ffmpeg.git
4 years agocheckasm: add vp9 MC tests.
Ronald S. Bultje [Tue, 15 Sep 2015 20:41:29 +0000 (16:41 -0400)]
checkasm: add vp9 MC tests.

4 years agoavfilter/vf_stereo3d: simplify blank calculations and check height
Paul B Mahol [Tue, 15 Sep 2015 18:55:47 +0000 (18:55 +0000)]
avfilter/vf_stereo3d: simplify blank calculations and check height

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agolavc/utils: Correctly show bit_rate >INT_MAX.
Carl Eugen Hoyos [Tue, 15 Sep 2015 18:56:23 +0000 (20:56 +0200)]
lavc/utils: Correctly show bit_rate >INT_MAX.

Fixes ticket #2089.

4 years agoavcodec/ffv1enc: Fix error message when the requested version does not support the...
Michael Niedermayer [Tue, 15 Sep 2015 18:27:14 +0000 (20:27 +0200)]
avcodec/ffv1enc: Fix error message when the requested version does not support the requested features

Found-by: "Peter B." <pb@das-werkstatt.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavfilter/vsrc_cellauto: use the name 's' for the pointer to the private context
Ganesh Ajjanagadde [Fri, 28 Aug 2015 05:31:02 +0000 (01:31 -0400)]
avfilter/vsrc_cellauto: use the name 's' for the pointer to the private context

Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
4 years agoREADME.md: avoid Github pull requests
Lou Logan [Mon, 14 Sep 2015 21:24:22 +0000 (13:24 -0800)]
README.md: avoid Github pull requests

Signed-off-by: Lou Logan <lou@lrcd.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Reviewed-by: Ganesh Ajjanagadde <gajjanag@mit.edu>
4 years agoavfilter/f_sendcmd: use the name 's' for the pointer to the private context
Ganesh Ajjanagadde [Wed, 9 Sep 2015 18:30:35 +0000 (14:30 -0400)]
avfilter/f_sendcmd: use the name 's' for the pointer to the private context

Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
4 years agoavfilter/af_stereowiden: fix 2nd argument for av_calloc
Paul B Mahol [Tue, 15 Sep 2015 16:27:52 +0000 (16:27 +0000)]
avfilter/af_stereowiden: fix 2nd argument for av_calloc

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agolavf: Switch probesize and max_analyze_duration to 64bit.
Carl Eugen Hoyos [Tue, 15 Sep 2015 15:36:38 +0000 (17:36 +0200)]
lavf: Switch probesize and max_analyze_duration to 64bit.

Both are only accessible through AVOptions.

4 years agolavf: Switch bitrate to 64bit unless compatibility with avconv was requested.
Carl Eugen Hoyos [Tue, 15 Sep 2015 15:29:38 +0000 (17:29 +0200)]
lavf: Switch bitrate to 64bit unless compatibility with avconv was requested.

Based on a patch by Steve Swanson, swanysteve at gmail.

Fixes ticket #2089.

4 years agolavc: Switch bitrate to 64bit unless compatibility with avconv was requested.
Michael Niedermayer [Tue, 15 Sep 2015 16:01:32 +0000 (18:01 +0200)]
lavc: Switch bitrate to 64bit unless compatibility with avconv was requested.

4 years agodoc/filters: mention newly added windowing functions
Paul B Mahol [Tue, 15 Sep 2015 15:22:33 +0000 (15:22 +0000)]
doc/filters: mention newly added windowing functions

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/vf_vidstabdetect: use the name 's' for the pointer to the private context
Ganesh Ajjanagadde [Mon, 7 Sep 2015 14:10:06 +0000 (10:10 -0400)]
avfilter/vf_vidstabdetect: use the name 's' for the pointer to the private context

Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
4 years agovp9: switch min_tile_cols location so it shifts up instead of down.
Ronald S. Bultje [Mon, 14 Sep 2015 18:50:47 +0000 (14:50 -0400)]
vp9: switch min_tile_cols location so it shifts up instead of down.

This fixes cases where the shifted number is 64, but we shifted non-
zero numbers away in the shift. The change makes behaviour consistent
with libvpx.

4 years agortpdec: add a trace when jitter buffer is full
Eloi BAIL [Mon, 14 Sep 2015 18:02:16 +0000 (14:02 -0400)]
rtpdec: add a trace when jitter buffer is full

This commit adds an error trace when jitter buffer
is full. It helps to understand leading decoding issues.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agortpdec: inform jitter buffer size
Eloi BAIL [Mon, 14 Sep 2015 18:02:17 +0000 (14:02 -0400)]
rtpdec: inform jitter buffer size

This commit print as AV_LOG_INFO the jitter buffer
size. It might be the default value or the value set by application.

Signed-off-by: Eloi BAIL <eloi.bail@savoirfairelinux.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavfilter/vf_xbr: use the name 's' for the pointer to the private context
Ganesh Ajjanagadde [Fri, 28 Aug 2015 05:37:48 +0000 (01:37 -0400)]
avfilter/vf_xbr: use the name 's' for the pointer to the private context

Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
4 years agoavfilter: add stereowiden filter
Paul B Mahol [Sat, 12 Sep 2015 04:01:18 +0000 (04:01 +0000)]
avfilter: add stereowiden filter

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/avf_showfreqs: add lanczos and gauss windowing functions
Paul B Mahol [Tue, 15 Sep 2015 09:13:53 +0000 (09:13 +0000)]
avfilter/avf_showfreqs: add lanczos and gauss windowing functions

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/avf_showcqt: draw text optionally
Paul B Mahol [Tue, 1 Sep 2015 12:33:08 +0000 (12:33 +0000)]
avfilter/avf_showcqt: draw text optionally

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavcodec/dxv: remove code that is never reached
Paul B Mahol [Sat, 12 Sep 2015 13:51:44 +0000 (13:51 +0000)]
avcodec/dxv: remove code that is never reached

The function is automatically called.

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavformat/avidec: Change (last)pos to 64bit
Michael Niedermayer [Tue, 15 Sep 2015 02:14:55 +0000 (04:14 +0200)]
avformat/avidec: Change (last)pos to 64bit

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavformat/avidec: Workaround broken initial frame
Michael Niedermayer [Tue, 15 Sep 2015 02:01:27 +0000 (04:01 +0200)]
avformat/avidec: Workaround broken initial frame

Fixes Ticket4851

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agolavu/avstring: switch AV_ESCAPE_FLAGs to shift-based formatting
Rodger Combs [Thu, 10 Sep 2015 16:12:15 +0000 (11:12 -0500)]
lavu/avstring: switch AV_ESCAPE_FLAGs to shift-based formatting

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agolavc: Fix standalone compilation of the avi demuxer.
Carl Eugen Hoyos [Mon, 14 Sep 2015 21:12:28 +0000 (23:12 +0200)]
lavc: Fix standalone compilation of the avi demuxer.

4 years agoavformat/hlsenc: Fix the method command line parameter
Michael Niedermayer [Mon, 14 Sep 2015 19:07:37 +0000 (21:07 +0200)]
avformat/hlsenc: Fix the method command line parameter

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agowmalossless: Warn when decoding 24 bit depth
Vittorio Giovara [Mon, 14 Sep 2015 17:15:44 +0000 (19:15 +0200)]
wmalossless: Warn when decoding 24 bit depth

See ticket #4134

4 years agoaaccoder_mips: Fix indentation
Timothy Gu [Mon, 14 Sep 2015 15:54:08 +0000 (08:54 -0700)]
aaccoder_mips: Fix indentation

4 years agoavformat: fix style after recent commits
Clément Bœsch [Mon, 14 Sep 2015 15:42:12 +0000 (17:42 +0200)]
avformat: fix style after recent commits

4 years agompjpegde: trim header name/value of MIME headers
Alex Agranovsky [Sat, 12 Sep 2015 23:07:53 +0000 (19:07 -0400)]
mpjpegde: trim header name/value of MIME headers

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agompjpeg: probe should require same constraints as packet reader - both proper content...
Alex Agranovsky [Sun, 13 Sep 2015 20:47:54 +0000 (16:47 -0400)]
mpjpeg: probe should require same constraints as packet reader - both proper content-type and content-size must be present

return AVPROBE_SCORE_MAX, rather than random positive number on success

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavcodec/mips/aaccoder_mips: Sync with generic aaccoder file.
Nedeljko Babic [Fri, 11 Sep 2015 13:16:16 +0000 (15:16 +0200)]
avcodec/mips/aaccoder_mips: Sync with generic aaccoder file.

Code in aaccoder_mips.c was not synced with changes in aaccoder.c for
some time.

That was cause for some fate-aac tests failing.

This patch fixes the problems.

Optimizations disabled in 933309a are enabled again.

Signed-off-by: Nedeljko Babic <nedeljko.babic@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavcodec/nvenc: merge compute initialQP sections
Agatha Hu [Fri, 11 Sep 2015 11:11:12 +0000 (13:11 +0200)]
avcodec/nvenc: merge compute initialQP sections

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
4 years agoavcodec/nvenc: Optimize nvenc parameters
Agatha Hu [Fri, 11 Sep 2015 09:07:10 +0000 (11:07 +0200)]
avcodec/nvenc: Optimize nvenc parameters

Add 3 more presets: fast, medium, slow.
Improve min/max QP calculation.

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
4 years agoavfilter/avf_showcqt: change fft left-right separation
Muhammad Faiz [Mon, 14 Sep 2015 09:23:58 +0000 (16:23 +0700)]
avfilter/avf_showcqt: change fft left-right separation

This simplifies the code

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavfilter/vsrc_mandelbrot: use the name 's' for the pointer to the private context
Ganesh Ajjanagadde [Sun, 13 Sep 2015 02:32:44 +0000 (22:32 -0400)]
avfilter/vsrc_mandelbrot: use the name 's' for the pointer to the private context

Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
4 years agoavfilter: add audio limiter filter
Paul B Mahol [Sat, 5 Sep 2015 19:12:58 +0000 (19:12 +0000)]
avfilter: add audio limiter filter

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/af_ladspa: support simpler syntax for controls
Paul B Mahol [Sat, 12 Sep 2015 07:31:49 +0000 (07:31 +0000)]
avfilter/af_ladspa: support simpler syntax for controls

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavformat/flvdec: Use the first index entry to find the first packet if there was...
Michael Niedermayer [Sun, 13 Sep 2015 11:53:05 +0000 (13:53 +0200)]
avformat/flvdec: Use the first index entry to find the first packet if there was a parsing error in the header

Fixes: unknow_codec.flv

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavformat/flvdec: Print last packet size at trace level
Michael Niedermayer [Sun, 13 Sep 2015 11:52:01 +0000 (13:52 +0200)]
avformat/flvdec: Print last packet size at trace level

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavformat/flvdec: Print terminator value found if it differs from AMF_END_OF_OBJECT...
Michael Niedermayer [Sun, 13 Sep 2015 10:33:49 +0000 (12:33 +0200)]
avformat/flvdec: Print terminator value found if it differs from AMF_END_OF_OBJECT in AMF_DATA_TYPE_MIXEDARRAY

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavformat/mpjpegdec: Allow mpjpeg headers parsing to succeed upon EOF, if required...
Alex Agranovsky [Sun, 13 Sep 2015 00:34:14 +0000 (20:34 -0400)]
avformat/mpjpegdec: Allow mpjpeg headers parsing to succeed upon EOF, if required headers are present

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoADPCM: Bump THP channel limit to 14
Rodger Combs [Fri, 11 Sep 2015 14:04:23 +0000 (09:04 -0500)]
ADPCM: Bump THP channel limit to 14

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoavfilter/vf_framerate: fix frame leak at uninit
Paul B Mahol [Sat, 12 Sep 2015 02:34:52 +0000 (02:34 +0000)]
avfilter/vf_framerate: fix frame leak at uninit

Reported-by: Andres Noetzli
Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/vf_framerate: unbreak flushing
Paul B Mahol [Fri, 11 Sep 2015 10:09:33 +0000 (10:09 +0000)]
avfilter/vf_framerate: unbreak flushing

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/vf_framerate: calculate delta from scaled pts instead of unscaled
Paul B Mahol [Thu, 10 Sep 2015 15:39:35 +0000 (15:39 +0000)]
avfilter/vf_framerate: calculate delta from scaled pts instead of unscaled

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter/vf_framerate: highbit depth support
Paul B Mahol [Mon, 7 Sep 2015 18:30:22 +0000 (18:30 +0000)]
avfilter/vf_framerate: highbit depth support

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter: add ocr filter
Paul B Mahol [Wed, 9 Sep 2015 13:43:35 +0000 (13:43 +0000)]
avfilter: add ocr filter

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agoavfilter: use ff_all_channel_counts() instead of ff_all_channel_layouts()
Paul B Mahol [Fri, 11 Sep 2015 18:51:10 +0000 (18:51 +0000)]
avfilter: use ff_all_channel_counts() instead of ff_all_channel_layouts()

Fixes playback of some files with ffplay.

Signed-off-by: Paul B Mahol <onemda@gmail.com>
4 years agolavf/img2dec: Improve jpeg auto-detection.
Carl Eugen Hoyos [Sat, 12 Sep 2015 17:00:02 +0000 (19:00 +0200)]
lavf/img2dec: Improve jpeg auto-detection.

Fixes ticket #4841.

4 years agoavformat/mov: use AV_OPT_TYPE_BOOL for export_{all,xmp} options
Clément Bœsch [Fri, 11 Sep 2015 19:20:39 +0000 (21:20 +0200)]
avformat/mov: use AV_OPT_TYPE_BOOL for export_{all,xmp} options

4 years agoavcodec/roqvideoenc: use AV_OPT_TYPE_BOOL for quake3_compat option
Clément Bœsch [Fri, 11 Sep 2015 19:17:57 +0000 (21:17 +0200)]
avcodec/roqvideoenc: use AV_OPT_TYPE_BOOL for quake3_compat option

4 years agoavfilter/treble: use AV_OPT_TYPE_BOOL for csg option
Clément Bœsch [Fri, 11 Sep 2015 19:17:05 +0000 (21:17 +0200)]
avfilter/treble: use AV_OPT_TYPE_BOOL for csg option

4 years agoavcodec/pgssubdec: use AV_OPT_TYPE_BOOL for forced_subs_only option
Clément Bœsch [Fri, 11 Sep 2015 19:16:08 +0000 (21:16 +0200)]
avcodec/pgssubdec: use AV_OPT_TYPE_BOOL for forced_subs_only option

4 years agoavdevice/v4l2: use AV_OPT_TYPE_BOOL for use_libv4l2 option
Clément Bœsch [Fri, 11 Sep 2015 19:14:59 +0000 (21:14 +0200)]
avdevice/v4l2: use AV_OPT_TYPE_BOOL for use_libv4l2 option

4 years agoavcodec/textdec: use AV_OPT_TYPE_BOOL for keep_ass_markup option
Clément Bœsch [Fri, 11 Sep 2015 19:11:36 +0000 (21:11 +0200)]
avcodec/textdec: use AV_OPT_TYPE_BOOL for keep_ass_markup option

4 years agoavcodec/libx264: use AV_OPT_TYPE_BOOL for fastfirstpass option
Clément Bœsch [Fri, 11 Sep 2015 19:10:14 +0000 (21:10 +0200)]
avcodec/libx264: use AV_OPT_TYPE_BOOL for fastfirstpass option

4 years agoavcodec/h264: use AV_OPT_TYPE_BOOL for is_avc option
Clément Bœsch [Fri, 11 Sep 2015 19:08:22 +0000 (21:08 +0200)]
avcodec/h264: use AV_OPT_TYPE_BOOL for is_avc option

4 years agoffprobe: use AV_OPT_TYPE_BOOL for writers options
Clément Bœsch [Fri, 11 Sep 2015 19:01:39 +0000 (21:01 +0200)]
ffprobe: use AV_OPT_TYPE_BOOL for writers options

4 years agoavdevice/caca: small trivial macros adjustments
Clément Bœsch [Fri, 11 Sep 2015 19:06:15 +0000 (21:06 +0200)]
avdevice/caca: small trivial macros adjustments

4 years agoavdevice/caca: remove space before ':' (English typo, cosmetics)
Clément Bœsch [Fri, 11 Sep 2015 19:06:12 +0000 (21:06 +0200)]
avdevice/caca: remove space before ':' (English typo, cosmetics)

4 years agoavdevice/caca: use AV_OPT_TYPE_BOOL for list_drivers option
Clément Bœsch [Fri, 11 Sep 2015 18:47:03 +0000 (20:47 +0200)]
avdevice/caca: use AV_OPT_TYPE_BOOL for list_drivers option

4 years agoavcodec: use AV_OPT_TYPE_BOOL for refcounted_frames, side_data_only_packets and skip_...
Clément Bœsch [Fri, 11 Sep 2015 18:45:15 +0000 (20:45 +0200)]
avcodec: use AV_OPT_TYPE_BOOL for refcounted_frames, side_data_only_packets and skip_alpha options

4 years agoconfigure: fix build on mingw64 target with msys2
Peter Belkner [Sat, 12 Sep 2015 13:23:52 +0000 (15:23 +0200)]
configure: fix build on mingw64 target with msys2

4 years agoMerge commit 'c0a49077ea4ff3a0ad30b9e33f1bb06ba9112aaa'
Hendrik Leppkes [Sat, 12 Sep 2015 13:18:05 +0000 (15:18 +0200)]
Merge commit 'c0a49077ea4ff3a0ad30b9e33f1bb06ba9112aaa'

* commit 'c0a49077ea4ff3a0ad30b9e33f1bb06ba9112aaa':
  asfdec: add more checks for size left in asf packet buffer

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '1b2412f50185447da4212f65f06e8d72a6daf06e'
Hendrik Leppkes [Sat, 12 Sep 2015 13:17:50 +0000 (15:17 +0200)]
Merge commit '1b2412f50185447da4212f65f06e8d72a6daf06e'

* commit '1b2412f50185447da4212f65f06e8d72a6daf06e':
  lavc: Fix compilation with --disable-everything --enable-parser=mpeg4video

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit 'f00f6d538dcbaa122eb5e2784f41f4a299296b7b'
Hendrik Leppkes [Sat, 12 Sep 2015 13:16:57 +0000 (15:16 +0200)]
Merge commit 'f00f6d538dcbaa122eb5e2784f41f4a299296b7b'

* commit 'f00f6d538dcbaa122eb5e2784f41f4a299296b7b':
  lavc: Sanitize header inclusion guards

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '6064f697a321174232a3fad351afb21150c3e9e5'
Hendrik Leppkes [Sat, 12 Sep 2015 13:12:05 +0000 (15:12 +0200)]
Merge commit '6064f697a321174232a3fad351afb21150c3e9e5'

* commit '6064f697a321174232a3fad351afb21150c3e9e5':
  lavc: Enable side data only packets by default

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoavformat/avienc: add muxer option "write_channel_mask"
Tobias Rapp [Thu, 10 Sep 2015 12:04:18 +0000 (14:04 +0200)]
avformat/avienc: add muxer option "write_channel_mask"

Allow writing an empty channel mask into the wave format header. Useful
if the input file contains an unknown channel layout.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
4 years agoasfdec: add more checks for size left in asf packet buffer
Alexandra Hájková [Mon, 7 Sep 2015 10:18:17 +0000 (12:18 +0200)]
asfdec: add more checks for size left in asf packet buffer

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agofate: readjust AAC encoder TNS test
Rostislav Pehlivanov [Sat, 12 Sep 2015 12:17:04 +0000 (13:17 +0100)]
fate: readjust AAC encoder TNS test

The recent commits change the value slightly. Even though it's
within the threshold it's better to risk as little as possible
especially when different systems, processors, FPUs and compilers
are involved.

Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
4 years agoaacenc_tns: readjust values for new TNS decision making
Rostislav Pehlivanov [Sat, 12 Sep 2015 12:15:02 +0000 (13:15 +0100)]
aacenc_tns: readjust values for new TNS decision making

Since TNS was fixed with the recent commits retweak the values
so it's more frequently used.
Still not enabled by default yet, though it's possible that it
will be made enabled by default in the near future.

Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
4 years agoaacenc_tns: encode coefficients directly and reenable compression
Rostislav Pehlivanov [Sat, 12 Sep 2015 12:10:28 +0000 (13:10 +0100)]
aacenc_tns: encode coefficients directly and reenable compression

This commit was made possible with the earlier commits since the
new quantization method basically means we're working always with
unsigned values. The specifications mention to use compression when
the first 2 bits are identical but they didn't mention if this should
happen before or after the conversion to signed values. Actually
they said nothing about conversion to signed values.

With this commit, coefficient compression usually always happens
which saves a lot of space, especially at extremely low bitrates
and doesn't change the quality at all.

Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
4 years agoaacenc_tns: redo coefficient quantization and decision making
Rostislav Pehlivanov [Sat, 12 Sep 2015 12:04:27 +0000 (13:04 +0100)]
aacenc_tns: redo coefficient quantization and decision making

This finally (and again) gets rid of basically everything the
specifications say about how TNS should be done. The main
problem used to be that a single filter was used for all
coefficients which despite being explicitly recommended by
the specifications usually sounds wrong, therefore it's
a corner case in the current TNS implementation.

This commit also changes the coefficient bit size, as apparently
it's better to use lower precision in case the windows are eight
short. This is apparently what fdk_aac uses, looking at the bit
stream and makes sense. Also the order when 8 SHORT windows happen
is important as 7 was too much and according to PSNR was worse
while 5 is just about correct.

Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
4 years agolavc: Fix compilation with --disable-everything --enable-parser=mpeg4video
Hendrik Schreiber [Thu, 10 Sep 2015 11:27:05 +0000 (13:27 +0200)]
lavc: Fix compilation with --disable-everything --enable-parser=mpeg4video

4 years agolavc: Sanitize header inclusion guards
Vittorio Giovara [Thu, 10 Sep 2015 18:26:09 +0000 (20:26 +0200)]
lavc: Sanitize header inclusion guards

Fix typos, add missing comments, and remove stray lines.

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
4 years agolavc: Enable side data only packets by default
Vittorio Giovara [Fri, 11 Sep 2015 15:47:29 +0000 (17:47 +0200)]
lavc: Enable side data only packets by default

Deprecate the now unused option, but temporarily retain the capability
to disable the now default behaviour.

Mention this change in the AVPacket documentation.

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
4 years agoMerge commit '1e0b8bf0b3b3b4247fb21e9839af342ae879607c'
Hendrik Leppkes [Sat, 12 Sep 2015 11:31:22 +0000 (13:31 +0200)]
Merge commit '1e0b8bf0b3b3b4247fb21e9839af342ae879607c'

* commit '1e0b8bf0b3b3b4247fb21e9839af342ae879607c':
  travis: fix dependencies

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '99404597201911de90cff2ef85f2d44176d39147'
Hendrik Leppkes [Sat, 12 Sep 2015 11:30:41 +0000 (13:30 +0200)]
Merge commit '99404597201911de90cff2ef85f2d44176d39147'

* commit '99404597201911de90cff2ef85f2d44176d39147':
  mmaldec: fix pkt_dts determination

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '87a051f97633010f71dfc1d23d806856499bf231'
Hendrik Leppkes [Sat, 12 Sep 2015 11:30:33 +0000 (13:30 +0200)]
Merge commit '87a051f97633010f71dfc1d23d806856499bf231'

* commit '87a051f97633010f71dfc1d23d806856499bf231':
  lavc: allow asynchronous decoders to return correct pkt_dts values

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit 'b7ab6e18eecad43593ad2a0e9fc9eba7f24751cb'
Hendrik Leppkes [Sat, 12 Sep 2015 11:29:29 +0000 (13:29 +0200)]
Merge commit 'b7ab6e18eecad43593ad2a0e9fc9eba7f24751cb'

* commit 'b7ab6e18eecad43593ad2a0e9fc9eba7f24751cb':
  mmaldec: disable timestamp interpolation

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '6b652c0273d79f2e0c52ad91450bd0737cf3c8a6'
Hendrik Leppkes [Sat, 12 Sep 2015 11:29:16 +0000 (13:29 +0200)]
Merge commit '6b652c0273d79f2e0c52ad91450bd0737cf3c8a6'

* commit '6b652c0273d79f2e0c52ad91450bd0737cf3c8a6':
  mmaldec: fix problems with flush logic

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit 'b84675d63aaede8f6944b901250a10456c5477e6'
Hendrik Leppkes [Sat, 12 Sep 2015 11:29:10 +0000 (13:29 +0200)]
Merge commit 'b84675d63aaede8f6944b901250a10456c5477e6'

* commit 'b84675d63aaede8f6944b901250a10456c5477e6':
  mmaldec: hack against buffering problems on broken input

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '5788623d29c3e806a7879210986110aced758dc2'
Hendrik Leppkes [Sat, 12 Sep 2015 11:25:38 +0000 (13:25 +0200)]
Merge commit '5788623d29c3e806a7879210986110aced758dc2'

* commit '5788623d29c3e806a7879210986110aced758dc2':
  jpeg2000: Split codeblock decoding from the main tile decoding

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit 'db53a2306f62f05faa67e6f3c60ee55a9b8e4776'
Hendrik Leppkes [Sat, 12 Sep 2015 11:16:13 +0000 (13:16 +0200)]
Merge commit 'db53a2306f62f05faa67e6f3c60ee55a9b8e4776'

* commit 'db53a2306f62f05faa67e6f3c60ee55a9b8e4776':
  jpeg2000: Do not warn about known and skippable markers

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '9b5a4a9cce3042558e107ae1ed30d9bf3d867a35'
Hendrik Leppkes [Sat, 12 Sep 2015 11:14:26 +0000 (13:14 +0200)]
Merge commit '9b5a4a9cce3042558e107ae1ed30d9bf3d867a35'

* commit '9b5a4a9cce3042558e107ae1ed30d9bf3d867a35':
  mmvideo: Make sure the rle does not write over the frame boundaries

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '41ed749fe987e60b0485fa721ad869590651324d'
Hendrik Leppkes [Sat, 12 Sep 2015 11:05:01 +0000 (13:05 +0200)]
Merge commit '41ed749fe987e60b0485fa721ad869590651324d'

* commit '41ed749fe987e60b0485fa721ad869590651324d':
  ogg: Do not try to use the parser if it is not present

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit '7b02cb29d9d60cdd5ef321043d11d02023e7dc8f'
Hendrik Leppkes [Sat, 12 Sep 2015 11:03:04 +0000 (13:03 +0200)]
Merge commit '7b02cb29d9d60cdd5ef321043d11d02023e7dc8f'

* commit '7b02cb29d9d60cdd5ef321043d11d02023e7dc8f':
  pixdesc: Document the component order

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agovp9: Fix emu[] edge overflow conditions for >8bpp/non-420.
Ronald S. Bultje [Fri, 11 Sep 2015 17:59:17 +0000 (13:59 -0400)]
vp9: Fix emu[] edge overflow conditions for >8bpp/non-420.

4 years agovp9: re-initialize internal buffers on bpp change also.
Ronald S. Bultje [Fri, 11 Sep 2015 17:24:48 +0000 (13:24 -0400)]
vp9: re-initialize internal buffers on bpp change also.

4 years agoMerge commit 'c7ed26ad6035793c7212c17545b0ded18b72db69'
Hendrik Leppkes [Sat, 12 Sep 2015 11:00:31 +0000 (13:00 +0200)]
Merge commit 'c7ed26ad6035793c7212c17545b0ded18b72db69'

* commit 'c7ed26ad6035793c7212c17545b0ded18b72db69':
  pixdesc: Add missing alpha flag for yuva420p9be

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agoMerge commit 'e336c51e6ffcdb93fbcf3c6153d378400608526b'
Hendrik Leppkes [Sat, 12 Sep 2015 10:54:25 +0000 (12:54 +0200)]
Merge commit 'e336c51e6ffcdb93fbcf3c6153d378400608526b'

* commit 'e336c51e6ffcdb93fbcf3c6153d378400608526b':
  pixdesc: Consistently order components

Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
4 years agotravis: fix dependencies
James Almer [Sat, 12 Sep 2015 06:23:45 +0000 (03:23 -0300)]
travis: fix dependencies

bc hasn't been needed since a982c5d74fbc7ff5bd2f2f73af61ae48e9b1bcc6

Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agommaldec: fix pkt_dts determination
wm4 [Tue, 8 Sep 2015 17:42:26 +0000 (19:42 +0200)]
mmaldec: fix pkt_dts determination

This also drops setting the frame->pts field. This is usually not set by
decoders, so this would be an inconsistency that's at worst a danger to
the API user.

It appears the buffer->dts field is normally not set by the MMAL
decoder, so don't use it. If it's ever going to be set by MMAL, we
don't know whether the value will be what we want.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agolavc: allow asynchronous decoders to return correct pkt_dts values
wm4 [Fri, 11 Sep 2015 16:36:56 +0000 (18:36 +0200)]
lavc: allow asynchronous decoders to return correct pkt_dts values

The generic code in utils.c sets the AVFrame.pkt_dts field from the
packet it was supposedly decoded. This does not have to be true for a
fully asynchronous decoder like mmaldec. It could be overwritten with an
incorrect value. Even if the decoder doesn't determine the DTS (but sets
it to AV_NOPTS_VALUE), it's impossible to determine a correct value in
utils.c.

Decoders can now be marked with FF_CODEC_CAP_SETS_PKT_DTS, in which case
utils.c won't overwrite the field. The decoders are expected to set this
field (even if they only set it to AV_NOPTS_VALUE).

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agommaldec: disable timestamp interpolation
wm4 [Tue, 8 Sep 2015 17:42:24 +0000 (19:42 +0200)]
mmaldec: disable timestamp interpolation

This MMAL feature fills in missing timestamps from the framerate set on
the input port. This is generally unwanted, since libavcodec decoders
merely pass through timestamps without ever "fixing" them. The framerate
is also unknown, and even the timebase doesn't have to be set.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agommaldec: fix problems with flush logic
wm4 [Tue, 8 Sep 2015 17:42:23 +0000 (19:42 +0200)]
mmaldec: fix problems with flush logic

Don't try to do a blocking wait for MMAL output if we haven't even sent
a single real packet, but only flush packets. Obviously we can't expect
to get anything back.

Additionally, don't send a flush packet to MMAL in the same case. It
appears the MMAL decoder will sometimes hang in mmal_vc_port_disable()
(called from ffmmal_close_decoder()), waiting for a reply from the GPU
which never arrives. Either MMAL disallows sending flush packets without
preceding real data, or it's a MMAL bug.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agommaldec: hack against buffering problems on broken input
wm4 [Tue, 8 Sep 2015 17:42:22 +0000 (19:42 +0200)]
mmaldec: hack against buffering problems on broken input

I can't come up with a nice way to handle this. It's hard to keep the
lock-stepped input/output in this case. You can't predict whether the
MMAL decoder will output a picture (because it's asynchronous), so
you have to assume in general that any packet could produce 0 or 1
frames. You can't continue to write input packets to the decoder,
because then you might get too many output frames, which you can't
get rid of because the lavc decoding API does not allow the decoder
to return an output frame without consuming an input frame (except
when flushing).

The ideal fix is a M:N decoding API (preferably asynchronous), which
would make this code potentially much cleaner. For now, this hack
will do.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
4 years agox86: port PSIGNW to cpuflags
James Almer [Sat, 12 Sep 2015 01:02:42 +0000 (22:02 -0300)]
x86: port PSIGNW to cpuflags

Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>
4 years agoffserver: remove redundant comment, clarify another one
Reynaldo H. Verdejo Pinochet [Sat, 12 Sep 2015 00:07:52 +0000 (17:07 -0700)]
ffserver: remove redundant comment, clarify another one

Signed-off-by: Reynaldo H. Verdejo Pinochet <reynaldo@osg.samsung.com>