Both may be set by user (libavformat) on decoding.

Fixes Ticket4665
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes Ticket4664

The changed fate tests lack red/blue shades and thus look correct
either way
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

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

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

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

Signed-off-by: Chris Spencer <spencercw@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

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

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

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

* commit '69b92f1b':
  qsvenc: properly handle asynchronous encoding

Conflicts:
	libavcodec/qsvenc.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit 'f5c4d38c':
  qsvdec: properly handle asynchronous decoding

Conflicts:
	libavcodec/qsvdec.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '6b15874f':
  af_resample: do not touch the timestamps if we are not resampling
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

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

* commit '6d592fbd':
  avconv: split creating and (re-)configuring complex filtergraphs

Conflicts:
	ffmpeg_filter.c
	ffmpeg_opt.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>

* commit 'be101bc1':
  avconv: create the complex filtergraphs earlier

Conflicts:
	ffmpeg_opt.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '1959351a':
  avconv: move the no streams failure to open_output_file()
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit 'ab7b0389':
  avconv: factor out the output stream initialization

Conflicts:
	ffmpeg.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit 'bd2ab27c':
  avconv: use read_file() for reading the 2pass stats

Conflicts:
	cmdutils.c
	cmdutils.h
	ffmpeg_opt.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '6d5d9246':
  avconv: move handling the 2pass logfile into avconv_opt

Conflicts:
	ffmpeg.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '59245e0c':
  avconv: set the encoding/decoding_needed flags earlier

Conflicts:
	ffmpeg.c
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

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

* commit '56c2f377':
  avconv: drop update_sample_fmt()

Conflicts:
	ffmpeg.c

No change as the removed code is not in FFmpeg
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit 'e61f3984':
  asfdec: make nb_sub to be unsigned int
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '2883ef34':
  asfdec: read the replicated data in a separate function
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '0989d3ad':
  asfdec: convert condition for the replicated data reading to be safer
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

* commit '40662728':
  asfdec: do not read replicated data when their length is 0
Merged-by: Michael Niedermayer <michael@niedermayer.cc>

Wait for async_depth frames before syncing.

Wait for async_depth frames before syncing.

This filter currently assumes that the input audio is continuous and
does some timestamps manipulation based on this assumption.

This is unnecessary if we are only converting the channel layout or the
sample format, without resampling. In such a case, just leave the
timestamps as they are.

The current code is less than straightforward due to the fact that
output streams can be created based on filtergraph definitions. This
change should make the code simpler and more readable. It will also be
useful in the future commits.

Since global options are processed before all the other options now, we
do not have to try creating the complex filtergraphs several times
anymore, it is enough to do it once after the input files are opened.

It is a better place for it, there is no reason to wait until
transcode_init().

Also, drop the now unused cmdutils_read_file(). There is no reason to
have two functions doing essentially the same thing.

It more logically belongs there.

This will be useful in the following commits.

There is only one decoder left that supports this (libopus, which is not
used by default since we have a native one) and this code goes against
the avconv design, since it propagates information back from the encoder
to decoder.