* qatar/master:
  aacenc: Fix number of coefficients used in a LFE channel.
  aacenc: Fix a segfault with grouped psymodel.
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The previous default value 0x0 was not good, since it is not even
valid.

In apply_unsharp(), when y is >= height, prevent out-of-buffer reading
from src, read from the last buffer line in src2 instead.

The check was implemented in the original unsharp libmpcodecs code and
lost in the port.

This also fixes output discrepancy between the two filters.

The delogo filter has been natively integrated into libavfilter,
simplify.

The ported filter supports named option parsing and more YUV formats.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>

The spec states:

 *  Only the lowest 12 spectral coefficients of any LFE may be non-zero

We were using the 12 lowest *bands*.

10l: Forgot about TYPE_LFE being after TYPE_CPE. Which causes a
segfault when encoding 5.1

Allow to cache more than one frame (e.g. for filters which return
more than one frame when avfilter_request_frame() is called on them),
and do not discard previously cached frames when a new one is added.

The new function provides a more flexible interface for peeking at a
FIFO buffer data.

In lavfi_read_header(), use the pad index designated in the inout for
linking an output to a sink, rather than always 0. Fix link creation
for filters with more than one output (e.g. the split filter).

Custom PAR num/denum are coded -1.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

FFmpeg writes data_size as AU_UNKNOWN_SIZE, make demuxer not
fail when data_size is set to this value.
Should fix trac issue #394.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

* qatar/master:
  swscale: add dithering to yuv2yuvX_altivec_real
  rv34: free+allocate buffer instead of reallocating it to preserve alignment
  h264: add missing brackets.
  swscale: use 15-bit intermediates for 9/10-bit scaling.
Merged-by: Michael Niedermayer <michaelni@gmx.at>

This is based on the original work by Baptiste Coudurier.

This is based on the original work by Baptiste Coudurier.

This is based on the original work by Baptiste Coudurier.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>

Also drop obfuscating CHROMA_WIDTH/HEIGHT macros.

More consistent with the original libmpcodecs code, and the name
"unsharpen" was confusing.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* commit 'b2c08787':
  Move x86util.asm from libavcodec/ to libavutil/.
  Move x86inc.asm to libavutil/.
  APIchanges: note error_recognition in lavf
  lavf: add support for error_recognition, use it in avidec, and bump minor API version
  avconv: change semantics of -map
  avconv: get rid of new* options.
  cmdutils: allow precisely specifying a stream for AVOptions.
  configure: add missing CFLAGS to fix building on the HURD
  libx264: Include hint for possible values for configuring libx264
  cmdutils: allow ':'-separated modifiers in option names.
  avconv: make -map_metadata work consistently with the other options
  avconv: remove deprecated options.
  avconv: make -map_chapters accept only the input file index.
  Make a copy of ffmpeg under a new name -- avconv.
  ffmpeg: add a warning stating that the program is deprecated.
  Add weighted motion compensation for RV40 B-frames
  RV3/4: calculate B-frame motion weights once per frame
  Move RV3/4-specific DSP functions into their own context
  mjpeg: propagate decode errors from ff_mjpeg_decode_sos and ff_mjpeg_decode_dqt
  h264: notice memory allocation failure

Conflicts:
	.gitignore
	Makefile
	cmdutils.c
	configure
	doc/ffplay.texi
	doc/ffprobe.texi
	doc/ffserver.texi
	libavcodec/libx264.c
	libavformat/avformat.h
	libavformat/avidec.c
	libavformat/version.h
	tests/lavf-regression.sh
	tests/lavfi-regression.sh
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

It just does that part in scalar form, I doubt using a vector store
over 2 array would speed it up particularly.

The function should be written to not use a scratch buffer.