The special cases in demuxers and decoders are a mess otherwise (and more
would be needed to support it fully)
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: rogerdpack <rogerpack2005@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: rogerdpack <rogerpack2005@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* qatar/master:
  vorbisdec: remove some pointless comments
  avprobe: add const to AVCodec pointer
  libavformat: add const to AVCodec pointers
  pthread: add const to AVCodec pointers

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

Signed-off-by: Mans Rullgard <mans@mansr.com>

Rework the transcode() function and its immediate annexes to have
the same structure as in avconv, while still maintaining proper
scheduling to avoid accumulation.

Using -filter_complex without inputs now works.

Fixes Ticket976
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Fixes Ticket971
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Use ceil on the max duration, rather than approximating it to the nearest
integer.

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

It seems the condition was flipped from what was intended
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

This may fix Ticket730
Reviewed-by: Nicolas George <nicolas.george@normalesup.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Without that, it would repeatedly produce empty packets
if the codec if forced to something invalid.

Fix trac ticket #1652.

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

The h264_vdpau decoder crashed if output colorspace was not 8-bit 420.
Add a check to error out instead (current hardware does not support
other colorspaces, so successful decoding is not possible).

Check implemented at a different place by michael, thus blame for bugs goes to michael
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Mans Rullgard <mans@mansr.com>

Signed-off-by: Mans Rullgard <mans@mansr.com>

Signed-off-by: Mans Rullgard <mans@mansr.com>

* qatar/master:
  mpegvideo_enc: don't use deprecated avcodec_encode_video().
  cmdutils: refactor -codecs option.
  avconv: make -shortest a per-output file option.
  lavc: add avcodec_descriptor_get_by_name().
  lavc: add const to AVCodec* function parameters.
  swf(dec): replace CODEC_ID with AV_CODEC_ID
  dvenc: don't use deprecated AVCODEC_MAX_AUDIO_FRAME_SIZE
  rtmpdh: Do not generate the same private key every time when using libnettle
  rtp: remove ff_rtp_get_rtcp_file_handle().
  rtsp.c: use ffurl_get_multi_file_handle() instead of ff_rtp_get_rtcp_file_handle()
  avio: add (ff)url_get_multi_file_handle() for getting more than one fd
  h264: vdpau: fix crash with unsupported colorspace
  amrwbdec: Decode the fr_quality bit properly

Conflicts:
	Changelog
	cmdutils.c
	cmdutils_common_opts.h
	doc/ffmpeg.texi
	ffmpeg.c
	ffmpeg.h
	ffmpeg_opt.c
	libavcodec/h264.c
	libavcodec/options.c
	libavcodec/utils.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

The simplifications are due to the recent option changes.

This is technically a major compatibility break, but seems the most
natural default and what users would expect without reading the docs.

An equivalent framestep filter has been natively integrated.

This filter is inspired upon libmpcodecs/vf_framestep.c, by Daniele
Forghieri.

Only-keyframe output is not supported, since that feature can be achieved
through the more versatile select filter.

Address trac ticket #1642.

This is useful for the M3U8 format incoming patch.

Allow factorization for the incoming patches, and avoid code/logic
duplication.

Decouple segment index wrapping and segment count logic. The counter
should not be reset since it detects the segment increasing timeframe.

This seems the expected behavior.

Address trac ticket #1599.

This allows the iteration callbacks to discover the internal class and
options, and show them when required.

This is required for iterating through the filter class (without
incurring into a loop).

This is due to the design of components iteration through AVClass
child_class_next() callback, which requires that two components cannot
share the same class.

Consistent with --enable-doc, and easier to type.