avformat/avienc: factor out avi_add_ientry()

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

lavc/hevc_ps: Fix default display window offsets for yuv422 and yuv444.

lavf/riffenc: Improve spec compliance; Fix WMP playback of AVI with xxpc chunks

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

tests/ref/vsynth: Remove unused file

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

avfilter/vf_waveform: add acolor filter

Useful in combination with color filter.

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/vf_bwdif: add x86 SIMD

Signed-off-by: Thomas Mundt <loudmax@yahoo.de>

lavc/libx265: Support GBR encoding.

lavc/hevc_ps: Support GBR decoding.

avfilter/vf_decimate: Use the correct frame in difference calculation

Fixes Ticket4964

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

Revert "avfilter/vf_decimate: Check that input parameters match"

ill commit a better solution

This reverts commit 30fe3fd52721c8c6566001192cd16be423ffc92b.

avfilter/vf_vectorscope: make it possible to override colorspace

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/smptebars: fix some 75% color values

They where slightly off.

Signed-off-by: Paul B Mahol <onemda@gmail.com>

fate: add missing dependencies for fate-h264-dts_5frames

Signed-off-by: James Almer <jamrial@gmail.com>

fate: add missing h264-dts_5frames ref file

Signed-off-by: James Almer <jamrial@gmail.com>

avformat/hlsenc: Add support for HLS PLAYLIST types EVENT and VOD

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

fate: Add ffprobe dependancy to fate-h264-dts_5frames

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

fate: add test for "Fix DTS for short H264 streams"

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

lavf/utils: Fix DTS for short H264 streams.

Fill DTS if all packets have been read in avformat_find_stream_info, and still
has_decode_delay_been_guessed returns false.

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

avformat/utils: factor update_dts_from_pts() out

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

Use avpriv_tempfile()

Should fix xvid/cache on windows with --enable-shared

May be related to Ticket 4780

Tested-by: Hendrik Leppkes <h.leppkes@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avutil/file: Move av_tempfile() to avutil/file_open ff_tempfile()

document the issue with av_tempfile()

Tested-by: Hendrik Leppkes <h.leppkes@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avcodec/alsdec: fix max bits in ltp prefix code

The maximum number of bits int the prefix code for
p(0) is 4. By setting it as 3, we were missing the
last 0 bit.
This fixes bug #4715 present on the trac.

Signed-off-by: Umair Khan <omerjerk@gmail.com>
Reviewed-by: Thilo Borgmann <thilo.borgmann@mail.de>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

configure: build fix for P5600 along option --disable-msa

Signed-off-by: Shivraj Patil <shivraj.patil@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avfilter/vf_waveform: add parade display mode

Rename old parade display mode to stacked.

Signed-off-by: Paul B Mahol <onemda@gmail.com>

vp9: fix a few signed integer left-shifts.

Fixes trac tickets 5127, 5129, 5130.

vp9_superframe: fix endianness of size markers.

vp9: add superframe merging bitstream filter.

Fixes ticket 4313.

lavf: allow BSFs to drop packets.

If pkt->size == 0 && pkt->side_data_elems == 0 after bsf->filter()
returns, the packet is considered dropped.

avfilter/vf_waveform: fix and extend millivolts grat lines

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/vf_waveform: fix ire8 for 25%

Signed-off-by: Paul B Mahol <onemda@gmail.com>

lavc/lpc: exploit even symmetry of window function

Yields 2x improvement in function performance, and boosts aac encoding
speed by ~ 4% overall. Sample benchmark (Haswell+GCC under -march=native):
after:
ffmpeg -i sin.flac -acodec aac -y sin_new.aac  5.22s user 0.03s system 105% cpu 4.970 total

before:
ffmpeg -i sin.flac -acodec aac -y sin_new.aac  5.40s user 0.05s system 105% cpu 5.162 total

Reviewed-by: Rostislav Pehlivanov <atomnuker@gmail.com>
Signed-off-by: Ganesh Ajjanagadde <gajjanag@gmail.com>

avformat/hlsenc: Fix passing options, regression since bc9a5965c815cf7fd998d8ce14a18b8e861dd9ce

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

avfilter/vf_waveform: fix indentation

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/vf_waveform add support for other scalers for graticule

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avutil/random_seed: Add the runtime in cycles of the main loop to the entropy pool

This should theoretically improve the randomness slightly

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

avutil/channel_layout: AV_CH_LAYOUT_6POINT1_BACK not reachable in parsing

Trying to make heads and tails out of DTS 6.1 I can across this typo.

I also noticed that this wiki page is incorrect or misleading, the
channel order for 6.1 given does not match the source code.  At the
least it should be clarified that the layout given does not apply to
DTS.  https://trac.ffmpeg.org/wiki/AudioChannelManipulation

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

lavf/movenc: Fix help output.

Fixes ticket #5323.

lavc/ffjni: fix uninitialized variable

lavf/mxfdec: Support Canopus codecs.

Fixes ticket #5316.

version.sh: Fix spurious rebuilds.

Regression since 104f8ea8.

avformat/concatdec: set safe mode to enabled instead of auto

This is safer, as a selected demuxer could still mean that it was auto-detected
by a user application

Reviewed-previously-by: Nicolas George <george@nsup.org>
Reviewed-previously-by: Andreas Cadhalpun <andreas.cadhalpun@googlemail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

lavf/avidec: Skip xxpc entries in index; fixes trac #5311

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

Add tests for functions in hash.c

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

mips: add support for R6

Understanding the mips32r6 and mips64r6 ISAs in the configure script is
not enough. In order to have full support for MIPS R6 in FFmpeg we need
to be able to build it, and for that we need to make sure we don't use
incompatible assembler code which makes the build fail. Ifdefing the
offending code is sufficient to fix the problem.

Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

configure: Check for msghdr struct.

Some (Solaris) systems apparently have an incompatible msghdr struct
breaking sctp protocol compilation.

Reported-by: mvelanka

lavf/mpeg: Identify sub-stream ID 0xa1 as mlp.

Fixes ticket #4786.
Auto-detection seems difficult, patch mostly confirmed by
http://dvd-audio.sourceforge.net/spec/aob.shtml

lavc/mjpegdec: Set sar for multiscope videos.

Fixes decoding of the files from ticket #4535 visually.

lavc/hevc_ps: Fix offset for yuv422 and yuv444.

Fixes ticket #4980.
Analyzed-by: kurosu and Hendrik
Reviewed-by: Ronald

Use correct msvc type specifiers for ptrdiff_t and size_t.

avfilter/vf_waveform: use intensity for other components too

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/ffjni: Fix occured typo

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

avcodec/ffjni: Fix ;;

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

aacenc_utils: Use temporary variable.

This ensures gcc does not create unnecessary
loads or stores and possibly even does not vectorize
the negation.
Speeds up mp3 to aac transcoding with default settings
by 10% when using "gcc (Debian 5.3.1-10) 5.3.1 20160224".

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

aacenc: use generational cache instead of resetting.

Approximately 11% faster transcoding from mp3 with
default settings.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

avfilter/vf_waveform: make it possible to draw dots instead of lines

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/vf_waveform: make sure that x/y for text position is positive

Signed-off-by: Paul B Mahol <onemda@gmail.com>

configure: build fix for P5600 with mips code restructuring

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

configure: add check_inline_asm_flags()

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

lavf/mp3dec: avoid printing useless message in default log level

"Skipping 0 bytes of junk" is useless to the user, and essentially
indicates a NOP. At 0 bytes, this message is now pushed back to
the verbose log level.

Signed-off-by: Moritz Barsnick <barsnick@gmx.net>

lavc/mjpegdec: avoid printing useless message in default log level

The change of bps from 0 doesn't contain any info useful to the
user. This message is now at info log level only if the original
value is !=0, otherwise pushed back to debug log level. The
original value is displayed additionally.

Signed-off-by: Moritz Barsnick <barsnick@gmx.net>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

lavf/avienc: Palette changing code only concerns AV_PIX_FMT_PAL8

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

avfilter/avf_showcqt: add performance debugging log

for easier development

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>

avfilter/vf_vectorscope: short for Magenta is Mg

Signed-off-by: Paul B Mahol <onemda@gmail.com>

ffserver: Fixed ffserver to support large ffm files

ffm_read_write_index returns a 64bit value,

Github: Closes #185

avcodec/nvenc: Fix typo and preset error message

avcodec/nvenc: Add encoder stats

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>

fate: Add test similar to ticket 1242

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

avformat/utils: fix dts from pts code in compute_pkt_fields() during ascending delay

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

lavf/riff: Add fourcc GTM4 from Telefactor digital audio for ASP.

avcodec: try to document timebase a bit more

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

avfilter/vf_vectorscope: draw color points names

Signed-off-by: Paul B Mahol <onemda@gmail.com>

lavc/utils: fix extra ASS sanity check in convert_sub_to_old_ass_form()

lavc/mediacodec: fix chroma width for yuv420p

avfilter/vf_waveform: draw graticule for color filter too

Signed-off-by: Paul B Mahol <onemda@gmail.com>

lavf/vplayerdec: support time durations with no ms specified

Example found in the wild:

0:00:03:25.000
0:01:47:A legend is sung
0:01:50:Of when England was young
0:01:53:And knights|were brave and bold
0:01:59:The good king had died

Reported-by: wm4

lavc/mjpegdec: avoid unneeded allocation if the frame is to be skipped

lavfi/extractplanes: Add RGB0 and friends as supported pix_fmts.

lavfi/extractplanes: Fix in_pixfmts.

If the original pix_fmt was >8 bit and not supported by the filter,
the filter system could choose a pix_fmt with different endianness
as input for extractplanes which broke the output because the output
always used the endianness of the original pix_fmt.

lavfi/extractplanes: Move endianness calculation up.

Needed for next commit.

lavc: add h264 mediacodec decoder

lavc: add JNI support

rtp: Fix play multiple multicast streams with the same port

We cannot play multiple multicast streams with the same port at the
same time. This is because both rtp and rtcp port are opened in
read-write mode, so they will not bind to the multicast address. Try
to make rtp port as read-only by default to solve this bug.

Signed-off-by: Zhao Zhili <wantlamy@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

avformat/ftp: Support response code 125 for STOR and RETR commands

This fixes a problem where ffmpeg would hang if there is already an open
data connection, and the server sends a 125 response code in reply to a
STOR or RETR command.

Signed-off-by: Raymond Hilseth <rhi@vizrt.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

lavf/img2dec: Use jpeg constants in jpeg_probe().

lavf/riffdec: Remove \n from avpriv_report_missing_feature().

Found-by: Clément Bœsch

lavc/cfhd: Remove one more \n from avpriv_report_missing_feature().

lavc/cfhd: Sanitize avpriv_report_missing_feature() usage.

avformat/rtpenc: Fix integer overflow in NTP_TO_RTP_FORMAT

RTCP synchronization packet was broken since commit in ffmpeg version > 2.8.3
(commit: e04b039b1528f4c7df5c2b93865651bfea168a19) Since this commit (2e814d0329aded98c811d0502839618f08642685)
"rtpenc: Simplify code by introducing a macro for rescaling NTP timestamps", NTP_TO_RTP_FORMAT
uses av_rescale_rnd() function to add the data to the packet.

This causes an overflow in the av_rescale_rnd() function and it will return INT64_MIN.
Causing the NTP stamp in the RTCP packet to have an invalid value.

Github: Closes #182

Reverting commit '2e814d0329aded98c811d0502839618f08642685' solves the problem.

avfilter/vf_vectorscope: add 12bit depth support

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/vf_histogram: add 12bit depth support

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/dca: clear X96 channels if nothing was decoded

The first X96 channel set can have more channels than core, causing X96
decoding to be skipped. Clear the number of decoded X96 channels to zero
in this rudimentary case.

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

mov - support seek in encrypted mp4

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

avfilter/vf_waveform: add text to graticule

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avfilter/vf_waveform: use nonsubsampled yuv output format for lowpass filter

Signed-off-by: Paul B Mahol <onemda@gmail.com>

avcodec/cfhd: Fixes cfhd_odd.mov which has a resolution of 496x241

In this case container width/height is better however.
Thanks to koda for the sample

avcodec/alsdec: treat quant_cof as a signed value

Signed-off-by: Umair Khan <omerjerk@gmail.com>
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

bitstream.c: improve init_vlc error messages.

Makes it far easier to spot the issue if e.g.
caused by a typo in the code table.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

avformat/avienc: Remove unneeded seekable tests

Reviewed-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

aacenc: avoid double in quantize_bands.

I cannot see any point whatsoever to use
double here instead of float, the results
are likely identical in all cases..
Using float allows for much more
efficient use of SIMD.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

avfilter/avf_showcqt: remove CoeffsType

seems not useful

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>