Also rearranges some functions for easier cleanup on failure.

reduces memory usage when the stream has fewer than MAX_CHANNELS

check frame header channel count against header/container channel count.

It is already done when using it to set sample_fmt.

Otherwise the delay expressed in has_b_frames increases with every
avcodec_close/avcodec_open.
Fixes fate-ea-dct with more than 1 thread.

Fixes Ticket592

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

Add whitespace.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Fixes Ticket561

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

Cleanup the options and add explicit fragment_size avoption.

* qatar/master: (22 commits)
  g722dec: check output buffer size before decoding
  g722dec: cosmetics: reindent/linewrap
  g722dec: remove the use of lowres for half-rate decoding.
  tta: check for extradata allocation failure in tta demuxer
  tta: check for allocation failure of decode_buffer
  tta: use correct frame_length calculation.
  tta: add support for decoding 24-bit sample format
  cosmetics: indentation
  tta: remove pointless braces
  tta: check output buffer size after adjusting frame length for last frame
  tta: fix reading of format in TTA header.
  tta: remove useless commented-out lines
  tta: check remaining bitstream size while reading unary value
  lavf: deprecate AVStream.stream_copy
  avconc: split choose_codec() to choose_decoder/choose_encoder.
  lavf: simplify by using FFMAX/FFMIN.
  mpegenc: add preload private option.
  cosmetics: simplify latm_decode_init
  latm: avoid unnecessary reinit of the aac decoder
  aacdec: initialize sbr context only in new channel elements
  ...

Conflicts:
	avconv.c
	libavcodec/resample.c
	libavcodec/tta.c
	libavformat/utils.c

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

This should be changed to LIBAVDEVICE_IDENT once it exists.

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>

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

Signed-off-by: Stefano Sabatini <stefasab@gmail.com>

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

Without this patch each displayed second is incremented 1 frame early,
second 0 is only 24 frames long where as every other second is 25 frames
long.

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

It is broken because an AVCodecContext can be opened/closed multiple
times, and sample_rate is getting divided by 2 each time that happens.

This removes the only use of lowres for audio.

using a floating-point calculation is not necessary.

Note that this will not work in most cases with avconv and avplay due to the
AVCODEC_MAX_AUDIO_FRAME_SIZE limit, but it will decode correctly if given a
large enough output buffer.

TTA does not support float at all, and format 2 is encrypted TTA.

It's only used in avconv, so it properly belongs to OutputStream struct
there.

Prevents -c copy from working for input streams and allows to move
stream_copy variable from AVStream to OutputStream.

Deprecate AVFormatContext.preload.

The situation was not clear when support was added but it is now:
CELT and Opus are really two different codecs.
The current code supports CELT via libcelt, not Opus.

Signed-off-by: Nicolas George <nicolas.george@normalesup.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

audio_resample can not reduce the number of channels

* cus/stable:
  ffplay: add frame drop statistics
  ffplay: consider estimated time of filter in early frame drop
  ffplay: reimplement early frame drop

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

* qatar/master:
  adpcm: use sign_extend()
  mpeg12: fix mpeg_decode_slice context parameter type
  Revert "mpeg12: move full_pel from MpegEncContext to Mpeg1Context"

Conflicts:
	libavcodec/mpeg12.c
	libavcodec/mpeg12.h

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

Signed-off-by: Marton Balint <cus@passwd.hu>

Signed-off-by: Marton Balint <cus@passwd.hu>

This patch reimplements early frame drop, it is now based on the current
difference between the master clock and the video clock, and the pts of the
current and the last displayed (or skipped) frame.  If the frame to be added to
the queue is late after decoding, then we drop it early because later we would
drop it anyway (unless it is the only frame in the picture queue).

The current approach has only one downside that I know of, it does not handle
well when the filters are changing significantly the pts of the frames, because
we compare pts values from filtered and unfiltered frames.

We also start using the pictq_mutex to ensure consistent video_current_pts,
video_current_pts_drift, frame_last_pts, frame_last_dropped_pts and
frame_last_dropped_pos values.

Signed-off-by: Marton Balint <cus@passwd.hu>