If http://www.mars.org/mailman/public/mad-dev/2001-May/000262.html is
to be believed, the last buffer must be followed by a number of 0 bytes
in order for the last frame to be decoded (at least in some cases).
Doing so seems to work here, fixing a missing 1152 samples when using
mp3parse before mad (not using mp3parse would yield the correct amount
of samples, if there's extra non-MP3 data after (eg, tag data)).
The mad mp3 decoder element shouldn't parse tags at all really, but we
have so far kept this code around for backwards-compatibility reasons
for people building manual pipelines for some reason. However, as it
turns out that code has never actually worked in 0.10 in practice,
since it only gets executed if mad_frame_decode() returns LOSTSYNC,
which doesn't actually seem to happen any more though because of the
preceding mad_header_decode(), which will discover and report the
sync loss if it runs into a tag and make mad_frame_decode() try to
resync right away.
Discovered this while trying to make it use gst_tag_list_from_id3v2_tag().
ID3 tags are usually handled by id3demux, and should be handled
by id3demux. Tag handling in mad based on libid3tag is very basic
and mostly unnecessary really, so just build this plugin without
ID3 tag support if libid3tag is not available.
Original commit message from CVS:
* ext/mad/gstmad.c: (gst_mad_src_event), (gst_mad_chain),
(gst_mad_change_state):
* ext/mad/gstmad.h:
track discont on incomming buffers and set discont on outgoing
buffers.
Pass unknown events upstreams instead of dropping them.
Original commit message from CVS:
* docs/plugins/gst-plugins-ugly-plugins-docs.sgml:
* docs/plugins/gst-plugins-ugly-plugins-sections.txt:
* ext/a52dec/gsta52dec.c:
* ext/amrnb/amrnbdec.c:
* ext/amrnb/amrnbenc.c:
* ext/amrnb/amrnbparse.c:
* ext/lame/gstlame.c:
* ext/mad/gstmad.c:
* ext/sidplay/gstsiddec.cc:
* gst/asfdemux/gstrtspwms.c:
* gst/mpegaudioparse/gstxingmux.c:
* gst/realmedia/rademux.c:
* gst/realmedia/rdtmanager.c:
* gst/realmedia/rtspreal.c:
* gst/synaesthesia/gstsynaesthesia.c:
Add missing elements to docs. Restore alphabetical order in section
file. Document mad (it was included in docs already).
Fix doc-markup: use convinience syntax for examples
(produces valid docbook), add several refsec2 when we have several
titles. Fix some types.
Original commit message from CVS:
* ext/mad/gstmad.c: (gst_mad_sink_event):
* gst/mpegaudioparse/gstmpegaudioparse.c:
(gst_mp3parse_sink_event):
Push EOS, FLUSH_STOP and NEWSEGMENT immediately instead
of dropping and leaking them.
Original commit message from CVS:
* ext/mad/gstmad.c: (gst_mad_dispose), (gst_mad_sink_event),
(gst_mad_chain):
* ext/mad/gstmad.h:
* gst/mpegaudioparse/gstmpegaudioparse.c: (gst_mp3parse_dispose),
(gst_mp3parse_sink_event), (gst_mp3parse_emit_frame):
* gst/mpegaudioparse/gstmpegaudioparse.h:
Cache all events except EOS if we still have to send a NEWSEGMENT
event. This will let TAG events be forwarded until after decodebin
to an encoder for example as decodebin only links the pads
after NEWSEGMENT. Fixes bug #518933.
Original commit message from CVS:
* ext/mad/gstid3tag.c: (gst_id3_tag_class_init):
Don't add pad templates for the base class but only in the
subclasses.
Original commit message from CVS:
* ext/mad/gstmad.c: (GST_TYPE_MAD_MODE),
(GST_TYPE_MAD_EMPHASIS), (GST_TAG_EMPHASIS), (CHECK_HEADER):
Call g_type_class_ref() from a thread-safe context and not
from the streaming thread to avoid GLib screwing up (see
bug #349410 and #493432). Comment out unused code.
Original commit message from CVS:
* ext/mad/gstmad.c: (mpg123_parse_xing_header):
* gst/mpegaudioparse/gstmpegaudioparse.c:
(gst_mp3parse_handle_first_frame):
Make sure that the Xing TOC starts with 0 and the entries
are increasing over time. Otherwise it's broken and should
be skipped. Fixes bug #507821.
Original commit message from CVS:
* configure.ac:
* ext/mad/gstmad.c: (gst_mad_chain):
Use the new buffer clipping function from gstaudio here and
require gst-plugins-base CVS.
Original commit message from CVS:
* ext/mad/gstmad.c: (_do_init), (gst_mad_init), (index_seek),
(normal_seek), (gst_mad_sink_event), (clip_outgoing_buffer),
(gst_mad_chain), (gst_mad_change_state):
* ext/mad/gstmad.h:
Implement buffer clipping and use GST_BOILERPLATE instead of manual
GType magic. Part one of bug #308312.
Original commit message from CVS:
2007-03-27 Julien MOUTTE <julien@moutte.net>
* ext/mad/gstmad.c: (gst_mad_chain): Generate a perfect
stream and try to adapt to incoming timestamps lack of
precision up to half a frame. This allows generating
a perfect stream even with muxed formats like FLV where
the timestamps are stored in milliseconds.
Original commit message from CVS:
* ext/mad/gstmad.c: (gst_mad_convert_sink), (gst_mad_convert_src):
When handling seeks, don't convert -1 start or stop values to a bogus
byte position - pass them upstream as -1.
Original commit message from CVS:
* ext/mad/gstid3tag.c:
* gst/mpegstream/gstrfc2250enc.h:
* gst/synaesthesia/gstsynaesthesia.c:
More G_OBJECT macro fixing.
Original commit message from CVS:
Patch by: Michal Benes <michal dot benes at itonis dot tv>
* ext/mad/gstmad.c: (gst_mad_chain):
Fix timestamping in mad by only activating a new timestamp when the
previous frame has been decoded. Fixes#350723.
Also clean up some of the non fatal warnings when the input buffer is
too small to decode a header.
