platform/upstream/libav.git
10 years agolavf: add AVFMT_FLAG_BITEXACT.
Anton Khirnov [Thu, 1 May 2014 08:43:10 +0000 (10:43 +0200)]
lavf: add AVFMT_FLAG_BITEXACT.

Use it instead of checking CODEC_FLAG_BITEXACT in the first stream's
codec context.

Using codec options inside lavf is fragile and can easily break when the
muxing codec context is not the encoding context.

10 years agolavc: add a native Opus decoder.
Anton Khirnov [Thu, 17 Apr 2014 10:51:03 +0000 (12:51 +0200)]
lavc: add a native Opus decoder.

Initial implementation by Andrew D'Addesio <modchipv12@gmail.com> during
GSoC 2012.

Completion by Anton Khirnov <anton@khirnov.net>, sponsored by the
Mozilla Corporation.

Further contributions by:
Christophe Gisquet <christophe.gisquet@gmail.com>
Janne Grunau <janne-libav@jannau.net>
Luca Barbato <lu_zero@gentoo.org>

10 years agobuild: do not append $(FFLIBS-) to $(FFLIBS) if $(NAME) is not defined
Janne Grunau [Tue, 13 May 2014 22:05:27 +0000 (00:05 +0200)]
build: do not append $(FFLIBS-) to $(FFLIBS) if $(NAME) is not defined

Avoids including disabled library Makefiles.

10 years agolavf: properly document the distinction between flags and ctx_flags
Anton Khirnov [Thu, 1 May 2014 08:03:54 +0000 (10:03 +0200)]
lavf: properly document the distinction between flags and ctx_flags

10 years agobuild: handle library dependencies in configure
Janne Grunau [Thu, 1 May 2014 11:18:57 +0000 (13:18 +0200)]
build: handle library dependencies in configure

Instead of setting FFLIBS in each library Makefile configure
exports FFLIBS-$library in config.mak.

10 years agoaarch64: assembler in clang-3.4 ignores the division by two
Janne Grunau [Tue, 13 May 2014 13:47:51 +0000 (15:47 +0200)]
aarch64: assembler in clang-3.4 ignores the division by two

Values are positive powers of two, so just replace it with right shift.

10 years agoavconv: Support VDA hwaccel
Anton Khirnov [Mon, 24 Mar 2014 01:02:33 +0000 (02:02 +0100)]
avconv: Support VDA hwaccel

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agolavc: Add new VDA hwaccel
Anton Khirnov [Mon, 24 Mar 2014 00:59:02 +0000 (01:59 +0100)]
lavc: Add new VDA hwaccel

It leverages the new hwaccel 1.2 features:

- get_buffer2 is never called
- the internal context is automatically initialized/deinitialized

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agovda: use hwaccel private data for internal bitstream buffer
Anton Khirnov [Thu, 6 Mar 2014 18:02:29 +0000 (19:02 +0100)]
vda: use hwaccel private data for internal bitstream buffer

10 years agolavu: Add a pixel format for new VDA hwaccel
Anton Khirnov [Thu, 6 Mar 2014 13:07:42 +0000 (14:07 +0100)]
lavu: Add a pixel format for new VDA hwaccel

The current hwaccel is broken and cannot be fixed in a compatible
way. It will be deprecated and replaced with a new one.

10 years agolavc: Add hwaccel private data and init/uninit callbacks
Anton Khirnov [Thu, 6 Mar 2014 17:37:42 +0000 (18:37 +0100)]
lavc: Add hwaccel private data and init/uninit callbacks

10 years agohwaccel: Support specific frame allocators
Luca Barbato [Thu, 8 May 2014 12:43:23 +0000 (14:43 +0200)]
hwaccel: Support specific frame allocators

It would reduce the boilerplate code users have to write.

10 years agohwaccel: Rename priv_data_size to frame_priv_data_size
Anton Khirnov [Thu, 6 Mar 2014 17:19:52 +0000 (18:19 +0100)]
hwaccel: Rename priv_data_size to frame_priv_data_size

This describes more accurately what this field is for.

10 years agolavc: document which parts of AVHWAccel are private.
Anton Khirnov [Thu, 6 Mar 2014 17:15:31 +0000 (18:15 +0100)]
lavc: document which parts of AVHWAccel are private.

10 years agolavc: set AVCodecContext.hwaccel in ff_get_format()
Anton Khirnov [Thu, 6 Mar 2014 17:01:05 +0000 (18:01 +0100)]
lavc: set AVCodecContext.hwaccel in ff_get_format()

This way each decoder does not have to do the same thing manually.

10 years agolavc: Add an internal wrapper around get_format()
Anton Khirnov [Thu, 6 Mar 2014 16:48:18 +0000 (17:48 +0100)]
lavc: Add an internal wrapper around get_format()

It will be useful in the following commits.

10 years agopcm-dvd: Fix 20bit decoding
nu774 [Fri, 9 May 2014 12:47:41 +0000 (21:47 +0900)]
pcm-dvd: Fix 20bit decoding

Increment the pointer as needed.

Bug-Id: 592

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agortmpproto: Check the buffer sizes when copying app/playpath strings
Martin Storsjö [Thu, 8 May 2014 12:12:23 +0000 (15:12 +0300)]
rtmpproto: Check the buffer sizes when copying app/playpath strings

As pointed out by Reimar Döffinger.

CC: libav-stable@libav.org
Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agortmpproto: Support alternative slist parameter in rtmp URLs
Uwe L. Korn [Mon, 5 May 2014 20:47:05 +0000 (21:47 +0100)]
rtmpproto: Support alternative slist parameter in rtmp URLs

Support the URL scheme where the playpath is in an RTMP URL is
passed as the slist argument and the app is given infront of the
query part of the URL:

rtmp://host[:port]/[app]?slist=[playpath]

(other arguments in the query part are stripped as they are not used)

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agolibfdk-aac: Relicense the library wrappers to 2-clause BSD
Martin Storsjö [Wed, 30 Apr 2014 06:44:31 +0000 (09:44 +0300)]
libfdk-aac: Relicense the library wrappers to 2-clause BSD

This should make it possible for Fraunhofer to contribute to these
wrappers - they didn't want to contribute to code under LGPL2.1 with
the "or any later version" clause (which allowed using the code
under the LGPL3 license).

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agolibfdk-aacenc: Enable 7.1 channel encoding
Jean First [Sun, 26 Jan 2014 00:31:53 +0000 (01:31 +0100)]
libfdk-aacenc: Enable 7.1 channel encoding

7.1(wide) and 7.1(wide-side) channel layouts are supported in
fdk-aac since the 0.1.3 release.

The earlier versions of fdk-aac didn't include any library
version defines in the public headers, thus checking for
the AACENCODER_LIB_VL0 define is enough to know that we're
building against a new enough version of fdk-aac.

This change includes contributions by Tim Walker,
Michael Niedermayer and Timothy Gu.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agoavconv: rename OutputStream.opts to OutputStream.encoder_opts
Anton Khirnov [Fri, 4 Apr 2014 17:38:04 +0000 (19:38 +0200)]
avconv: rename OutputStream.opts to OutputStream.encoder_opts

This makes it more clear what is this variable for exactly.

10 years agoavconv: remove a useless line
Anton Khirnov [Sun, 6 Apr 2014 05:28:51 +0000 (07:28 +0200)]
avconv: remove a useless line

Doing anything with that field when we are doing streamcopy (so the
encoder is not opened) should have absolutely no effect.

10 years agoavconv: rename 'codec' to 'enc_ctx'
Anton Khirnov [Fri, 4 Apr 2014 17:59:48 +0000 (19:59 +0200)]
avconv: rename 'codec' to 'enc_ctx'

This more clearly describes what is the variable used for.

Also, move its declaration into the block where it is actually used.

10 years agoavconv: rename 'icodec' to 'dec_ctx'
Anton Khirnov [Fri, 4 Apr 2014 17:59:48 +0000 (19:59 +0200)]
avconv: rename 'icodec' to 'dec_ctx'

This more clearly describes what is the variable used for.

10 years agoavconv: rename InputStream.opts to InputStream.decoder_opts
Anton Khirnov [Fri, 4 Apr 2014 17:38:04 +0000 (19:38 +0200)]
avconv: rename InputStream.opts to InputStream.decoder_opts

This makes it more clear what is this variable for exactly.

10 years agoavconv: simplify exit_program() by using more local vars
Anton Khirnov [Tue, 25 Mar 2014 11:38:31 +0000 (12:38 +0100)]
avconv: simplify exit_program() by using more local vars

10 years agoprobe: Support AAC_LATM in set_codec_from_probe_data
Luca Barbato [Sat, 26 Apr 2014 19:10:30 +0000 (21:10 +0200)]
probe: Support AAC_LATM in set_codec_from_probe_data

10 years agolatm: Add a probe function
Luca Barbato [Sat, 26 Apr 2014 18:32:47 +0000 (20:32 +0200)]
latm: Add a probe function

10 years agomathematics: K&R formatting cosmetics
Luca Barbato [Thu, 1 May 2014 12:54:06 +0000 (14:54 +0200)]
mathematics: K&R formatting cosmetics

10 years agoswscale: Fix an undefined behaviour
Luca Barbato [Thu, 1 May 2014 22:21:23 +0000 (00:21 +0200)]
swscale: Fix an undefined behaviour

Prevent a division by zero down the codepath.

Sample-Id: 00001721-google
Reported-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
10 years agovp9: Read the frame size as unsigned
Luca Barbato [Thu, 1 May 2014 22:18:04 +0000 (00:18 +0200)]
vp9: Read the frame size as unsigned

Sample-Id: 00001723-google
Reported-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
10 years agoaac: K&R formatting cosmetics
Luca Barbato [Sat, 26 Apr 2014 17:02:36 +0000 (19:02 +0200)]
aac: K&R formatting cosmetics

10 years agoflv: K&R formatting cosmetics
Luca Barbato [Sat, 26 Apr 2014 12:09:55 +0000 (14:09 +0200)]
flv: K&R formatting cosmetics

10 years agolavc: deprecate AVCodecContext.codec_name
Anton Khirnov [Tue, 29 Apr 2014 14:44:08 +0000 (16:44 +0200)]
lavc: deprecate AVCodecContext.codec_name

It is undocumented and has no real use.

10 years agolavc: do not use AVCodecContext.codec_name in avcodec_string()
Anton Khirnov [Tue, 29 Apr 2014 14:42:07 +0000 (16:42 +0200)]
lavc: do not use AVCodecContext.codec_name in avcodec_string()

That field will be deprecated.

10 years agormdec: do not export anything to AVCodecContext.codec_name
Anton Khirnov [Tue, 29 Apr 2014 14:40:55 +0000 (16:40 +0200)]
rmdec: do not export anything to AVCodecContext.codec_name

That field will be deprecated and the value that is written there is not
particularly useful.

10 years agoasfenc: use codec descriptors instead of AVCodecs to write codec info
Anton Khirnov [Tue, 29 Apr 2014 14:13:38 +0000 (16:13 +0200)]
asfenc: use codec descriptors instead of AVCodecs to write codec info

Also, stop using AVCodecContext.codec_name as fallback, since it will be
deprecated.

Changes the result of the lavf-asf test (and its associated seektest),
since 'msmpeg4v3' gets written instead of just 'msmpeg4'.

10 years agooggparseskeleton: do not use AVCodecContext.codec_name
Anton Khirnov [Tue, 29 Apr 2014 13:42:38 +0000 (15:42 +0200)]
oggparseskeleton: do not use AVCodecContext.codec_name

That field is undocumented, of dubious use, and will be deprecated.

10 years agomov: export stsd Compressorname in metadata
Anton Khirnov [Tue, 29 Apr 2014 13:37:52 +0000 (15:37 +0200)]
mov: export stsd Compressorname in metadata

Stop using the undocumented to-be-deprecated AVCodecContext.codec_name
field.

10 years agolavc: move CODEC_FLAG_MV0 to mpegvideo
Anton Khirnov [Tue, 29 Apr 2014 06:00:03 +0000 (08:00 +0200)]
lavc: move CODEC_FLAG_MV0 to mpegvideo

10 years agolavc: make the xvid-specific "gmc" flag a private option of libxvid
Anton Khirnov [Tue, 29 Apr 2014 06:08:43 +0000 (08:08 +0200)]
lavc: make the xvid-specific "gmc" flag a private option of libxvid

10 years agolavc: move CODEC_FLAG_NORMALIZE_AQP to mpegvideo
Anton Khirnov [Tue, 29 Apr 2014 06:00:03 +0000 (08:00 +0200)]
lavc: move CODEC_FLAG_NORMALIZE_AQP to mpegvideo

10 years agolavc: deprecate CODEC_FLAG_INPUT_PRESERVED
Anton Khirnov [Tue, 29 Apr 2014 05:50:16 +0000 (07:50 +0200)]
lavc: deprecate CODEC_FLAG_INPUT_PRESERVED

The addition of reference-counted frames makes it pointless.

10 years agomatroska: add the Opus mapping
Anton Khirnov [Sun, 27 Apr 2014 11:40:11 +0000 (13:40 +0200)]
matroska: add the Opus mapping

10 years agomatroskaenc: write CodecDelay
Anton Khirnov [Tue, 29 Apr 2014 10:26:38 +0000 (12:26 +0200)]
matroskaenc: write CodecDelay

10 years agolavfi: add Bauer stereo-to-binaural audio filter
Alessandro Ghedini [Tue, 29 Apr 2014 16:53:16 +0000 (18:53 +0200)]
lavfi: add Bauer stereo-to-binaural audio filter

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agomatroskadec: read the CodecDelay element
Anton Khirnov [Tue, 29 Apr 2014 10:03:13 +0000 (12:03 +0200)]
matroskadec: read the CodecDelay element

10 years agolibrtmp: Map native options to librtmp ones when possible
Luca Barbato [Sat, 26 Apr 2014 11:32:03 +0000 (13:32 +0200)]
librtmp: Map native options to librtmp ones when possible

Makes the two protocols nearly seamless.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agohls: Sync the file number with the start sequence
Luca Barbato [Tue, 29 Apr 2014 16:20:17 +0000 (18:20 +0200)]
hls: Sync the file number with the start sequence

10 years agohls: Factor the sequence computation
Luca Barbato [Tue, 29 Apr 2014 16:37:01 +0000 (18:37 +0200)]
hls: Factor the sequence computation

10 years agohls: Print start_number as first sequence value
Luca Barbato [Tue, 29 Apr 2014 08:13:34 +0000 (10:13 +0200)]
hls: Print start_number as first sequence value

The option now behaves as expected.

10 years agosegment: Report the current media sequence
Luca Barbato [Tue, 29 Apr 2014 08:28:11 +0000 (10:28 +0200)]
segment: Report the current media sequence

Useful for debugging mostly.

10 years agohls: Report the current media sequence
Luca Barbato [Tue, 29 Apr 2014 08:07:03 +0000 (10:07 +0200)]
hls: Report the current media sequence

Useful for debugging mostly.

10 years agoavconv_dxva2: define all used GUIDs directly instead of relying on the dxva2api.h...
Hendrik Leppkes [Mon, 28 Apr 2014 21:38:05 +0000 (23:38 +0200)]
avconv_dxva2: define all used GUIDs directly instead of relying on the dxva2api.h header

This avoids issues with inconsistent headers with different versions of mingw-w64 and older MSVC variants.

Fixes FATE builds of MSVC 2010.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agoh264: drop commented out cruft
Anton Khirnov [Tue, 29 Apr 2014 09:13:43 +0000 (11:13 +0200)]
h264: drop commented out cruft

10 years agompegvideo: remove unused MpegEncContext.b4_stride
Anton Khirnov [Thu, 5 Dec 2013 20:17:00 +0000 (21:17 +0100)]
mpegvideo: remove unused MpegEncContext.b4_stride

10 years agompegvideo: move the MpegEncContext fields used from arm asm to the beginning
Anton Khirnov [Thu, 5 Dec 2013 20:14:40 +0000 (21:14 +0100)]
mpegvideo: move the MpegEncContext fields used from arm asm to the beginning

This should reduce the frequency with which the offsets need to be
updated.

10 years agohls: Add an option to prepend a baseurl to the playlist entries
Luca Barbato [Tue, 18 Mar 2014 20:19:00 +0000 (21:19 +0100)]
hls: Add an option to prepend a baseurl to the playlist entries

Useful to generate playlists with absolute paths.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agosegment: Add an option to prepend a string to the list entries
Enrique Arizón Benito [Tue, 18 Mar 2014 20:18:59 +0000 (21:18 +0100)]
segment: Add an option to prepend a string to the list entries

Useful to generate lists with absolute urls.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agoconfigure: rework dxva in avconv handling
Anton Khirnov [Mon, 28 Apr 2014 21:00:38 +0000 (23:00 +0200)]
configure: rework dxva in avconv handling

Move the dxva check after the dependencies have been detected, make sure
the ole32 library exists.

Fixes build on non-Windows.

10 years agoavconv: add support for DXVA2 decoding
Hendrik Leppkes [Tue, 22 Apr 2014 13:22:53 +0000 (15:22 +0200)]
avconv: add support for DXVA2 decoding

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohwaccel: fix dxva2 & vaapi loop filter parameters
Michael Niedermayer [Tue, 22 Apr 2014 13:22:52 +0000 (15:22 +0200)]
hwaccel: fix dxva2 & vaapi loop filter parameters

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agodxva2_h264: add a workaround for old Intel GPUs
Hendrik Leppkes [Tue, 22 Apr 2014 13:22:51 +0000 (15:22 +0200)]
dxva2_h264: add a workaround for old Intel GPUs

Old Intel GPUs expect the reference frame index to the actual surface,
instead of the index into RefFrameList as specified by the spec.

This workaround should be set when using one of the "ClearVideo" decoder
devices.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agodxva2_h264: set the correct ref frame index in the long slice struct
Hendrik Leppkes [Tue, 22 Apr 2014 13:22:50 +0000 (15:22 +0200)]
dxva2_h264: set the correct ref frame index in the long slice struct

The latest H.264 DXVA specification states that the index in this
structure should refer to a valid entry in the RefFrameList of the picture
parameter structure, and not to the actual surface index.

Fixes H.264 DXVA2 decoding on recent Intel GPUs (tested on Sandy and Ivy)

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agolavc: eliminate tb_unreliable()
Anton Khirnov [Fri, 25 Apr 2014 14:33:27 +0000 (16:33 +0200)]
lavc: eliminate tb_unreliable()

If framerate estimation is enabled, it makes sense to always apply it,
instead of limiting it to a few specific cases.

10 years agovc1: Do not return an error when skipping b frames
Alessandro Ghedini [Mon, 21 Apr 2014 17:10:52 +0000 (19:10 +0200)]
vc1: Do not return an error when skipping b frames

This caused mpv (and possibly others) to fallback to software decoding after
seeking a VC1 stream.

Bug-Id: 667

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agomov: Emit the correct tags for clcp tracks
Miles Gould [Mon, 21 Apr 2014 13:46:22 +0000 (14:46 +0100)]
mov: Emit the correct tags for clcp tracks

Bug-Id: 664

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agomov: Write prof section of tapt tag
Aidan Skinner [Mon, 21 Apr 2014 13:47:38 +0000 (14:47 +0100)]
mov: Write prof section of tapt tag

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agofate: Add fic-in-avi test
Derek Buitenhuis [Wed, 23 Apr 2014 16:15:13 +0000 (12:15 -0400)]
fate: Add fic-in-avi test

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agoaarch64: NEON optimized FIR audio resampling
Janne Grunau [Mon, 21 Apr 2014 14:13:28 +0000 (16:13 +0200)]
aarch64: NEON optimized FIR audio resampling

Optimized for the default filter length 16.

30% faster opus silk decoding.

10 years agolavr: define ResampleContext in resample.h
Janne Grunau [Mon, 21 Apr 2014 14:12:21 +0000 (16:12 +0200)]
lavr: define ResampleContext in resample.h

Required for arch optimized resampling.

10 years agolavu: add CHK_OFFS as AV_CHECK_OFFSET to check struct member offsets
Janne Grunau [Wed, 23 Apr 2014 10:19:59 +0000 (12:19 +0200)]
lavu: add CHK_OFFS as AV_CHECK_OFFSET to check struct member offsets

10 years agoflv: Do not mangle dts values for negative cts
Luca Barbato [Wed, 23 Apr 2014 19:20:17 +0000 (21:20 +0200)]
flv: Do not mangle dts values for negative cts

Some applications really mean to send negative pts.

10 years agoflv: Warn only once
Luca Barbato [Wed, 23 Apr 2014 19:19:27 +0000 (21:19 +0200)]
flv: Warn only once

No point in sending the message multiple time.

10 years agoflv: Improve log messages
Luca Barbato [Wed, 23 Apr 2014 19:04:20 +0000 (21:04 +0200)]
flv: Improve log messages

Messages should start with a capital letter and possibly end with a "."
if they are statements.

10 years agoon2avc: change a comment at #endif to match actual define
Kostya Shishkov [Wed, 23 Apr 2014 18:02:37 +0000 (20:02 +0200)]
on2avc: change a comment at #endif to match actual define

10 years agoOn2 AVC decoder
Kostya Shishkov [Sun, 9 Jun 2013 18:03:33 +0000 (20:03 +0200)]
On2 AVC decoder

10 years agofic: Make warning message more accurate
Derek Buitenhuis [Wed, 23 Apr 2014 16:53:00 +0000 (17:53 +0100)]
fic: Make warning message more accurate

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agofic: Remove redundant clips
Derek Buitenhuis [Wed, 23 Apr 2014 16:18:36 +0000 (12:18 -0400)]
fic: Remove redundant clips

The equations can't overflow or underflow anyway.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agofic: Simplify alpha blending
Derek Buitenhuis [Tue, 22 Apr 2014 20:42:11 +0000 (16:42 -0400)]
fic: Simplify alpha blending

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agoavutil: move av_get_time_base_q() after include rational.h
Michael Niedermayer [Tue, 22 Apr 2014 20:44:14 +0000 (22:44 +0200)]
avutil: move av_get_time_base_q() after include rational.h

Fix compilation

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agoavutil: Add av_get_time_base_q()
Derek Buitenhuis [Tue, 22 Apr 2014 05:26:07 +0000 (07:26 +0200)]
avutil: Add av_get_time_base_q()

This fixes usage of AV_TIME_BASE_Q in C++ applications, which
cannot use compound literals directly in their code.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agofic: Support rendering cursors
Derek Buitenhuis [Sat, 19 Apr 2014 22:46:52 +0000 (18:46 -0400)]
fic: Support rendering cursors

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agofic: Use proper quantization matrix index
Derek Buitenhuis [Sat, 19 Apr 2014 19:40:35 +0000 (15:40 -0400)]
fic: Use proper quantization matrix index

The matrices are not zigzagged.

Fixes artefacting.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agoaarch64: NEON vorbis_inverse_coupling
Janne Grunau [Sun, 20 Apr 2014 15:57:36 +0000 (17:57 +0200)]
aarch64: NEON vorbis_inverse_coupling

From the ARMv7 NEON version. 16 times faster as the C version, overall
more than 12% faster vorbis decoding on Apple's A7.

10 years agoaarch64: NEON fixed/floating point MPADSP apply_window
Janne Grunau [Sat, 19 Apr 2014 16:17:23 +0000 (18:17 +0200)]
aarch64: NEON fixed/floating point MPADSP apply_window

30%/25% (fixed/float) faster mp3 decoding on Apple's A7. The floating
point decoder is approximately 7% faster.

10 years agoaarch64: NEON float to s16 audio conversion
Janne Grunau [Wed, 16 Apr 2014 21:47:32 +0000 (23:47 +0200)]
aarch64: NEON float to s16 audio conversion

10 years agoaarch64: NEON float (i)MDCT
Janne Grunau [Tue, 15 Apr 2014 16:35:57 +0000 (18:35 +0200)]
aarch64: NEON float (i)MDCT

Approximately as fast as the ARM NEON version on Apple's A7.

10 years agoaarch64: NEON float FFT
Janne Grunau [Wed, 26 Mar 2014 14:20:42 +0000 (15:20 +0100)]
aarch64: NEON float FFT

Approximately as fast as the ARM NEON version on Apple's A7.

10 years agolavf: do not use the parser duration for video
Anton Khirnov [Sun, 13 Apr 2014 05:05:27 +0000 (07:05 +0200)]
lavf: do not use the parser duration for video

The parser has no way of knowing video duration, and therefore no video
parsers set it.

10 years agoavconv: always reset packet pts after decoding an audio frame
Anton Khirnov [Sat, 12 Apr 2014 19:55:46 +0000 (21:55 +0200)]
avconv: always reset packet pts after decoding an audio frame

Currently, if a decoder sets AVFrame.pts, we'd send the same timestamp
to it twice, which is wrong.

10 years agolavc: improve AVCodecContext.delay doxy
Anton Khirnov [Fri, 11 Apr 2014 15:47:51 +0000 (17:47 +0200)]
lavc: improve AVCodecContext.delay doxy

10 years agotxd: do not set the codec timebase.
Anton Khirnov [Sun, 13 Apr 2014 08:43:45 +0000 (10:43 +0200)]
txd: do not set the codec timebase.

It is not supposed to be changed from outside of lavc.
Set the stream timebase and average framerate instead.

10 years agompegvideo: Use the current_picture pts
Michael Niedermayer [Fri, 18 Apr 2014 21:11:31 +0000 (23:11 +0200)]
mpegvideo: Use the current_picture pts

The picture slot can be recycled by select_input_picture and
only current_picture is populated with the valid pts.

Unbreak timestamps when in cbr mode.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agoavcodec/fic: Fix return value check
Michael Niedermayer [Wed, 16 Apr 2014 00:06:37 +0000 (02:06 +0200)]
avcodec/fic: Fix return value check

Fixes part of Ticket3466
Found-by: Andrey_Karpov / PVS-Studio
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agoavcodec/fic: clear slice_data
Michael Niedermayer [Sat, 15 Feb 2014 16:19:59 +0000 (17:19 +0100)]
avcodec/fic: clear slice_data

Fixes artifacts
Fixes use of freed memory

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agoavcodec/fic: fix slice checks
Michael Niedermayer [Sat, 15 Feb 2014 16:19:32 +0000 (17:19 +0100)]
avcodec/fic: fix slice checks

fix integer overflows

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agofic: use correct IDCT
Kostya Shishkov [Tue, 1 Apr 2014 19:00:47 +0000 (21:00 +0200)]
fic: use correct IDCT

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
10 years agomvc: Specify the pixel format for the mv-mvc* tests
Martin Storsjö [Sat, 19 Apr 2014 15:51:38 +0000 (18:51 +0300)]
mvc: Specify the pixel format for the mv-mvc* tests

Also set the RGBA pixel format correctly as the native endian format,
which is what it returns.

This fixes the tests on big endian.

Signed-off-by: Martin Storsjö <martin@martin.st>