First, container stores only DTS and not PTS as it was believed.
Second, multiple frames in a packet store timestamp instead of position
after the frame length.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

This removes out-of-edge motion compensation artifacts (easily spotted green
blocks in avplay, gray blocks in transcoding), for example here:
http://samples.libav.org/samples/real/tv_watching_t1.rmSigned-off-by: Diego Biurrun <diego@biurrun.de>

This reverts commit 4f4f3384.

This commit has some ugly corner cases and needs to be discussed
further.

It allows to check if a given codec can be written into a container.

Metadata currently is written only at the start of the file in normal
cases, when transcoding from a rtmp source metadata could be
written later and the offset recorded can exceed 32bit.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

ID3v1 are legacy tags with several limitations; furthermore
avconv/ffmpeg writes the tags in UTF-8 which probably has near-0
software support.

Add a -write_id3v1 option to be able to turn it on; disabled by default.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

MpegTSContext's AVClass member can be NULL.

New streams are always added to the last output file.

Move find_codec_or_die(), choose_codec(), new_output_stream() and
read_avserver_streams() below transcode() where they belong.

It's only used in opt_output_file(), so move it right above
opt_output_file().

parse_forced_key_frames() is only used in new_video_stream(), so move it
right above new_video_stream().

Its only reason for its existence was ffmpeg's inability to properly assign
AVOptions to streams. Now this is not a problem anymore, so 'ab' should
go.

It's often mistakenly used as 'same quality', emphasize that it's not
true in the manual.

The new option doesn't depend on its placement wrt -new* options (which
don't exist anymore) and works in a similar way as per-stream AVOptions.

-[vas]codec remain as aliases to -codec:[vas]

Nothing except x264 uses them and native x264 presets should be used
with private options.

Signed-off-by: Anton Khirnov <anton@khirnov.net>

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

AVFormatParameters are converted into corresponding private options in
av_open_input_file/stream() compat wrappers, so accessing them from
demuxers is redundant.

Get rid of AVFormatParameters usage.

Custom PAR num/denum are in 1-256 range.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Signed-off-by: Diego Biurrun <diego@biurrun.de>

Signed-off-by: Diego Biurrun <diego@biurrun.de>

Signed-off-by: Diego Biurrun <diego@biurrun.de>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

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

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.

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