Performance improvements:

quant_bands:
with:     681 decicycles in quant_bands, 8388453 runs,    155 skips
without: 1190 decicycles in quant_bands, 8388386 runs,    222 skips
Around 42% for the function

Twoloop coder:

abs_pow34:
with/without: 7.82s/8.17s
Around 4% for the entire encoder

Both:
with/without: 7.15s/8.17s
Around 12% for the entire encoder

Fast coder:

abs_pow34:
with/without: 3.40s/3.77s
Around 10% for the entire encoder

Both:
with/without: 3.02s/3.77s
Around 20% faster for the entire encoder
Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
Tested-by: Michael Niedermayer <michael@niedermayer.cc>
Reviewed-by: James Almer <jamrial@gmail.com>

Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>

Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>

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

return AVERROR code on error
Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>

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

This code is not correct for git master
Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

test for flt to s16
should pass on correct rounding to nearest
Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>

Reviewed-by: Jan Sebechlebsky <sebechlebskyjan@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes demuxing the sample file from github pull request 197,
the size of its idx1 tag is 6171936 bytes, followed by a JUNK
tag of 9505704 bytes.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>

this is fontconfig pattern
Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>

do not rebuild when gain/gain_entry command are
equal with old gain/gain_entry
Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>

The problem was introduced in commit 1273bc6d.
Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>

Signed-off-by: Rick Kern <kernrj@gmail.com>

Handles insertion into existing SEI NAL unit, inserts emulation
prevention bytes.
Signed-off-by: Rick Kern <kernrj@gmail.com>

Prevents encode callback from running after codec is closed.
Fixes a crash when an error is returned.
Signed-off-by: Rick Kern <kernrj@gmail.com>

Signed-off-by: Rick Kern <kernrj@gmail.com>

mov_finalize_stsd_codec parses stream information from the ALAC extradata,
so run it after the extradata processing is completed in mov_read_stsd.

Fixes playback of 96kHz ALAC streams muxed by qaac or the reference alac encoder.
Fixes trac ticket #5826

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

Partially revert "avformat/matroskadec: set aspect ratio only when DisplayWidth and DisplayHeight are in pixels"

The code works just fine regardless of unit, so only make sure DisplayUnit
is not "unknown".
Found-by: Nicolas George <george@nsup.org>
Reviewed-by: Nicolas George <george@nsup.org>
Signed-off-by: James Almer <jamrial@gmail.com>

It's container level information on some formats (Matroska, MXF, yuv4mpeg), so
it should be printed at higher log levels than debug.
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>

We don't currently support values 1 (centimeters), 2 (inches) or 3 (DAR),
only the default value 0 (pixels) which doesn't need to be written.

The fate refs are updated as unknown SAR is now signaled in the output
files with the addition of the new element.
Reviewed-by: Carl Eugen Hoyos <ceffmpeg@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>

A missing DisplayUnit element or one with the default value of 0 means
DisplayWidth and DisplayHeight should be interpreted as pixels.

The current code setting st->sample_aspect_ratio is wrong when DisplayUnit
is anything else.
Reviewed-by: Carl Eugen Hoyos <ceffmpeg@gmail.com>
Signed-off-by: James Almer <jamrial@gmail.com>

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

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

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

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

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

Tested functions: av_fifo_generic_peek(), av_fifo_grow()
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

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

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

since:
3e5e5bdf Merge commit '398f015f'
Signed-off-by: James Zern <jzern@google.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

The calculation of width/height should round up, not round down to
prevent setting width or height to 0.

Also image->comps[compno].w is unsigned (at least in openjpeg2), so the
calculation could silently wrap around without the explicit cast to int.
Reviewed-by: Michael Bradshaw <mjbshaw@gmail.com>
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>