Sebastian Dröge [Sat, 12 Jun 2010 07:02:29 +0000 (09:02 +0200)]
[MOVED FROM BAD 050/134] vp8enc: Implement multipass encoding
Fixes bug #621348.
Sebastian Dröge [Mon, 14 Jun 2010 13:56:24 +0000 (15:56 +0200)]
[MOVED FROM BAD 049/134] vp8enc: Set VP8E_SET_CPUUSED to 0
This setting controls how much CPU can be used by the encoder, specified
in fractions of 16. Negative values mean strict enforcement of this
while positive values are adaptive.
The default value is -4, which means that we're not running as fast
as possible and probably are wasting some quality. 0 is the recommended
default by libvpx upstream.
Sebastian Dröge [Mon, 14 Jun 2010 13:51:30 +0000 (15:51 +0200)]
[MOVED FROM BAD 048/134] vp8enc: Use VPX defines for REALTIME, GOOD/BEST quality deadlines instead of our own
These are the values used for the speed property.
Tim-Philipp Müller [Thu, 3 Jun 2010 09:49:40 +0000 (10:49 +0100)]
[MOVED FROM BAD 047/134] vp8enc: fix printf format warning in log message
gstvp8enc.c:564: error: format ‘%d’ expects type ‘int’, but argument 8 has type ‘size_t’
gstvp8enc.c:744: error: format ‘%d’ expects type ‘int’, but argument 8 has type ‘size_t’
Tim-Philipp Müller [Fri, 3 Jul 2009 15:08:38 +0000 (16:08 +0100)]
[MOVED FROM BAD 046/134] basevideo, vp8: guard unstable API with GST_USE_UNSTABLE_API
Add some guards and fat warnings to the header files with still unstable
API, so people who just look at the installed headers know that it
actually is unstable API.
Merging previous commit into current codebase.
David Schleef [Tue, 1 Jun 2010 22:54:51 +0000 (15:54 -0700)]
[MOVED FROM BAD 045/134] basevideo: Move base video from vp8 to gst-libs
Sebastian Dröge [Wed, 26 May 2010 04:52:15 +0000 (06:52 +0200)]
[MOVED FROM BAD 044/134] vp8: Use VPX_PLANE_* instead of PLANE_*
Sebastian Dröge [Mon, 24 May 2010 09:04:02 +0000 (11:04 +0200)]
[MOVED FROM BAD 043/134] vp8: Add compatilibity defines to work with older versions of libvpx too
Philip Jägenstedt [Sun, 23 May 2010 07:28:13 +0000 (09:28 +0200)]
[MOVED FROM BAD 042/134] vp8dec: s/IMG_FMT_I420/VPX_IMG_FMT_I420/
This corresponds to upstream libvpx commit
6cd4a10e167203d1deb79abf60ee72599e97891b
Sebastian Dröge [Sat, 22 May 2010 10:55:45 +0000 (12:55 +0200)]
[MOVED FROM BAD 041/134] vp8enc: Allow a maximum keyframe distance of 0, i.e. all frames are keyframes
Sebastian Dröge [Sat, 22 May 2010 06:45:35 +0000 (08:45 +0200)]
[MOVED FROM BAD 040/134] vp8dec: Set decoder deadline from the QoS information
Sebastian Dröge [Fri, 28 May 2010 14:35:12 +0000 (16:35 +0200)]
[MOVED FROM BAD 039/134] vp8enc: Move debug output one line above where the packet is still valid
Sebastian Dröge [Fri, 28 May 2010 13:53:30 +0000 (15:53 +0200)]
[MOVED FROM BAD 038/134] vp8enc: Correctly ignore non-frame packets from the encoder
Fixes bug #619916.
Sebastian Dröge [Sat, 22 May 2010 05:44:27 +0000 (07:44 +0200)]
[MOVED FROM BAD 037/134] basevideodecoder: Take the frame duration into account when calculating the earliest time
This formula is used in many other elements too.
Fixes bug #619318.
Sebastian Dröge [Sat, 22 May 2010 05:35:01 +0000 (07:35 +0200)]
[MOVED FROM BAD 036/134] basevideodecoder: Reset QoS values when necessary
Sebastian Dröge [Sat, 22 May 2010 07:35:24 +0000 (09:35 +0200)]
[MOVED FROM BAD 035/134] vp8enc: Use GST_VIDEO_CAPS_YUV(I420) instead of handwritten I420 caps for the pad template
Fixes bug #619344.
Philip Jägenstedt [Fri, 21 May 2010 18:53:36 +0000 (20:53 +0200)]
[MOVED FROM BAD 034/134] vp8dec: drop late frames after decoding them
This saves a memcpy, which is always something.
Philip Jägenstedt [Fri, 21 May 2010 19:28:29 +0000 (21:28 +0200)]
[MOVED FROM BAD 033/134] vp8enc: threads property
Increasing from 1 to 2 threads on an Thinkpad X60s decreased encode time
in a test from ~24 s to ~19 s, so this is quite useful.
Ideally we should let 0 be the default and automatically match the number
of CPU cores (or something).
Philip Jägenstedt [Fri, 21 May 2010 13:17:46 +0000 (15:17 +0200)]
[MOVED FROM BAD 032/134] vp8enc: add mode property to switch between CBR/VBR
Always using CBR when bitrate is used isn't that great, VBR mode
can produce meaningful results too.
Sebastian Dröge [Fri, 21 May 2010 08:54:57 +0000 (10:54 +0200)]
[MOVED FROM BAD 031/134] vp8dec: Only enable postprocessing if the decoder supports it
Philip Jägenstedt [Fri, 21 May 2010 06:23:58 +0000 (08:23 +0200)]
[MOVED FROM BAD 030/134] vp8: typo: s/HAVE_VP8_DECODER/HAVE_VP8_ENCODER/
Fixup for bug #619172.
Philip Jägenstedt [Fri, 21 May 2010 06:13:06 +0000 (08:13 +0200)]
[MOVED FROM BAD 029/134] vp8: move #ifdef HAVE_VP8_ENCODER/DECODER
Otherwise we'll try including e.g. <vpx/vp8cx.h> which doesn't exist.
Sebastian Dröge [Thu, 20 May 2010 18:06:09 +0000 (20:06 +0200)]
[MOVED FROM BAD 028/134] vp8enc: Write GStreamer element and version in the vorbiscomment vendor string
Sebastian Dröge [Thu, 20 May 2010 14:49:03 +0000 (16:49 +0200)]
[MOVED FROM BAD 027/134] vp8: Only enable the encoder or decoder if it's available in libvpx
Fixes bug #619172.
Philip Jägenstedt [Thu, 20 May 2010 08:19:54 +0000 (10:19 +0200)]
[MOVED FROM BAD 026/134] vp8: exlcude dec/enc based on CONFIG_VP8_DECODER/ENCODER
This may not be very autotoolish, but works with libvpx in the state
that libvpx is actually in. Moved the debug init to the elements
themselves to minimize amount of #ifdefs
Philip Jägenstedt [Thu, 20 May 2010 07:24:53 +0000 (09:24 +0200)]
[MOVED FROM BAD 025/134] vp8enc: Limit max-latency to 25 to match libvpx
From libvpx/vp8/encoder/onyx_int.h:
#define MAX_LAG_BUFFERS (CONFIG_REALTIME_ONLY? 1 : 25)
While we don't need to be tied to what libvpx does internally, it
doesn't make sense to pretend to support longer frame lags than are
actually possible.
Sebastian Dröge [Thu, 20 May 2010 07:56:25 +0000 (09:56 +0200)]
[MOVED FROM BAD 024/134] vp8: Undef HAVE_CONFIG_H before including libvpx headers
A public libvpx header includes private headers if this is
defined, causing compilation failures because the private headers
are not installed of course.
Sebastian Dröge [Thu, 20 May 2010 06:53:12 +0000 (08:53 +0200)]
[MOVED FROM BAD 023/134] vp8enc: Some more minor adjustments for the Ogg mapping
Sebastian Dröge [Wed, 19 May 2010 21:02:19 +0000 (23:02 +0200)]
[MOVED FROM BAD 022/134] vp8dec: Fix memory leak
Sebastian Dröge [Wed, 19 May 2010 19:34:42 +0000 (21:34 +0200)]
[MOVED FROM BAD 021/134] vp8enc: Adjust Ogg mapping for the changes
Sebastian Dröge [Wed, 19 May 2010 16:12:18 +0000 (18:12 +0200)]
[MOVED FROM BAD 020/134] vp8dec: Add properties to control the VP8 decoder post processing feature
This is disabled by default for now.
Sebastian Dröge [Wed, 19 May 2010 15:16:54 +0000 (17:16 +0200)]
[MOVED FROM BAD 019/134] vp8enc: Rename keyframe-interval to max-keyframe-distance
And use default settings for buffer sizes until we expose this
somehow.
Sebastian Dröge [Wed, 19 May 2010 15:13:17 +0000 (17:13 +0200)]
[MOVED FROM BAD 018/134] vp8: Improve error handling and debug output
Sebastian Dröge [Wed, 19 May 2010 12:46:48 +0000 (14:46 +0200)]
[MOVED FROM BAD 017/134] vp8: Use correct strides and plane offsets for GStreamer
Sebastian Dröge [Tue, 18 May 2010 12:47:54 +0000 (14:47 +0200)]
[MOVED FROM BAD 016/134] vp8enc: Implement GstTagSetter interface
Sebastian Dröge [Tue, 18 May 2010 12:33:49 +0000 (14:33 +0200)]
[MOVED FROM BAD 015/134] vp8enc: Fix setting of the keyframe flag on encoded frames
Sebastian Dröge [Tue, 18 May 2010 12:30:15 +0000 (14:30 +0200)]
[MOVED FROM BAD 014/134] vp8enc: Post an error message on the bus if encoder initialization fails
Sebastian Dröge [Tue, 18 May 2010 12:28:55 +0000 (14:28 +0200)]
[MOVED FROM BAD 013/134] vp8dec: Fix memory leaks and fail if initializing the decoder fails
David Schleef [Tue, 18 May 2010 09:44:54 +0000 (02:44 -0700)]
[MOVED FROM BAD 012/134] vp8enc: Set timebase
Also misc cleanup.
Sebastian Dröge [Sun, 16 May 2010 08:36:12 +0000 (10:36 +0200)]
[MOVED FROM BAD 011/134] vp8dec: Fix decoding of invisible frames
Sebastian Dröge [Fri, 14 May 2010 12:26:34 +0000 (14:26 +0200)]
[MOVED FROM BAD 010/134] vp8enc: Update the latency when initializing the encoder
Sebastian Dröge [Fri, 14 May 2010 12:02:53 +0000 (14:02 +0200)]
[MOVED FROM BAD 009/134] vp8dec: Correctly initialize stream info before peeking at the stream
Otherwise peeking will fail and we'll get invalid values
Sebastian Dröge [Fri, 14 May 2010 09:01:29 +0000 (11:01 +0200)]
[MOVED FROM BAD 008/134] vp8dec: Make sure to pass a keyframe as first frame to the decoder, copy output frames only once and require width/height/etc on the input caps
Sebastian Dröge [Fri, 14 May 2010 08:30:18 +0000 (10:30 +0200)]
[MOVED FROM BAD 007/134] vp8enc: Add support for invisible frames and the Ogg mapping
David Schleef [Fri, 14 May 2010 08:14:46 +0000 (01:14 -0700)]
[MOVED FROM BAD 006/134] vp8dec: Fix reset after seeking
Also remove some unused code.
Sebastian Dröge [Thu, 13 May 2010 19:19:32 +0000 (21:19 +0200)]
[MOVED FROM BAD 005/134] vp8enc: Set frame numbers as buffer offsets
Sebastian Dröge [Thu, 13 May 2010 19:18:08 +0000 (21:18 +0200)]
[MOVED FROM BAD 004/134] vp8enc: Always get as many frames as possible from the encoder
Sebastian Dröge [Thu, 13 May 2010 19:08:03 +0000 (21:08 +0200)]
[MOVED FROM BAD 003/134] vp8enc: Fill the oldest pending frame instead of the newest
Sebastian Dröge [Thu, 13 May 2010 18:20:32 +0000 (20:20 +0200)]
[MOVED FROM BAD 002/134] vp8enc: Correctly set delta unit flag for non-keyframes
David Schleef [Thu, 13 May 2010 08:04:04 +0000 (01:04 -0700)]
[MOVED FROM BAD 001/134] vp8: Add encoder/decoder
Christian Fredrik Kalager Schaller [Sat, 15 Sep 2012 20:16:52 +0000 (22:16 +0200)]
Update spec file with F18 name change and add deinterlacer
Mark Nauwelaerts [Sat, 15 Sep 2012 17:06:06 +0000 (19:06 +0200)]
use gst_element_factory_get_metadata to replace obsolete API
Mark Nauwelaerts [Fri, 14 Sep 2012 15:55:16 +0000 (17:55 +0200)]
replace _get_caps_reffed with _get_caps
Mark Nauwelaerts [Fri, 14 Sep 2012 15:07:26 +0000 (17:07 +0200)]
replace gst_element_class_set_details_simple with gst_element_class_set_metadata
Wim Taymans [Fri, 14 Sep 2012 11:30:37 +0000 (13:30 +0200)]
fix more caps
Tim-Philipp Müller [Fri, 14 Sep 2012 01:57:44 +0000 (02:57 +0100)]
Back to development
Tim-Philipp Müller [Fri, 14 Sep 2012 01:48:43 +0000 (02:48 +0100)]
Release 0.11.94
Tim-Philipp Müller [Fri, 14 Sep 2012 00:50:44 +0000 (01:50 +0100)]
po: update translations
Tim-Philipp Müller [Fri, 14 Sep 2012 00:46:14 +0000 (01:46 +0100)]
docs: update docs
Tim-Philipp Müller [Thu, 13 Sep 2012 23:47:38 +0000 (00:47 +0100)]
tests: push stream-start and segment events in wavpackenc test
Wim Taymans [Thu, 13 Sep 2012 08:56:27 +0000 (10:56 +0200)]
v4l2: remove unused properties
Wim Taymans [Thu, 13 Sep 2012 08:15:54 +0000 (10:15 +0200)]
pulsesrc: disable reconfigure
See https://bugzilla.gnome.org/show_bug.cgi?id=683902
Jan Schmidt [Tue, 11 Sep 2012 05:09:59 +0000 (22:09 -0700)]
deinterlace: Don't treat every custom-downstream event as EOS
Don't fall through to the EOS handling after receiving a
custom-downstream event.
Stefan Sauer [Wed, 12 Sep 2012 19:05:44 +0000 (21:05 +0200)]
collectpads: remove gst_collect_pads_add_pad_full
Rename gst_collect_pads_add_pad_full() to gst_collect_pads_add_pad() and fix all
invocations.
Mark Nauwelaerts [Wed, 12 Sep 2012 15:14:46 +0000 (17:14 +0200)]
udp: add include for IPPROTO_*
Mark Nauwelaerts [Wed, 12 Sep 2012 14:39:08 +0000 (16:39 +0200)]
udp: properly match braces and cpp directives
Fixes compilation where IPV6_TCLASS not defined.
Edward Hervey [Wed, 12 Sep 2012 12:42:07 +0000 (14:42 +0200)]
shapewipe: Use default query handler where needed
And clean up get_caps code while I'm at it
Wim Taymans [Wed, 12 Sep 2012 11:28:07 +0000 (13:28 +0200)]
deinterlace: improve framerate transform
Handle G_MAXINT in the framerates better. If we cannot double or divide the
framerate, clamp to the smallest/largest possible value we can express instead
of failing.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=683861
Wim Taymans [Wed, 12 Sep 2012 11:17:54 +0000 (13:17 +0200)]
deinterlace: small cleanup
Youness Alaoui [Fri, 7 Sep 2012 21:20:57 +0000 (17:20 -0400)]
videomixer2: Adding nv12 and nv21 support
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=683841
Michael Smith [Wed, 12 Sep 2012 08:18:53 +0000 (10:18 +0200)]
qtdemux: add support for prores
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=683839
Tim-Philipp Müller [Tue, 11 Sep 2012 23:16:31 +0000 (00:16 +0100)]
tests: fix most of the rganalysis unit tests
Before the element would post messages on the bus itself, now
the sinks do that based on the tag events they receive. But
since we don't have proper sink elements in these unit tests,
but just dangling pads, we have to post the tag messages the
test checks for ourselves.
Down from 52/55 failing to 7/52 failing.
Mark Nauwelaerts [Tue, 11 Sep 2012 15:36:51 +0000 (17:36 +0200)]
ext, gst: only activate in pull mode if upstream is seekable
Wim Taymans [Tue, 11 Sep 2012 13:38:23 +0000 (15:38 +0200)]
v4l2: disable renegotiation
We can't yet wait for the bufferpool to DRAIN before starting renegotiation so
disable it for now.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=682770
Mark Nauwelaerts [Tue, 11 Sep 2012 10:48:39 +0000 (12:48 +0200)]
tests: rtpbin: port to the new GLib thread API
Mark Nauwelaerts [Tue, 11 Sep 2012 10:36:56 +0000 (12:36 +0200)]
directsoundsink: port to the new GLib thread API
Wim Taymans [Tue, 11 Sep 2012 09:59:54 +0000 (11:59 +0200)]
qtdemux: don't reset segment
Don't reset the segment because we need the values for accumulation. the segment
is reset at start and after a flushing seek. Fixes some problems with files with
quicktime segments.
Wim Taymans [Mon, 10 Sep 2012 15:14:37 +0000 (17:14 +0200)]
tests: fix id3demux test
Mark Nauwelaerts [Mon, 10 Sep 2012 12:31:02 +0000 (14:31 +0200)]
gst: adjust comment style
Mark Nauwelaerts [Mon, 10 Sep 2012 12:30:42 +0000 (14:30 +0200)]
avidemux: remove defunct commented code
Wim Taymans [Mon, 10 Sep 2012 11:35:15 +0000 (13:35 +0200)]
pulsesrc: consider stream alive when not connected yet
When we start and renegotiate, there is a moment where the stream is created but
not yet connected. Make sure all functions deal with this situation correctly
instead of erroring out.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=681247
Wim Taymans [Mon, 10 Sep 2012 10:15:25 +0000 (12:15 +0200)]
pulsesrc: don't fail when not negotiated yet
When get_time is called but we are not yet negotiated, return 0 instead of
posting an error. It's possible that the base class is still negotiating when
our get_time is called.
Wim Taymans [Mon, 10 Sep 2012 09:32:25 +0000 (11:32 +0200)]
update for audio base src api change
Tim-Philipp Müller [Sun, 9 Sep 2012 23:42:52 +0000 (00:42 +0100)]
video/x-3ivx and video/x-xvid -> video/mpeg,mpegversion=4
If it ever turns out that we really must use thoe specific
fourccs and not the generic one, we can still add a flavor
field to the caps later.
Daniela [Fri, 7 Sep 2012 14:15:42 +0000 (16:15 +0200)]
rtspsrc: avoid leak
When setup fails, make sure to cleanup afterwards.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=673509
Mark Nauwelaerts [Fri, 7 Sep 2012 13:23:44 +0000 (15:23 +0200)]
rtpamrdepay: unmap rtp buffer
... thereby plugging a memleak.
Mark Nauwelaerts [Fri, 7 Sep 2012 12:13:17 +0000 (14:13 +0200)]
tests: rtp-payloading: adjust to modified bufferlist semantics
... now implemented by buffer memory blocks.
Mark Nauwelaerts [Fri, 7 Sep 2012 12:11:39 +0000 (14:11 +0200)]
rtph264pay: avoid crashing on NULL access in debug message
Mark Nauwelaerts [Fri, 7 Sep 2012 12:11:02 +0000 (14:11 +0200)]
rtph263ppay: plug caps leak
Wim Taymans [Thu, 6 Sep 2012 15:09:20 +0000 (17:09 +0200)]
deinterlace: remove redundant _set_allocation call
Mark Nauwelaerts [Thu, 6 Sep 2012 15:05:00 +0000 (17:05 +0200)]
tests: deinterlace: do not leak deinterlace pads
Mark Nauwelaerts [Thu, 6 Sep 2012 15:04:39 +0000 (17:04 +0200)]
deinterlace: plug some leaks
Wim Taymans [Thu, 6 Sep 2012 14:49:02 +0000 (16:49 +0200)]
deinterlace: reuse core function for GCD
Mark Nauwelaerts [Thu, 6 Sep 2012 14:31:00 +0000 (16:31 +0200)]
deinterlace: support filter in getcaps
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.