Commit Graph

39671 Commits

Author SHA1 Message Date
Nicolas George
423047ea31 vf_drawtext: fix memory leak of glyph data. 2012-04-10 15:16:12 +02:00
Michael Niedermayer
24ab1abfb6 resample: support double precission resampling
This commit is dedicated to the audiophiles who can hear it when a
needle is dropped on the moon.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-10 13:48:55 +02:00
Michael Niedermayer
3d9338b1c2 swr: support int32 and float as internal sample formats
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-10 13:19:29 +02:00
Michael Niedermayer
7f1ae79d38 swr: support float & int32 in the resampler
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-10 13:18:49 +02:00
Michael Niedermayer
605bcf6101 swr-test: Add newline to test output
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-10 13:06:00 +02:00
Michael Niedermayer
2f0801527d resample: update copyright years, they where incorrect.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-10 12:35:37 +02:00
Michael Niedermayer
6bfb3042b3 Merge remote-tracking branch 'qatar/master'
* qatar/master:
  avconv: use default alignment for audio buffer
  avcodec: use align == 0 for default alignment in avcodec_fill_audio_frame()
  avutil: use align == 0 for default alignment in audio sample buffer functions
  avutil: allow NULL linesize in av_samples_fill_arrays() and av_samples_alloc()
  avconv: remove OutputStream.picref.
  avconv: only set SAR once on the decoded frame.
  avcodec: validate the channel layout vs. channel count for decoders
  audioconvert: make av_get_channel_layout accept composite names.
  avutil: add av_get_packed_sample_fmt() and av_get_planar_sample_fmt()

Conflicts:
	doc/APIchanges
	ffmpeg.c
	libavcodec/utils.c
	libavcodec/version.h
	libavutil/audioconvert.c
	libavutil/audioconvert.h
	libavutil/avutil.h
	libavutil/samplefmt.c
	libavutil/samplefmt.h

Merged-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-09 23:02:33 +02:00
Reimar Döffinger
7f66912f1a Fix gif regression test on big-endian.
Decode output must be converted to rgb24 to avoid CRC difference
due to palette being stored in machine endianness.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 21:55:52 +02:00
Michael Niedermayer
3532a87a25 exr: optimize exr_halflt2uint()
30% faster

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-09 20:55:09 +02:00
Reimar Döffinger
cabc0ac72f Add gif and xbm regression tests.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 17:21:35 +02:00
Reimar Döffinger
6458a08700 vf_scale: ensure the palette is set for output.
Since those are pseudo-palette formats, swscale does not write
into data[1], swscale must initialize the palette properly itself.
This lead to frames that actually decoded as all-gray before.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 17:21:35 +02:00
Reimar Döffinger
24eac3cff5 Mark GRAY8 format as pseudo-paletted.
This fixes that the GIF encoder crashes with it because
it has no palette.
And the arguments for the pseudopalette apply to gray8 as
much as to RGB8 etc.
In addition the changes required in lavfi should be needed anyway
when adding support for RGB8 etc.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 17:21:35 +02:00
Reimar Döffinger
bbaf941e29 scale: fix slice rendering with conversion between pal/non-pal.
We can't use whether the input format is paletted to decide that
the output format has a palette in data[1], too, that makes no sense.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 17:21:35 +02:00
Reimar Döffinger
438f3ef821 RV20: try keeping aspect during resolution changes.
Resolution changes are usually only used to scale with
network bandwidth, the (full) resolution specified in the
RM header really is authoritative.
While I am not sure this is the best solution, it is a
conservative approach that still should fix the most
common cases.
In particular, it fixes aspect with the sample from trac
issue  (in MPlayer, ffplay seems to just ignore
sample aspect changes in mid-playback).

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 17:21:34 +02:00
Reimar Döffinger
1d130328ed aacdec: update debug code to always print the processed data.
When decoding LATM, this function will not process extradata
but a different buffer.
It seems this was forgotten to update when LATM support
was added.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 17:21:34 +02:00
Justin Ruggles
679a973e97 avconv: use default alignment for audio buffer 2012-04-09 10:20:35 -04:00
Justin Ruggles
c58846f3a8 avcodec: use align == 0 for default alignment in avcodec_fill_audio_frame()
Use default alignment in audio_get_buffer()
2012-04-09 10:20:35 -04:00
Justin Ruggles
0109a09dc3 avutil: use align == 0 for default alignment in audio sample buffer functions 2012-04-09 10:20:35 -04:00
Justin Ruggles
18ed3788b0 avutil: allow NULL linesize in av_samples_fill_arrays() and av_samples_alloc() 2012-04-09 10:20:35 -04:00
Anton Khirnov
b6c4518a87 avconv: remove OutputStream.picref.
It's only used inside transcode_video() and there's no point in
preserving it between subsequent calls. So use a local variable instead.
2012-04-09 05:13:00 +02:00
Anton Khirnov
695ec04e6a avconv: only set SAR once on the decoded frame.
No point in repeating the assignment for each output stream.
2012-04-09 05:10:38 +02:00
Reimar Döffinger
ed68fd45d3 Add v408 codec regression test.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 00:46:43 +02:00
Justin Ruggles
1337de0c4b avcodec: validate the channel layout vs. channel count for decoders
Set avctx->channel_layout to 0 if the channel count does not match
avctx->channels.
2012-04-08 18:45:58 -04:00
Nicolas George
b2db35995f audioconvert: make av_get_channel_layout accept composite names.
Signed-off-by: Nicolas George <nicolas.george@normalesup.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
2012-04-08 18:35:49 -04:00
Reimar Döffinger
c311afca0a Add roqaudio regression test.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-09 00:19:06 +02:00
Justin Ruggles
4d693b023c avutil: add av_get_packed_sample_fmt() and av_get_planar_sample_fmt()
Based on a patch by Clément Bœsch <ubitux@gmail.com>
2012-04-08 15:13:28 -04:00
Michael Niedermayer
5d6a40bc74 Merge remote-tracking branch 'qatar/master'
* qatar/master:
  rtsp: Don't use av_malloc(0) if there are no streams
  rtsp: Don't use uninitialized data if there are no streams
  vaapi: mpeg2: fix slice_vertical_position calculation.
  hwaccel: mpeg2: decode first field, if requested.
  cosmetics: Fix indentation
  rtsp: Don't expose the MS-RTSP RTX data stream to the caller

Merged-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 20:55:11 +02:00
Reimar Döffinger
4f043c8a11 Allow running fate-wmv8-x8intra manually.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 20:27:35 +02:00
Reimar Döffinger
a5efc8c9bf Allow manually running disabled FATE tests.
This makes it easy to allow people to run tests that are disabled
(e.g. because they are broken) without having to hack Makefiles
by adding the test name to FATE_TESTS-no.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 20:27:35 +02:00
Reimar Döffinger
03ef5047d1 Fix nellymoser encoder crash with hardcoded tables.
Use the correct init function to avoid crashes due to writing
to a rodata location.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 20:26:25 +02:00
Reimar Döffinger
a7b1420ca2 Update reference for disabled fate-wmv8-x8intra test to new format.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 20:01:07 +02:00
Michael Niedermayer
1d80c8db27 ffv1: set slice geometry based on user specified slices.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 17:03:38 +02:00
Michael Niedermayer
8dfbc1c5cf ffv1: allow enabling of version 2 by using some of its features.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 16:48:34 +02:00
Michael Niedermayer
1b295a17d7 ffv1: fix version=2 chroma handling
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 16:01:37 +02:00
Michael Niedermayer
5612359501 ffv1: add a minor version field beginning with ffv1.3
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 16:01:36 +02:00
Martin Storsjö
9294f538e9 rtsp: Don't use av_malloc(0) if there are no streams
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-04-08 15:51:33 +03:00
Martin Storsjö
62c3c8ca78 rtsp: Don't use uninitialized data if there are no streams
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-04-08 15:51:32 +03:00
Reimar Döffinger
069cf86d32 Fix side-data memleak also for audio.
This uses the same code as in decode_video also in decode_audio.
Should fix valgrind FATE failures for nellymoser encode test.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 13:06:13 +02:00
Gwenole Beauchesne
58e1032d8f vaapi: mpeg2: fix slice_vertical_position calculation.
VASliceParameterBufferMPEG2.slice_vertical_position shall express
the slice vertical position from the original bitstream. The HW
decoder will correctly decode to the right line computed from the
appropriate top_field_first and is_first_field flags.

This patch aligns with DXVA's definition, which is what most HW and
drivers expect. In particular, Intel PowerVR (Cedarview et al.) and
NVIDIA (through VA-to-VDPAU layer). Since it looks more complex to fix
binary drivers, I aligned the Intel Gen driver (Sandy Bridge et al.)
to this behaviour, while maintaining compatibility with codec layers
not providing this patch yet.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-04-08 13:10:20 +03:00
Gwenole Beauchesne
9cb150c9ab hwaccel: mpeg2: decode first field, if requested.
If user opted to present fields as they come, then the first field
picture needs to be submitted to the HW for decoding. In particular,
this fixes MPEG-2 decoding of interlaced streams.

Tested on Intel Cedar Trail, Sandy Bridge and Ivy Bridge platforms.
Someone reported on the ffmpeg-devel@ list this also works on DXVA
(Windows) and other Linux platforms (NVIDIA, through the VA wrapper).

This also means a similar patch to non-hwaccel VDPAU may be necessary.

Note: I believe the SLICE_FLAG_ALLOW_FIELD is useless since the first
field shall always be submitted to the HW anyway. Nobody uses HW accels
(dxva, vaapi, vdpau, etc.) without that flag though.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-04-08 13:08:52 +03:00
Martin Storsjö
2ce7f4d4e6 cosmetics: Fix indentation
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-04-08 12:04:29 +03:00
Martin Storsjö
456001486e rtsp: Don't expose the MS-RTSP RTX data stream to the caller
This avoids exposing a dummy AVStream which won't get any data
and which will make avformat_find_stream_info wait for info about
this stream.

Signed-off-by: Martin Storsjö <martin@martin.st>
2012-04-08 12:04:22 +03:00
Thierry Foucu
cb93946820 mpegtsenc: Add support to mux s302 stream
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 09:44:14 +02:00
Michael Niedermayer
352a67e6f8 configure: fix - vs _ error
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-08 09:43:38 +02:00
Reimar Döffinger
58c25724b7 Add nellymoser "fuzzy" FATE test.
Since we cannot specify decode parameters (and also because
it is better in principle) the 1-channel reference file
needs to be enabled, too.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 02:43:56 +02:00
Reimar Döffinger
e9841505d2 Allow setting CODEC_FLAG2_CHUNKS via the name "chunks".
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 02:43:56 +02:00
Reimar Döffinger
d506ef05a0 Fixes for third argument do_image_formats.
Remove it from two places where it is useless, do not apply
it to the encode command and make it apply to the output
instead of the input of the decode command.
Should fix the dpx test.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-08 02:43:56 +02:00
Stefano Sabatini
f0bb1a599c ffprobe: use avbprint API
Simplify, increase robustness.
2012-04-08 00:02:21 +02:00
Michael Niedermayer
6101e5322f Merge remote-tracking branch 'qatar/master'
* qatar/master:
  rtpdec_asf: Set the no_resync_search option for the chained asf demuxer
  asfdec: Add an option for not searching for the packet markers
  cosmetics: Clean up the tiffenc pix_fmts declaration to match the style of others
  cosmetics: Align codec declarations
  cosmetics: Convert mimic.c to utf-8
  avconv: remove an unused function parameter.
  avconv: remove now pointless variables.
  avconv: drop support for building without libavfilter.
  nellymoserenc: fix crash due to memsetting the wrong area.
  libavformat: Only require first packet to be known for audio/video streams
  avplay: Don't try to scale timestamps if the tb isn't set

Conflicts:
	Changelog
	configure
	ffmpeg.c
	libavcodec/aacenc.c
	libavcodec/bmpenc.c
	libavcodec/dnxhddec.c
	libavcodec/dnxhdenc.c
	libavcodec/ffv1.c
	libavcodec/flacenc.c
	libavcodec/fraps.c
	libavcodec/huffyuv.c
	libavcodec/libopenjpegdec.c
	libavcodec/mpeg12enc.c
	libavcodec/mpeg4videodec.c
	libavcodec/pamenc.c
	libavcodec/pgssubdec.c
	libavcodec/pngenc.c
	libavcodec/qtrleenc.c
	libavcodec/rawdec.c
	libavcodec/sgienc.c
	libavcodec/tiffenc.c
	libavcodec/v210dec.c
	libavcodec/wmv2dec.c
	libavformat/utils.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
2012-04-07 22:41:37 +02:00
Reimar Döffinger
5012c07336 Support detecting and demuxing EIA-608 subtitles in mov.
The format is slightly proprietary.
DVDs use a format of
code byte (0x00, 0x01, 0xfe or 0xff), two data bytes
MOV uses instead
cdat/cdt2 atom, two data bytes
Auto-detecting and supporting both in one decoder is trivial,
so a single codec ID is used.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
2012-04-07 22:11:21 +02:00