://git.0pointer.de/?p=pulseaudio.git;a=commit;h=f27a50691c8fe45bac7dd6b21fac91a359def3a1
http://git.0pointer.de/?p=pulseaudio.git;a=commit;h=2501687579e359d5032a4d165b2ffc8f5b1b8ba6

14 years agoavidemux: use segment_full when we can
Wim Taymans [Tue, 27 Oct 2009 17:07:18 +0000 (18:07 +0100)]
avidemux: use segment_full when we can

Use segment_full so that we can pass the applied rate to the segment values. We
will change the applied rate when we implement skip mode.

14 years agowavenc: Fix buffer offset by moving length incrementation
Robert Swain [Sat, 17 Oct 2009 23:16:06 +0000 (00:16 +0100)]
wavenc: Fix buffer offset by moving length incrementation

14 years agoosxvideosink: Create the video NSView in READY->PAUSED rather than NULL->READY
Michael Smith [Sat, 24 Oct 2009 01:31:14 +0000 (18:31 -0700)]
osxvideosink: Create the video NSView in READY->PAUSED rather than NULL->READY

14 years agoosxvideo: explicitly link to GST_LIBS
Michael Smith [Sat, 24 Oct 2009 01:28:22 +0000 (18:28 -0700)]
osxvideo: explicitly link to GST_LIBS

14 years agoAdd dependencies of gstriff to things that link to gstriff, needed on Win32.
Michael Smith [Sat, 24 Oct 2009 01:09:43 +0000 (18:09 -0700)]
Add dependencies of gstriff to things that link to gstriff, needed on Win32.

14 years agortp examples: remove executable bits from C files.
Michael Smith [Sat, 24 Oct 2009 00:25:17 +0000 (17:25 -0700)]
rtp examples: remove executable bits from C files.

14 years agotests: disable all jitterbuffer tests for now
Tim-Philipp Müller [Fri, 23 Oct 2009 10:21:44 +0000 (11:21 +0100)]
tests: disable all jitterbuffer tests for now

Since even the one enabled seems to fail.

14 years agotests: also include the new test for prev commit
Stefan Kost [Thu, 22 Oct 2009 10:39:58 +0000 (13:39 +0300)]
tests: also include the new test for prev commit

14 years agotests: add a jitterbuffer test
Stefan Kost [Thu, 22 Oct 2009 10:19:07 +0000 (13:19 +0300)]
tests: add a jitterbuffer test

Tests pushing a few buffers in various order and asserting the order sent by the
jitterbuffer. Contains two disabled tests that need more work.

14 years agomatroskamux: Dirac "muxing" units end on EOS too
Sebastian Dröge [Thu, 22 Oct 2009 10:30:14 +0000 (12:30 +0200)]
matroskamux: Dirac "muxing" units end on EOS too

A Dirac muxing unit are all non-picture, non-end-of-sequence
packets up to and including the first picture or eos packet.

See http://www.diracvideo.org/wiki/index.php/ContainerFormatMappingGuidelines

14 years agoavidemux: fix compilation with debugging disabled
Tim-Philipp Müller [Thu, 22 Oct 2009 01:09:08 +0000 (02:09 +0100)]
avidemux: fix compilation with debugging disabled

total_idx is always evaluated.

14 years agocacasink: minor cleanups for header.
Priit Laes [Mon, 19 Oct 2009 18:59:46 +0000 (21:59 +0300)]
cacasink: minor cleanups for header.

Use G_BEGIN_DECLS macros, remove unused variables and fix typo.
See #599018.

14 years agocacasink: exit properly when invalid driver has been selected.
Priit Laes [Mon, 19 Oct 2009 18:59:23 +0000 (21:59 +0300)]
cacasink: exit properly when invalid driver has been selected.

See #599018.

14 years agoavidemux: Stop scanning at the last entry... and not the one before :)
Edward Hervey [Tue, 20 Oct 2009 16:23:28 +0000 (18:23 +0200)]
avidemux: Stop scanning at the last entry... and not the one before :)

This ensures we actually push out everything

14 years agoqtdemux: unpack more information into image/x-j2c caps
Andy Wingo [Tue, 20 Oct 2009 15:20:55 +0000 (17:20 +0200)]
qtdemux: unpack more information into image/x-j2c caps

* gst/qtdemux/qtdemux_fourcc.h: Add new fourccs for use by the mj2
  unpacker.
* gst/qtdemux/qtdemux.c (qtdemux_parse_trak): Unpack JPEG2000 component
  mapping and channel definitions from the jp2h header. Will add
  component-map and channel-definitions elements to the caps if the
  component maps or channel definitions are nonstandard, where standard
  order means RGB, 444 packed YUV, or greyscale, with no alpha channel.

Fixes #598915.

14 years agotests: include stdio.h for sscanf
Stefan Kost [Tue, 20 Oct 2009 14:33:41 +0000 (17:33 +0300)]
tests: include stdio.h for sscanf

14 years agoFix the StreamVolume interface not being advertised
Bastien Nocera [Mon, 19 Oct 2009 14:21:57 +0000 (15:21 +0100)]
Fix the StreamVolume interface not being advertised

gst_pulsesink_interface_supported() was missing a check for it.

https://bugzilla.gnome.org/show_bug.cgi?id=598933

14 years agolevel: code cleanup
Stefan Kost [Fri, 16 Oct 2009 18:14:14 +0000 (21:14 +0300)]
level: code cleanup

Use gdouble instead of double. Calculate falloff_time once instead of twice.

14 years agoavidemux: MEMDUMP the junk blobs
Edward Hervey [Sun, 18 Oct 2009 13:52:02 +0000 (15:52 +0200)]
avidemux: MEMDUMP the junk blobs

It will only actually pull the junk blobs from upstream if the memdump
level is activated

14 years agoavidemux: Some avi files have INFO lists in the headers.
Edward Hervey [Sun, 18 Oct 2009 13:51:34 +0000 (15:51 +0200)]
avidemux: Some avi files have INFO lists in the headers.

14 years agoavidemux: Don't seek on empty streams
Edward Hervey [Sun, 18 Oct 2009 14:02:01 +0000 (16:02 +0200)]
avidemux: Don't seek on empty streams

14 years agoavidemux: Ensure _calculate_durations_from_index only uses valid streams
Edward Hervey [Sun, 18 Oct 2009 13:50:39 +0000 (15:50 +0200)]
avidemux: Ensure _calculate_durations_from_index only uses valid streams

14 years agoavidemux: Only call convert function if we have strf.auds
Edward Hervey [Sun, 18 Oct 2009 13:49:29 +0000 (15:49 +0200)]
avidemux: Only call convert function if we have strf.auds

14 years agoavidemux: Use first indexed stream for seeking.
Edward Hervey [Sun, 18 Oct 2009 13:48:06 +0000 (15:48 +0200)]
avidemux: Use first indexed stream for seeking.

In the future, main_stream can be adjusted to contain the optimal stream
as mentionned in the FIXME line 3440

14 years agoavidemux: Only expose streams that actually have something in it.
Edward Hervey [Sun, 18 Oct 2009 13:46:48 +0000 (15:46 +0200)]
avidemux: Only expose streams that actually have something in it.

This guarantees that in pull-mode, all streams have a valid index to
work with.

14 years agoavidemux: Properly mark presence of index.
Edward Hervey [Sun, 18 Oct 2009 13:40:37 +0000 (15:40 +0200)]
avidemux: Properly mark presence of index.

Instead of blindly saying we have an index, only do so if we have a
non-empty index.

14 years agopulse: never apply volume more than once
Lennart Poettering [Sat, 17 Oct 2009 00:18:53 +0000 (02:18 +0200)]
pulse: never apply volume more than once

Generally decisions on the volume of the stream should be done inside of
PA, not inside of Gst. Only PA knows how volumes translate between
devices and s on.

This patch makes sure that all volumes set via the volume property are
only applied *once* to the underlying stream. After applying them the
client side will not store them anymore. This should make sure that
really only user-triggered volume changes are forwarded to server, but
the client never tries to save/restore the volume internally.

Fixes bug #595231.

14 years agopulsesink: Initialize gettext for the translated strings in plugin_init()
Sebastian Dröge [Sat, 17 Oct 2009 06:55:16 +0000 (08:55 +0200)]
pulsesink: Initialize gettext for the translated strings in plugin_init()

14 years agopulse: use 'performer' as a fallback for 'artist' tag
Lennart Poettering [Fri, 16 Oct 2009 22:10:30 +0000 (00:10 +0200)]
pulse: use 'performer' as a fallback for 'artist' tag

14 years agopulse: when constructing a stream title from tag data make sure it is translatable
Lennart Poettering [Fri, 16 Oct 2009 22:09:36 +0000 (00:09 +0200)]
pulse: when constructing a stream title from tag data make sure it is translatable

14 years agopulse: loop while connecting to server
Lennart Poettering [Fri, 16 Oct 2009 22:06:15 +0000 (00:06 +0200)]
pulse: loop while connecting to server

pthread does not guarantee that there are no spurious condition variable
wakeups, neither does pa_threaded_mainloop_xxx() which is a wrapper
around it. So we need to loop around the _wait() function to make sure
we get the right wakeup.

Also, unify the order of the wait loops across the file.

14 years agopulse: mainloop creation can fail too, so handle that
Lennart Poettering [Fri, 16 Oct 2009 22:05:10 +0000 (00:05 +0200)]
pulse: mainloop creation can fail too, so handle that

14 years agopulse: adjust CHECK_DEAD_GOTO macro to glib style
Lennart Poettering [Fri, 16 Oct 2009 22:03:06 +0000 (00:03 +0200)]
pulse: adjust CHECK_DEAD_GOTO macro to glib style

14 years agopulse: make a few things smaller by making them bitfields
Lennart Poettering [Fri, 16 Oct 2009 15:28:42 +0000 (17:28 +0200)]
pulse: make a few things smaller by making them bitfields

14 years agopulse: bump minimum libpulse version to 0.9.10
Lennart Poettering [Fri, 16 Oct 2009 15:26:41 +0000 (17:26 +0200)]
pulse: bump minimum libpulse version to 0.9.10

Older versions than 0.9.10 are really really old and buggy. Drop
compatibility with them. Nobody should run anything that old.

Also see: https://bugzilla.gnome.org/show_bug.cgi?id=595029

14 years agodebugutils: register pushfilesrc element
Mark Nauwelaerts [Fri, 16 Oct 2009 16:18:31 +0000 (18:18 +0200)]
debugutils: register pushfilesrc element

14 years agoavimux: support (some) VBR audio muxing
Mark Nauwelaerts [Fri, 16 Oct 2009 15:28:09 +0000 (17:28 +0200)]
avimux: support (some) VBR audio muxing

AVI format can handle VBR audio provided audio chunks are of fixed duration
(cfr fixed duration video frames).  Apply this approach to (always) parsed
raw AAC and (if parsed) to MPEG-1/2 audio.

See #368681.

14 years agojpegdec: fix branch hints
Tim-Philipp Müller [Fri, 16 Oct 2009 12:41:45 +0000 (13:41 +0100)]
jpegdec: fix branch hints

Remove inappropriate branching hints and add some new ones.

14 years agojpegdec: fix regression in indirect decode path
Tim-Philipp Müller [Fri, 16 Oct 2009 11:33:04 +0000 (12:33 +0100)]
jpegdec: fix regression in indirect decode path

Revert variable name back to what it was before the G_LIKELY was
added (in commit 69c24fb9). The code works better that way.

14 years agojpegdec: fix regression with certain formats
Tim-Philipp Müller [Fri, 16 Oct 2009 01:47:38 +0000 (02:47 +0100)]
jpegdec: fix regression with certain formats

Fix regression introduced by previous commit (#598517).

14 years agojpegdec: don't use decompress structure members we shouldn't be using
Tim-Philipp Müller [Thu, 15 Oct 2009 18:49:55 +0000 (19:49 +0100)]
jpegdec: don't use decompress structure members we shouldn't be using

14 years agojpegdec: remove some unused members from jpegdec instance structure
Tim-Philipp Müller [Wed, 14 Oct 2009 16:53:52 +0000 (17:53 +0100)]
jpegdec: remove some unused members from jpegdec instance structure

14 years agobuild: use gst-glib-gen.mak to fix the glib build rules.
Stefan Kost [Fri, 16 Oct 2009 08:53:38 +0000 (11:53 +0300)]
build: use gst-glib-gen.mak to fix the glib build rules.

The build rules in glib-gen.mak were using pattern rules in a non save way.

14 years agoAutomatic update of common submodule
Stefan Kost [Fri, 16 Oct 2009 07:15:35 +0000 (10:15 +0300)]
Automatic update of common submodule

From 85d1530 to 0702fe1

14 years agoavidemux: adjust flow return aggregation to updated loop_data
Mark Nauwelaerts [Thu, 15 Oct 2009 19:04:02 +0000 (21:04 +0200)]
avidemux: adjust flow return aggregation to updated loop_data

In particular, each stream is now treated separately, and one stream's
EOS should not lead to overall EOS.

14 years agoqtdemux: check some more atom sizes prior to parsing
Mark Nauwelaerts [Thu, 15 Oct 2009 09:52:35 +0000 (11:52 +0200)]
qtdemux: check some more atom sizes prior to parsing

14 years agortsp: handle events in TCP mode
Wim Taymans [Thu, 15 Oct 2009 11:19:13 +0000 (13:19 +0200)]
rtsp: handle events in TCP mode

We need to handle events in TCP mode so that we can reply to the LATENCY event
with TRUE.

14 years agoavidemux: add missing argument in debug message
Mark Nauwelaerts [Thu, 15 Oct 2009 09:24:45 +0000 (11:24 +0200)]
avidemux: add missing argument in debug message

14 years agoflvmux: Use loop test to prevent timeout on slow machines
Marvin Schmidt [Wed, 14 Oct 2009 16:58:06 +0000 (18:58 +0200)]
flvmux: Use loop test to prevent timeout on slow machines

Partially fixes bug #597739.

14 years agortspsrc: forward events into the rtpbin
Wim Taymans [Wed, 14 Oct 2009 14:15:48 +0000 (16:15 +0200)]
rtspsrc: forward events into the rtpbin

Only catch the SEEK event on the srcpad and let other events enter the rtpbin.

14 years agomatroskademux: Fix late tags finding
Thiago Santos [Wed, 14 Oct 2009 14:33:24 +0000 (11:33 -0300)]
matroskademux: Fix late tags finding

Use the correct taglist variable when notifying of late tags.

14 years agoavidemux: use GstIndex for (limited) seeking in push mode
Mark Nauwelaerts [Wed, 14 Oct 2009 11:09:03 +0000 (13:09 +0200)]
avidemux: use GstIndex for (limited) seeking in push mode

... but disable this for now.  Although it basically works fine,
user experience might be shaky (depending on taste), since there
is no keyframe info in push mode.

14 years agoavidemux: add GstIndex support
Mark Nauwelaerts [Wed, 14 Oct 2009 11:08:47 +0000 (13:08 +0200)]
avidemux: add GstIndex support

14 years agoavidemux: also determine duration in push mode
Mark Nauwelaerts [Wed, 14 Oct 2009 09:55:33 +0000 (11:55 +0200)]
avidemux: also determine duration in push mode

14 years agoqtdemux: add GstIndex support
Mark Nauwelaerts [Wed, 14 Oct 2009 09:54:44 +0000 (11:54 +0200)]
qtdemux: add GstIndex support

14 years agov4l2src: Set duration on buffers
Thiago Santos [Wed, 14 Oct 2009 10:38:26 +0000 (07:38 -0300)]
v4l2src: Set duration on buffers

Use framerate to estimate duration of buffers.

Fixes #590362

14 years agortpptdemux: only forward the lost-event to the last seen pt-number
Håvard Graff [Wed, 14 Oct 2009 10:28:55 +0000 (12:28 +0200)]
rtpptdemux: only forward the lost-event to the last seen pt-number

forward all events on all pads except for the PacketLost event, which we want to
forward to the last seen pt pad.

Fixes #598377

14 years agopulsesink: set desired minreq value to segsize/latency-time
René Stadler [Tue, 6 Oct 2009 19:28:50 +0000 (22:28 +0300)]
pulsesink: set desired minreq value to segsize/latency-time

If we let the daemon decide freely by passing -1, we end up always getting 20ms.
We want to set this value because in some cases we want to select a higher
latency-time in order to save power.

Fixes #597601

14 years agoAutomatic update of common submodule
Edward Hervey [Wed, 14 Oct 2009 08:41:21 +0000 (10:41 +0200)]
Automatic update of common submodule

From a3e3ce4 to 85d1530

14 years agotests/pipeline/flac: Fix build on macosx 10.5
Edward Hervey [Tue, 13 Oct 2009 16:33:34 +0000 (18:33 +0200)]
tests/pipeline/flac: Fix build on macosx 10.5

14 years agoavidemux: demote some warnings to debug
Wim Taymans [Tue, 13 Oct 2009 16:19:32 +0000 (18:19 +0200)]
avidemux: demote some warnings to debug

14 years agoavi: add new avi flag we might want to use
Wim Taymans [Tue, 13 Oct 2009 15:47:40 +0000 (17:47 +0200)]
avi: add new avi flag we might want to use

14 years agoavimux: calculate suggested buffer size
Wim Taymans [Tue, 13 Oct 2009 15:46:22 +0000 (17:46 +0200)]
avimux: calculate suggested buffer size

Calculate the suggested buffer size based on the largest chunk in the file.

See #597847

14 years agoavimux: add jpeg2000 to allowed caps
Wim Taymans [Tue, 13 Oct 2009 15:45:14 +0000 (17:45 +0200)]
avimux: add jpeg2000 to allowed caps

14 years agoavidemux: add debug for the superindex offsets
Wim Taymans [Tue, 13 Oct 2009 15:41:13 +0000 (17:41 +0200)]
avidemux: add debug for the superindex offsets

14 years agoqtdemux: Fix uninitialized variable warning
Jan Schmidt [Tue, 13 Oct 2009 15:02:37 +0000 (16:02 +0100)]
qtdemux: Fix uninitialized variable warning

Fix another bogus may-be-used-uninitialized warning in qtdemux