]> git.sesse.net Git - ffmpeg/log
ffmpeg
8 years agoswresample: add exact_rational option
Muhammad Faiz [Sat, 11 Jun 2016 22:19:20 +0000 (05:19 +0700)]
swresample: add exact_rational option

give high quality resampling
as good as with linear_interp=on
as fast as without linear_interp=on
tested visually with ffplay
ffplay -f lavfi "aevalsrc='sin(10000*t*t)', aresample=osr=48000, showcqt=gamma=5"
ffplay -f lavfi "aevalsrc='sin(10000*t*t)', aresample=osr=48000:linear_interp=on, showcqt=gamma=5"
ffplay -f lavfi "aevalsrc='sin(10000*t*t)', aresample=osr=48000:exact_rational=on, showcqt=gamma=5"

slightly speed improvement
for fair comparison with -cpuflags 0
audio.wav is ~ 1 hour 44100 stereo 16bit wav file
ffmpeg -i audio.wav -af aresample=osr=48000 -f null -
        old         new
real    13.498s     13.121s
user    13.364s     12.987s
sys      0.131s      0.129s

linear_interp=on
        old         new
real    23.035s     23.050s
user    22.907s     22.917s
sys      0.119s     0.125s

exact_rational=on
real    12.418s
user    12.298s
sys      0.114s

possibility to decrease memory usage if soft compensation is ignored

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
8 years agolavu/intmath.h: fix compilation with msvc10.
Matt Oliver [Mon, 6 Jun 2016 07:02:34 +0000 (17:02 +1000)]
lavu/intmath.h: fix compilation with msvc10.

Signed-off-by: Matt Oliver <protogonoi@gmail.com>
8 years agolavf/os_support.h: Fix for unicode filenames on windows.
Matt Oliver [Mon, 6 Jun 2016 07:04:39 +0000 (17:04 +1000)]
lavf/os_support.h: Fix for unicode filenames on windows.

Fixes #819 #5256 #5281

Signed-off-by: Matt Oliver <protogonoi@gmail.com>
8 years agoTimeline function for the eq filter
Ilya87 [Wed, 1 Jun 2016 04:49:51 +0000 (07:49 +0300)]
Timeline function for the eq filter

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/h264_slice: Add casts to suppress warnings
Michael Niedermayer [Sun, 12 Jun 2016 19:28:10 +0000 (21:28 +0200)]
avcodec/h264_slice: Add casts to suppress warnings

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat/redspark: deobfuscate header decrypt code
James Almer [Sun, 12 Jun 2016 20:26:11 +0000 (17:26 -0300)]
avformat/redspark: deobfuscate header decrypt code

Signed-off-by: James Almer <jamrial@gmail.com>
8 years agoavformat/redspark: remove av_malloc usage
James Almer [Sun, 12 Jun 2016 19:30:32 +0000 (16:30 -0300)]
avformat/redspark: remove av_malloc usage

Signed-off-by: James Almer <jamrial@gmail.com>
8 years agolavc/videotoolbox: fix sps/pps mistake introduced in 1534ef87
Clément Bœsch [Sun, 12 Jun 2016 18:45:43 +0000 (20:45 +0200)]
lavc/videotoolbox: fix sps/pps mistake introduced in 1534ef87

8 years agolavc/videotoolbox: fixes compilation after 1534ef87
Clément Bœsch [Sun, 12 Jun 2016 18:36:24 +0000 (20:36 +0200)]
lavc/videotoolbox: fixes compilation after 1534ef87

8 years agolavc/vda_h264: fixes compilation after 1534ef87
Clément Bœsch [Sun, 12 Jun 2016 18:12:07 +0000 (20:12 +0200)]
lavc/vda_h264: fixes compilation after 1534ef87

8 years agoavcodec/iff: Fix bytestream advance
Michael Niedermayer [Sun, 12 Jun 2016 18:13:36 +0000 (20:13 +0200)]
avcodec/iff: Fix bytestream advance

Fixes Ticket5585

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoMerge commit 'c8dcff0cdb17d0aa03ac729eba12d1a20f1f59c8'
Clément Bœsch [Sun, 12 Jun 2016 14:06:58 +0000 (16:06 +0200)]
Merge commit 'c8dcff0cdb17d0aa03ac729eba12d1a20f1f59c8'

* commit 'c8dcff0cdb17d0aa03ac729eba12d1a20f1f59c8':
  h264: factor out calculating the POC count into a separate file

Merged-by: Clément Bœsch <u@pkh.me>
8 years agoMerge commit '113aeee6aed35cb786a9f6d69b0cb210f498b9da'
Clément Bœsch [Sun, 12 Jun 2016 13:37:58 +0000 (15:37 +0200)]
Merge commit '113aeee6aed35cb786a9f6d69b0cb210f498b9da'

* commit '113aeee6aed35cb786a9f6d69b0cb210f498b9da':
  h264_parser: move the H264DSPContext to the parser context

H264Context is kept locally as it is currently needed for logging
(h->avctx as log context) and for is_avc and nal_length_size. These
later fields will later be obtained when the extradata parsing is
decoupled from the decoder. This code will be updated accordingly when
that commit is merged.

Merged-by: Clément Bœsch <u@pkh.me>
8 years agofate: add afade test
Petru Rares Sincraian [Sat, 11 Jun 2016 11:30:09 +0000 (13:30 +0200)]
fate: add afade test

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agolavc/h264: remove unused ff_h264_init_dequant_tables prototype
Clément Bœsch [Sun, 12 Jun 2016 11:38:23 +0000 (13:38 +0200)]
lavc/h264: remove unused ff_h264_init_dequant_tables prototype

Unused since 1534ef87 / 3176217c.

8 years agoMerge commit '3176217c60ca7828712985092d9102d331ea4f3d'
Clément Bœsch [Sun, 12 Jun 2016 11:24:27 +0000 (13:24 +0200)]
Merge commit '3176217c60ca7828712985092d9102d331ea4f3d'

* commit '3176217c60ca7828712985092d9102d331ea4f3d':
  h264: decouple h264_ps from the h264 decoder

Main changes:

- a local GetBitContext is created for the various
  ff_h264_decode_seq_parameter_set() attempts

- just like the old code, remove_sps() is adjusted so it doesn't remove
  the pps.

  Fixes decode with Ticket #631
  http://ffmpeg.org/pipermail/ffmpeg-user/attachments/20111108/dae58f17/attachment.mp4
  but see next point as well.

- ff_h264_update_thread_context() is updated to work even when SPS
  isn't set as it breaks current skip_frame code. This makes sure we
  can still decode the sample from ticket #631 without the need for
  -flags2 +chunks. (Thanks to Michael)

- keep {sps,pps}_ref pointers that stay alive even when the active
  pps/sps get removed from the available lists (patch by michaelni with
  additionnal frees in ff_h264_free_context() from mateo)

- added a check on sps in avpriv_h264_has_num_reorder_frames() to fix
  crashes with mpegts_with_dvbsubs.ts from Ticket #4074
  http://samples.ffmpeg.org/ffmpeg-bugs/trac/ticket4074/mpegts_with_dvbsubs.ts

- in h264_parser.c:h264_parse(), after the ff_h264_decode_extradata() is
  called, the pps and sps from the local parser context are updated with
  the pps and sps from the used h264context. This fixes fate-flv-demux.

- in h264_slice.c, "PPS changed between slices" error is not triggered
  anymore in one condition as it makes fate-h264-xavc-4389 fails with
  THREADS=N (Thanks to Michael)

Merged-by: Clément Bœsch <clement@stupeflix.com>
Merged-by: Michael Niedermayer <michael@niedermayer.cc>
Merged-by: Matthieu Bouron <matthieu.bouron@stupeflix.com>
8 years agoavcodec/sheervideo: add 10-bit interlaced RGB(A) support
Paul B Mahol [Mon, 30 May 2016 17:33:31 +0000 (19:33 +0200)]
avcodec/sheervideo: add 10-bit interlaced RGB(A) support

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/utvideodec: add support for UQRG and UQRA formats
Paul B Mahol [Sat, 11 Jun 2016 21:44:06 +0000 (23:44 +0200)]
avcodec/utvideodec: add support for UQRG and UQRA formats

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agolavu/file_open: Use current directory for temporary files also on Android.
Carl Eugen Hoyos [Sat, 11 Jun 2016 20:41:18 +0000 (22:41 +0200)]
lavu/file_open: Use current directory for temporary files also on Android.

Fixes ticket #5620.

8 years agolavc/libutvideodec: Do not set bits_per_raw_sample.
Carl Eugen Hoyos [Sat, 11 Jun 2016 20:38:41 +0000 (22:38 +0200)]
lavc/libutvideodec: Do not set bits_per_raw_sample.

It is not always 8 but never different from the pix_fmt property.

8 years agoavcodec/utvideodec: fix multiple slices for UQY2 and other issues
Paul B Mahol [Sat, 11 Jun 2016 19:31:49 +0000 (21:31 +0200)]
avcodec/utvideodec: fix multiple slices for UQY2 and other issues

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavfilter/vf_histogram: indent histogram options
Dave Rice [Sat, 11 Jun 2016 04:55:30 +0000 (00:55 -0400)]
avfilter/vf_histogram: indent histogram options

8 years agoavfilter/vf_histogram: shortcuts for histogram options
Dave Rice [Sat, 11 Jun 2016 04:54:59 +0000 (00:54 -0400)]
avfilter/vf_histogram: shortcuts for histogram options

8 years agoavcodec/utvideo: add support for UQY2
Paul B Mahol [Sat, 11 Jun 2016 16:17:41 +0000 (18:17 +0200)]
avcodec/utvideo: add support for UQY2

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/mpc8: Correct end truncation
Michael Niedermayer [Sat, 11 Jun 2016 12:28:24 +0000 (14:28 +0200)]
avcodec/mpc8: Correct end truncation

Fixes Ticket5478

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agolavc/videotoolboxenc: set extradata when opening codec
Rick Kern [Sun, 29 May 2016 02:32:55 +0000 (22:32 -0400)]
lavc/videotoolboxenc: set extradata when opening codec

VideoToolbox doesn't supply parameter sets until the first frame is done
encoding. This spins up a temporary encoder and encodes a single frame to
get this data.

Signed-off-by: Rick Kern <kernrj@gmail.com>
8 years agolavc/videotoolboxenc: fix allocated buffer size
Rick Kern [Sun, 29 May 2016 02:28:04 +0000 (22:28 -0400)]
lavc/videotoolboxenc: fix allocated buffer size

Allocates a buffer with padding, and with enough room for start codes when
length codes are smaller.

Signed-off-by: Rick Kern <kernrj@gmail.com>
8 years agolibavformat/flvenc: support for codec configuration change mid stream
Ivan [Fri, 10 Jun 2016 02:12:10 +0000 (22:12 -0400)]
libavformat/flvenc: support for codec configuration change mid stream

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agolibavformat/flvenc: refactoring: extracted method for writing codec headers
Ivan [Fri, 10 Jun 2016 02:12:09 +0000 (22:12 -0400)]
libavformat/flvenc: refactoring: extracted method for writing codec headers

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat/mp3dec: Increase probe score slightly when the whole data from begin to...
Michael Niedermayer [Sat, 11 Jun 2016 00:41:45 +0000 (02:41 +0200)]
avformat/mp3dec: Increase probe score slightly when the whole data from begin to end is mp3

Improves score for 1000-frames-of-noise-encoded-with-lame.mp3 without file extension

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agovaapi_encode_h26[45]: Reject bitrate targets higher than 2^31
Mark Thompson [Thu, 2 Jun 2016 18:06:08 +0000 (19:06 +0100)]
vaapi_encode_h26[45]: Reject bitrate targets higher than 2^31

8 years agoavcodec/cfhd: Set dimensions unconditionally
Michael Niedermayer [Fri, 10 Jun 2016 18:22:02 +0000 (20:22 +0200)]
avcodec/cfhd: Set dimensions unconditionally

Fixes Ticket5215

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/sheervideo: add support for 10-bit interlaced YCbCr(A) 4:2:2
Paul B Mahol [Fri, 10 Jun 2016 14:09:12 +0000 (16:09 +0200)]
avcodec/sheervideo: add support for 10-bit interlaced YCbCr(A) 4:2:2

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/sheervideo: fix argx format support
Paul B Mahol [Fri, 10 Jun 2016 11:04:29 +0000 (13:04 +0200)]
avcodec/sheervideo: fix argx format support

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoswscale: add input support for gbrap10 pixel format
Paul B Mahol [Fri, 10 Jun 2016 11:01:26 +0000 (13:01 +0200)]
swscale: add input support for gbrap10 pixel format

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavutil: add 10-bit planar RGB with alpha
Paul B Mahol [Fri, 10 Jun 2016 10:48:09 +0000 (12:48 +0200)]
avutil: add 10-bit planar RGB with alpha

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoffmpeg: Add cuvid hwaccel support
Timo Rothenpieler [Sat, 4 Jun 2016 17:07:21 +0000 (19:07 +0200)]
ffmpeg: Add cuvid hwaccel support

8 years agoavcodec/cuvid: add cuvid decoder
Timo Rothenpieler [Sat, 4 Jun 2016 17:06:52 +0000 (19:06 +0200)]
avcodec/cuvid: add cuvid decoder

8 years agoavcodec/mpegvideo: Do not clear the parse context during init
Michael Niedermayer [Fri, 10 Jun 2016 13:38:26 +0000 (15:38 +0200)]
avcodec/mpegvideo: Do not clear the parse context during init

It is allocated before, this cannot work
Fixes Ticket5613

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoswscale: fix ring buffer size when scaling slices of a frame
Pedro Arthur [Thu, 9 Jun 2016 00:38:49 +0000 (21:38 -0300)]
swscale: fix ring buffer size when scaling slices of a frame

The ring buffer size should be able to store input lines
when there is not enough lines to output a single line.

8 years agoswscale: fix crash with swscale-test when using slices
Pedro Arthur [Thu, 9 Jun 2016 00:36:16 +0000 (21:36 -0300)]
swscale: fix crash with swscale-test when using slices

8 years agoavcodec/sheervideo: add support for 10-bit interlaced YCbCr(A) 4:4:4(:4)
Paul B Mahol [Mon, 30 May 2016 17:33:31 +0000 (19:33 +0200)]
avcodec/sheervideo: add support for 10-bit interlaced YCbCr(A) 4:4:4(:4)

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoMAINTAINERs cleanup (remove myself from things i de facto dont maintain)
Michael Niedermayer [Fri, 10 Jun 2016 07:59:21 +0000 (09:59 +0200)]
MAINTAINERs cleanup (remove myself from things i de facto dont maintain)

x86 is maintained entirely by others these days
ML, mostly too

remove myself from a few spots that have other maintainers and where i
just dont know the code that well anyway to do an ideal job

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/sheervideo: fix prediction for ybyr format
Paul B Mahol [Fri, 10 Jun 2016 07:52:38 +0000 (09:52 +0200)]
avcodec/sheervideo: fix prediction for ybyr format

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/sheervideo: fix predictions for c82p format
Paul B Mahol [Thu, 9 Jun 2016 22:24:30 +0000 (00:24 +0200)]
avcodec/sheervideo: fix predictions for c82p format

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavfilter/af_loudnorm: add dual_mono option
Kyle Swanson [Tue, 7 Jun 2016 16:55:02 +0000 (11:55 -0500)]
avfilter/af_loudnorm: add dual_mono option

Signed-off-by: Kyle Swanson <k@ylo.ph>
8 years agoavformat/utils: Do not overwrite, but use sample_fmt from context
Michael Niedermayer [Thu, 9 Jun 2016 09:22:01 +0000 (11:22 +0200)]
avformat/utils: Do not overwrite, but use sample_fmt from context

Fixes Ticket 3759

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat/utils: Open decoder even if there are no packets if parameters are missing
Michael Niedermayer [Thu, 9 Jun 2016 17:11:18 +0000 (19:11 +0200)]
avformat/utils: Open decoder even if there are no packets if parameters are missing

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat/utils: Initialize st in loop
Michael Niedermayer [Thu, 9 Jun 2016 17:09:36 +0000 (19:09 +0200)]
avformat/utils: Initialize st in loop

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/sheervideo: supports some other 8bit formats
Paul B Mahol [Thu, 9 Jun 2016 13:30:27 +0000 (15:30 +0200)]
avcodec/sheervideo: supports some other 8bit formats

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/sheervideo: add interlaced YCbCr(A) 4:2:2:4 8-bit support
Paul B Mahol [Mon, 30 May 2016 17:33:31 +0000 (19:33 +0200)]
avcodec/sheervideo: add interlaced YCbCr(A) 4:2:2:4 8-bit support

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavutil/opt: Fix max/min checking in av_opt_set() for AV_OPT_TYPE_VIDEO_RATE
Michael Niedermayer [Thu, 9 Jun 2016 03:01:26 +0000 (05:01 +0200)]
avutil/opt: Fix max/min checking in av_opt_set() for AV_OPT_TYPE_VIDEO_RATE

Fixes Ticket5618

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat: Fix max value of AV_OPT_TYPE_VIDEO_RATE
Michael Niedermayer [Thu, 9 Jun 2016 03:00:52 +0000 (05:00 +0200)]
avformat: Fix max value of AV_OPT_TYPE_VIDEO_RATE

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavfilter: Fix max value of AV_OPT_TYPE_VIDEO_RATE
Michael Niedermayer [Thu, 9 Jun 2016 03:00:22 +0000 (05:00 +0200)]
avfilter: Fix max value of AV_OPT_TYPE_VIDEO_RATE

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavdevice: Fix max value of AV_OPT_TYPE_VIDEO_RATE
Michael Niedermayer [Thu, 9 Jun 2016 02:52:52 +0000 (04:52 +0200)]
avdevice: Fix max value of AV_OPT_TYPE_VIDEO_RATE

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/sheervideo: add interlaced YCbCr(A) 4:4:4:4 8-bit support
Paul B Mahol [Mon, 30 May 2016 17:33:31 +0000 (19:33 +0200)]
avcodec/sheervideo: add interlaced YCbCr(A) 4:4:4:4 8-bit support

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agox86/showcqt: use three operand format for some instructions
James Almer [Wed, 8 Jun 2016 22:37:08 +0000 (19:37 -0300)]
x86/showcqt: use three operand format for some instructions

Fixes failures with yasm 1.1.0 and older

Signed-off-by: James Almer <jamrial@gmail.com>
8 years agox86/showcqt: add missing preprocessor checks
James Almer [Wed, 8 Jun 2016 22:30:22 +0000 (19:30 -0300)]
x86/showcqt: add missing preprocessor checks

Old yasm/nasm versions don't support some of these

Signed-off-by: James Almer <jamrial@gmail.com>
8 years agoaacsbr: reduce element type mismatch warning severity
Rostislav Pehlivanov [Wed, 8 Jun 2016 20:01:04 +0000 (21:01 +0100)]
aacsbr: reduce element type mismatch warning severity

All HE-AAC samples with an LFE channel make this warning get spammed on
every frame. Turning off SBR for LFE channels makes sense (since it has
much less coefficients than normal channels do), so this error print is
of no value in this case.
It makes sense to keep the error in other cases, hence why it's still
around, degraded to warning severity since the decoder will still
attempt to decode without SBR.

Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
8 years agoavfilter/src_movie: call open_stream after guess_channel_layout
Muhammad Faiz [Thu, 2 Jun 2016 07:50:46 +0000 (14:50 +0700)]
avfilter/src_movie: call open_stream after guess_channel_layout

fix error 'Channel layout change is not supported' when
opening wav file

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
8 years agoavfilter/avf_showcqt: set range on fps/rate/r option
Muhammad Faiz [Wed, 8 Jun 2016 00:44:03 +0000 (07:44 +0700)]
avfilter/avf_showcqt: set range on fps/rate/r option

see Ticket #5618

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
8 years agox86/aacdec: use HADDPS macro
James Almer [Wed, 8 Jun 2016 17:18:18 +0000 (14:18 -0300)]
x86/aacdec: use HADDPS macro

Signed-off-by: James Almer <jamrial@gmail.com>
8 years agoavutil/x86util: move haddps sse emulation from showcqt
James Almer [Wed, 8 Jun 2016 17:18:00 +0000 (14:18 -0300)]
avutil/x86util: move haddps sse emulation from showcqt

Signed-off-by: James Almer <jamrial@gmail.com>
8 years agoMAINTAINERS: Add myself as maintainer for VAAPI encoders
Mark Thompson [Mon, 6 Jun 2016 21:10:03 +0000 (22:10 +0100)]
MAINTAINERS: Add myself as maintainer for VAAPI encoders

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/h264: Put context_count check back
Michael Niedermayer [Wed, 8 Jun 2016 10:32:57 +0000 (12:32 +0200)]
avcodec/h264: Put context_count check back

Fixes assertion failure

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavfilter/avf_showcqt: cqt_calc optimization on x86
Muhammad Faiz [Sat, 4 Jun 2016 07:33:05 +0000 (14:33 +0700)]
avfilter/avf_showcqt: cqt_calc optimization on x86

on x86_64:
        time    PSNR
plain   3.303   inf
SSE     1.649   107.087535
SSE3    1.632   107.087535
AVX     1.409   106.986771
FMA3    1.265   107.108437

on x86_32 (PSNR compared to x86_64 plain):
        time    PSNR
plain   7.225   103.951979
SSE     1.827   105.859282
SSE3    1.819   105.859282
AVX     1.533   105.997661
FMA3    1.384   105.885377

FMA4 test is not available

Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
8 years agoavformat/matroskadec: force 48kHz sample rate when rescaling Opus inital padding
James Almer [Mon, 6 Jun 2016 04:14:11 +0000 (01:14 -0300)]
avformat/matroskadec: force 48kHz sample rate when rescaling Opus inital padding

Mkvtoolnix stores the sample rate of the original stream as reported by the
"OpusHead" stream header instead of 48kHz, the actual sample rate of the Opus
stream.
Ignoring the stored sample rate and forcing 48kHz preserves the correct initial
padding when remuxing such files.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>
8 years agoavcodec/h264: Fix off by 1 context count
Michael Niedermayer [Tue, 7 Jun 2016 20:17:38 +0000 (22:17 +0200)]
avcodec/h264: Fix off by 1 context count

Fixes fate-h264-xavc-4389 with slice threads

8 years agoavcodec/sheervideo: fix order of operations for raw lines
Paul B Mahol [Tue, 7 Jun 2016 21:14:07 +0000 (23:14 +0200)]
avcodec/sheervideo: fix order of operations for raw lines

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/sheervideo: add support for 8-bit interlaced (A)RGB
Paul B Mahol [Mon, 30 May 2016 17:33:31 +0000 (19:33 +0200)]
avcodec/sheervideo: add support for 8-bit interlaced (A)RGB

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec: add BitJazz SheerVideo decoder
Paul B Mahol [Mon, 30 May 2016 17:33:31 +0000 (19:33 +0200)]
avcodec: add BitJazz SheerVideo decoder

Signed-off-by: Paul B Mahol <onemda@gmail.com>
8 years agoavcodec/alsdec: Check r to prevent out of array read
Michael Niedermayer [Tue, 7 Jun 2016 18:50:38 +0000 (20:50 +0200)]
avcodec/alsdec: Check r to prevent out of array read

No testcase known

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agodoc/filters: remove outdated showwavespic example
Lou Logan [Tue, 7 Jun 2016 18:52:15 +0000 (10:52 -0800)]
doc/filters: remove outdated showwavespic example

Use "colors" option instead.

Signed-off-by: Lou Logan <lou@lrcd.com>
8 years agoavformat/mpegts: Fix probing of mpegts with invalid ASC
Michael Niedermayer [Tue, 7 Jun 2016 13:46:08 +0000 (15:46 +0200)]
avformat/mpegts: Fix probing of mpegts with invalid ASC

Fixes Ticket5566

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat/rawdec: Fix avg_framerate for h264
Michael Niedermayer [Fri, 3 Jun 2016 21:59:36 +0000 (23:59 +0200)]
avformat/rawdec: Fix avg_framerate for h264

The framerate is 25 which is a fixed default and is wrong undo the 1 line
change which caused this regression
Only the avg_frame rate setting is removed
The timebase update is not done as there was a objection (see ML)
Fixes Ticket 5444

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agolavc/mediacodec: improve error messages
Matthieu Bouron [Wed, 1 Jun 2016 10:05:49 +0000 (12:05 +0200)]
lavc/mediacodec: improve error messages

8 years agolavc/mediacodec: bypass width/height restrictions when looking for a decoder
Matthieu Bouron [Tue, 31 May 2016 09:00:12 +0000 (11:00 +0200)]
lavc/mediacodec: bypass width/height restrictions when looking for a decoder

Codec width/height restrictions seem hardcoded at the OMX level and
seem arbitrary. Bypassing those restrictions allows a device to decode
streams at higher resolutions.

For example it allows a Nexus 5 to decode h264 streams with a resolution
higher than 1920x1080.

8 years agolavc/mediacodec: rename MediaFormat field clazz to mediaformat_class for consistency
Matthieu Bouron [Mon, 6 Jun 2016 14:05:46 +0000 (16:05 +0200)]
lavc/mediacodec: rename MediaFormat field clazz to mediaformat_class for consistency

8 years agolavc/mediacodec: move struct declarations at the begin
Matthieu Bouron [Mon, 6 Jun 2016 13:56:10 +0000 (15:56 +0200)]
lavc/mediacodec: move struct declarations at the begin

8 years agolavc/mediacodec: rename jfields_mapping to jni_amediacodeclist_mapping
Matthieu Bouron [Mon, 6 Jun 2016 13:28:36 +0000 (15:28 +0200)]
lavc/mediacodec: rename jfields_mapping to jni_amediacodeclist_mapping

8 years agolavc/mediacodec: do not delete a local reference twice in case of error
Matthieu Bouron [Tue, 31 May 2016 09:37:11 +0000 (11:37 +0200)]
lavc/mediacodec: do not delete a local reference twice in case of error

8 years agoavcodec/utils: check skip_samples signedness
Michael Niedermayer [Tue, 7 Jun 2016 01:48:09 +0000 (03:48 +0200)]
avcodec/utils: check skip_samples signedness

Fixes Ticket5528

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavfilter/af_firequalizer: add zero_phase option
Muhammad Faiz [Sun, 5 Jun 2016 23:28:09 +0000 (06:28 +0700)]
avfilter/af_firequalizer: add zero_phase option

simply by substracting pts to compensate delay
also handle AV_NOPTS_VALUE

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
8 years agoavdevice/dshow: Fix frame rate
Michael Niedermayer [Tue, 7 Jun 2016 00:32:13 +0000 (02:32 +0200)]
avdevice/dshow: Fix frame rate

Fixes: Ticket5526
Tested-by: kgp700
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoChangelog : add OpenExr improvements
Martin Vignali [Mon, 6 Jun 2016 20:48:24 +0000 (22:48 +0200)]
Changelog : add OpenExr improvements

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/nvenc: Check capabilities for interlaced encoding
Timo Rothenpieler [Mon, 6 Jun 2016 19:19:29 +0000 (21:19 +0200)]
avcodec/nvenc: Check capabilities for interlaced encoding

8 years agoavformat/mpegts: Do not trust BSSD descriptor, it is sometimes not an S302M stream
Michael Niedermayer [Mon, 6 Jun 2016 19:43:01 +0000 (21:43 +0200)]
avformat/mpegts: Do not trust BSSD descriptor, it is sometimes not an S302M stream

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/exr: indent b44 uncompress function
Martin Vignali [Sat, 4 Jun 2016 13:47:53 +0000 (15:47 +0200)]
avcodec/exr: indent b44 uncompress function

8 years agoavcodec/exr: fix decoding of B44 exr when all channel doesnt have the same pixel...
Martin Vignali [Sat, 4 Jun 2016 13:43:48 +0000 (15:43 +0200)]
avcodec/exr: fix decoding of B44 exr when all channel doesnt have the same pixel type

8 years agoavcodec/exr: move channel_line_size to thread data
Martin Vignali [Sat, 4 Jun 2016 13:38:59 +0000 (15:38 +0200)]
avcodec/exr: move channel_line_size to thread data

In order to be used by b44 uncompress.

8 years agoavcodec/exr: improve pxr24 uncompress
Martin Vignali [Sat, 4 Jun 2016 12:24:08 +0000 (14:24 +0200)]
avcodec/exr: improve pxr24 uncompress

Fix pxr24 uncompress when all channels doesnt have the same pixel type.
The expected length after zip decoding, is now calculated channel by channel.

8 years agoavcodec/exr: remove unneed scanline_size var
Martin Vignali [Sat, 4 Jun 2016 12:18:48 +0000 (14:18 +0200)]
avcodec/exr: remove unneed scanline_size var

8 years agoavcodec/exr: fix tile decoding when all channels doesnt have the same pixel type
Martin Vignali [Sat, 4 Jun 2016 12:06:39 +0000 (14:06 +0200)]
avcodec/exr: fix tile decoding when all channels doesnt have the same pixel type

Also simplify the tile code, sharing more code with the scanline mode.

8 years agoavcodec/exr: indent the if (layer_match) part
Martin Vignali [Sat, 4 Jun 2016 15:45:31 +0000 (17:45 +0200)]
avcodec/exr: indent the if (layer_match) part

8 years agoavcodec/exr: fix layer detection
Martin Vignali [Sat, 4 Jun 2016 15:42:20 +0000 (17:42 +0200)]
avcodec/exr: fix layer detection

Only test a channel if the layer name match.
Avoid to try to mix channel between the main layer (rgba layer),
and the layer request by the user.

8 years agoavcodec/bmp_parser: Check fsize
Michael Niedermayer [Mon, 6 Jun 2016 12:32:48 +0000 (14:32 +0200)]
avcodec/bmp_parser: Check fsize

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/bmp_parser: reset state
Michael Niedermayer [Mon, 6 Jun 2016 12:30:40 +0000 (14:30 +0200)]
avcodec/bmp_parser: reset state

Fixes part of ticket 5598

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/bmp_parser: Fix remaining size
Michael Niedermayer [Mon, 6 Jun 2016 12:27:20 +0000 (14:27 +0200)]
avcodec/bmp_parser: Fix remaining size

Fixes part of ticket 5598

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavcodec/bmp_parser: Fix frame_start_found in cross frame cases
Michael Niedermayer [Mon, 6 Jun 2016 12:01:43 +0000 (14:01 +0200)]
avcodec/bmp_parser: Fix frame_start_found in cross frame cases

Fixes part of ticket 5598

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
8 years agoavformat/matroskadec: Fix rounding error with codec_delay
Michael Niedermayer [Mon, 6 Jun 2016 02:23:16 +0000 (04:23 +0200)]
avformat/matroskadec: Fix rounding error with codec_delay

Fixes Ticket5509

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>