Commit graph

7849 commits

Author SHA1 Message Date
David Schleef
3c0170aa4b faac: Fix channel layout
Adjusted the channel mappings to agree with the default channel mappings
in the spec, which is presumably what faac is using.  Fixes #607105.
2010-01-16 15:34:33 -08:00
David Schleef
82c07135d4 flite: handle samplesperbuffer 2010-01-16 15:32:02 -08:00
David Schleef
43147671c1 flite: Add flitetestsrc
Flite is a speech synthesizer library (BSD-ish license).  This element
negotiates a (possibly) multichannel audio stream, and then continually
outputs buffers naming the number and position of the channel.  It's
kinda crappy, but I needed it for testing purposes.
2010-01-16 13:03:24 -08:00
Sebastian Dröge
b95eda9c9a frei0r: Make sure that property names always start with a letter
Otherwise GLib throws assertions
2010-01-16 17:02:52 +01:00
Sebastian Dröge
e476ff9883 shapewipe: Replace floating point arithmetic in the inner processing loops by integer arithmetic 2010-01-16 16:52:11 +01:00
Sebastian Dröge
007b667366 faac: Only accept specific channel layouts as required by AAC
Fixes bug #607105.
2010-01-16 13:05:58 +01:00
Sebastian Dröge
be9ca4d5e1 examples: Fix build and signal connection for GtkBuilder stuff 2010-01-16 11:41:20 +01:00
Sebastian Dröge
cf89e526e3 configure: Check for gmodule-export-2.0 for GtkBuilder 2010-01-16 11:26:15 +01:00
Sebastian Dröge
55c48789b7 icles: Link with GTK+ 2010-01-16 11:23:00 +01:00
Sebastian Dröge
92b7c0ad15 configure: Require GTK+ >= 2.12 for GtkBuilder
GTK+ 2.12 is already older than 2 years so this shouldn't
be a problem for anybody. Also it's only used in examples.
2010-01-16 11:16:05 +01:00
Javier Jardón
878ccf4a56 tests: Port from libglade to GtkBuilder
Fixes bug #601108.
2010-01-16 11:14:55 +01:00
Julien Moutte
f33344ab0c mpeg4videoparse: don't set an invalid framerate.
When fixed_vop_rate is not set we can not set a framerate based on
vop_time_increment_resolution as it would most likely be wrong.
Don't set any framerate on the caps in that case.
2010-01-15 21:34:02 +01:00
Thiago Santos
774b7b33cc qtmux: Actually use new caps info on renegotiation
Following the previous qtmux commit, this patch tries
to use the new info added to the caps to fill the 'trak'
atom's fields and children atoms. This way qtmux will
use the late added 'codec_data' when h264parse adds
it in the following pipeline:

videotestsrc num-buffers=200 ! x264enc byte-stream=true ! \
h264parse output-format=0 ! qtmux ! \
filesink location=test.mov
2010-01-14 19:09:44 -03:00
Thiago Santos
0b4c0890a5 qtmux: Do caps renegotiation when it only adds fields
Qtmux can accept caps renegotiation if the new caps is a
superset of the old one, meaning upstream added new info to
the caps. This patch still doesn't make qtmux update any
atoms info from the new info, but at least it doesn't
reject the new caps anymore.

A pipeline that reproduces this use case is:

videotestsrc num-buffers=200 ! x264enc byte-stream=true ! \
h264parse output-format=0 ! qtmux ! \
filesink location=test.mov
2010-01-14 19:09:44 -03:00
Sebastian Dröge
bb757ad0bb assrender: Improve debug output a bit and simplify code 2010-01-14 12:59:21 +01:00
Sebastian Dröge
d85be17a00 assrender: Make sure that the video buffer is writable before writing to it 2010-01-14 12:57:04 +01:00
Wim Taymans
9356b9df10 apexsink: change volume property to a double
Change the volume property from int to a double so that playbin2 can use it.
Keep the volume internally as a percent for now with the default 75% mapped to
a volume of 1.0.
some cleanups, add some FIXME and comments.

Fixes #606878
2010-01-14 12:49:39 +01:00
Mark Nauwelaerts
dbc1f9f2d6 qtmux: provide request pads under wider conditions
Fixes #606859.
2010-01-13 19:30:45 +01:00
Thiago Santos
5a605b883b qtmux: Only accept avc-sample h264
qtmux and mp4mux should only accept h264 in avc-sample
format
2010-01-13 10:35:57 -03:00
Thiago Santos
eba1357244 h264parse: Add stream-format to caps
Adds the new stream-format field to h264parse caps

Fixes #606662
2010-01-13 10:35:57 -03:00
Tim-Philipp Müller
e15d4c44aa build: include cruft.mak and add recently-merged plugins to cruft list 2010-01-13 00:42:21 +00:00
Tim-Philipp Müller
4be0225870 mmssrc: post error message in addition to redirect when connect fails
When we can't connect to the server, we should still post a proper
error message on the bus, so that applications that don't handle the
redirect messsage get a proper error message (and a hint to look for
the redirect).

Fixes #606636.
2010-01-12 23:55:20 +00:00
Thiago Santos
ba06eddb40 h264parse: Correctly rewrite all nal prefixes
When converting to byte-stream do not forget to rewrite
all nal prefixes instead of only the first one.
2010-01-12 18:18:39 -03:00
Mark Nauwelaerts
8be4e3725f faac: improve faac configuration
In particular, align configuration some more with reference binary faac.

Fixes #606726.
2010-01-12 19:46:01 +01:00
Thiago Santos
6e1f2cf392 mpegtsmux: Fix mistake on previous commit
Use the correct variables and variable format in the warning
message
2010-01-12 15:21:29 -03:00
Edward Hervey
58b9479423 flacparse: Initialize variables.
Fixes build on $#@*( macosx
2010-01-12 18:55:53 +01:00
Edward Hervey
c41fbca13c tests: Add neonhttpsrc to the list of elements ignored for state-change
Setting neonhttpsrc to PAUSED/PLAYING requires a running http server on
localhost.
2010-01-12 17:05:13 +01:00
Thiago Santos
e6ab078785 h264parse: remove codec_data if output is bytestream
codec_data should be removed from the src pad caps if the
output is in bytestream format

Fixes #606657
2010-01-12 09:58:48 -03:00
Thiago Santos
ddeb6e17fd mpegtsmux: Do not crash on misinterpreted h264
Avoid crashing when bytestream h264 is interpreted as avc format
h264

Fixes #606657
2010-01-12 09:37:34 -03:00
b3f0b029f3 win32: Include config.h before anything else. Fix mpegdemux LIBADD
Because config.h defines __MSVCRT_VERSION__, which should be defined
before inclusion of any system header.

Also fixes mpegdemux Makefile.am LIBADD typo.

Fixes #606665
2010-01-12 12:31:56 +01:00
Руслан Ижбулатов
d61abce8e4 librfb: port rfbsrc to MinGW
Fixes #606677
2010-01-12 00:29:57 +00:00
Thiago Santos
08d13fd191 aacparse: Also add stream-format to template caps
Do not forget to add stream-format to template caps
off aacparse
2010-01-11 13:42:49 -03:00
Thiago Santos
a428ad82c7 Rename aac's stream-format 'none' to 'raw'
Renames aac's stream-format from previous commits from none to
raw
2010-01-11 13:42:48 -03:00
Руслан Ижбулатов
01f514160f configure: include <windows.h> into msacm.h conftest 2010-01-11 16:46:42 +01:00
Thiago Santos
2d647b6ea8 aacparse: update tests to stream-format changes
Updates aacparse unit tests to check for stream-format
correctness as well.
2010-01-11 12:10:02 -03:00
Thiago Santos
7cfe3643db aacparse: Add stream-format to output caps
Adds stream-format field to output caps
2010-01-11 11:01:10 -03:00
Thiago Santos
5b8b363310 qtmux: Only accept stream-format='none' aac
Only accept raw aac streams (stream-format=none) to avoid
generating invalid files.

Fixes #604925
2010-01-11 11:01:10 -03:00
Thiago Santos
635b0c1129 faac: Add format to output caps
Adds 'stream-format' to output caps to inform if the stream is
adts or raw. This makes it possible for muxers to reject
input streams on caps if they don't support adts instead of
generating a invalid file.

Fixes #604925
2010-01-11 11:01:09 -03:00
Tim-Philipp Müller
c31559841a resindvd: fix unaligned access in this copy of mpegdemux as well
See #606371.
2010-01-11 01:03:38 +00:00
Tim-Philipp Müller
ee65cabbb3 vmnc: use existing macros to avoid alignment issues 2010-01-11 01:01:46 +00:00
Tim-Philipp Müller
904a4e0f48 rfbsrc: use existing macros to avoid alignment issues 2010-01-11 00:58:12 +00:00
Tim-Philipp Müller
64c6ee0b4e asfmux: fix alignment and endianness issue 2010-01-11 00:24:59 +00:00
Tim-Philipp Müller
0bd4d4e27b zbar: use GST_MAKE_FOURCC to create FOURCC 2010-01-11 00:16:18 +00:00
Linqiang Pu
d68c2d4ab9 mpegdemux: fix allignment issue
Use GST_READ_UINT32_BE instead of GUINT32_FROM_BE to
fix int allignment issues on ARM

Fixes #606371
2010-01-10 10:09:55 -03:00
Marc-André Lureau
67e5d76d08 mpegtsmux: only video can have non-delta-unit 2010-01-08 16:06:38 +00:00
Thomas Vander Stichele
5fab65aafb mpegtsmux: mark delta-unit on outgoing buffers
Signed-off-by: Marc-André Lureau <mlureau@flumotion.com>
2010-01-08 13:07:53 +01:00
Mark Nauwelaerts
4067cd13c2 h264parse: optimize src caps updating 2010-01-07 19:00:14 +01:00
Mark Nauwelaerts
51aaa30fab h264parse: fix framerate calculation
... and also remove a superfluous duplicate debug statement.
2010-01-07 19:00:00 +01:00
Mark Nauwelaerts
997ec65cdf h264parse: fix regression in output buffer timestamping
Specifically, properly mark a buffer (NALU or AU) to start a frame or not.
2010-01-07 18:59:58 +01:00
Michael Smith
e2da34e6b6 mpeg4videoparse: avoid crash if vos < 5 bytes. 2010-01-06 11:20:41 -08:00