For stereo we need to use intermediate planar buffers, but mono does not need
to be deinterleaved so the output buffer can be used directly.

ENOTSUP is not defined on some systems

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

Luma slices in YUV420 colourspace should have height in multiple of two since
they have the same line of chrominance data corresponding to pair of them.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

This fixes false positives of has_codec_delay_been_guessed() for
streams where not every input picture generates an output picture,
such as interlaced H264.

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

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

Restore behavior of identifying files with huge id3 tags as
mp3 at AVPROBE_SCORE_MAX/4. This was broken in r25378 and subsequently
removed in r25929.

This only encourages our users to put id3v2 tags on non-mp3 files to opt
into extension based probing.

Locking the decoder against channel config changes in
parse_adts_frame_header() seems to be unnecessary and
streams with channel config changes are reported.

The sample in http://roundup.libav.org/issue999 still works.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>

This avoids having to convert all input data from int16 to float, which is used
internally for encoding.