Commit graph

1723 commits

Author SHA1 Message Date
Vincent Penquerc'h
c7282a5718 alsasrc: fail gracefully when ALSA does not give timestamps
https://bugzilla.gnome.org/show_bug.cgi?id=660170
2011-10-03 11:14:09 +02:00
Vincent Penquerc'h
15dc839467 textoverlay: add YV12 support
Basically the same as I420, just with chroma planes swapped.

https://bugzilla.gnome.org/show_bug.cgi?id=660604
2011-10-01 19:18:02 +01:00
Vincent Penquerc'h
c956c5fd00 oggstream: only use information from skeleton if we have nothing better
The codec setup headers are a lot more likely to have correct information,
especially as it's easy to remux a skeleton in a file where streams don't
have the same parameters (I've even seen a file with two skeletons).

Still, this is useful in the case we have a codec we can't decode, so we
can at least (theoretically) convert granpos to time, so we discard this
information if the codec setup has already provided it.

This fixes playback on (at lesat) the original archive.org encoding of
"The Night of the Living Dead" (now replaced by another encoding).

https://bugzilla.gnome.org/show_bug.cgi?id=612443
2011-09-19 23:21:23 +01:00
Sebastian Dröge
d094913a61 textoverlay: Protect against accessing the NULL parent of the pads during shutdown
Fixes bug #658901.
2011-09-19 09:34:08 +02:00
Tim-Philipp Müller
15d8082a55 oggdemux: remove superfluous check in newsegment event handler
If we get a newsegment event from upstream, we can be quite
sure we're not operating pull-based.
2011-09-16 20:14:39 +01:00
Tim-Philipp Müller
049e275632 oggdemux: minor printf format fix 2011-09-16 20:11:56 +01:00
Vincent Penquerc'h
89fc5b4bd8 oggdemux: fix wedge when seeking twice quickly in push mode
This could happen when testing with navseek, and pressing
right and left at roughly the same time. The current chain
is temporarily moved away, and this caused the flush events
not to be sent to the source pads, which would cause the
data queues downstream to reject incoming data after the
seek, and shut down, wedging the pipeline.

Now, I can't really decide whether this is a nasty steaming
hack or a good fix, but it certainly does fix the issue, and
does not seem to break anything else so far.

https://bugzilla.gnome.org/show_bug.cgi?id=621897
2011-09-16 20:07:33 +01:00
Vincent Penquerc'h
0173afa38c oggdemux: implement push mode seeking
This patch implements seeking in push mode (eg, over the net)
in Ogg, using the double bisection method.
As a side effect, it also fixes duration determination of network
streams, by seeking to the end to check the actual duration.

Known issues:
- Getting an EOS while seeking stops the streaming task, I can't
  find a way to prevent this (eg, by issuing a seek in the event
  handler).
- Seeking twice in a VERY short succession with playbin2 fails
  for streams with subtitles, we end up pushing in a dataqueue
  which is flushing. Rare in normal use AFAICT.
- Seeking is slow on slow links - byte ranges guesses could be
  made better, decreasing the number of required requests
- If no granule position is found in the last 64 KB of a stream,
  duration will be left unknown (should be pretty rare)

https://bugzilla.gnome.org/show_bug.cgi?id=621897
2011-09-16 19:47:10 +01:00
Vincent Penquerc'h
6704b37fc3 oggdemux: do not propagate discontinuities in sparse streams
The first packet of a sparse stream may arrive after an initial
delay in the stream. If ogg_stream_packetout reports a discontinuity
in a sparse stream, do not propagate it to other streams in the
chain unnecessarily.

https://bugzilla.gnome.org/show_bug.cgi?id=621897
2011-09-14 23:20:01 +01:00
Thomas Vander Stichele
d223a6dd59 theoraenc: Fix descriptions of properties 2011-09-11 14:22:59 -04:00
Vincent Penquerc'h
cea0ac790f theoraenc: do not automatically override quality when using target bitrate
If both quality and bitrate are set, libtheora will try to meet
both constraints, causing it to prefer emitting a smaller number
of good frames, to emitting the full number of frames that would
not meet the requested quality. This causes a slideshow effect
when the bitrate is low and the quality is high. And the default
theoraenc is high (48/63).

So only set quality when it is requested, and leave it unset
otherwise.

https://bugzilla.gnome.org/show_bug.cgi?id=658443
2011-09-07 13:27:33 +02:00
Vincent Penquerc'h
02763b2fa2 theoraenc: fix caps leak
https://bugzilla.gnome.org/show_bug.cgi?id=657333
2011-08-25 15:17:01 +01:00
Vincent Penquerc'h
fd218c4ec2 oggmux: fix leaks in skeleton writing
https://bugzilla.gnome.org/show_bug.cgi?id=563251
2011-08-25 08:32:26 +02:00
Vincent Penquerc'h
29038283bc oggmux: generate message headers from received tags
Some message headers can be deduced from tags (eg, "Language").

https://bugzilla.gnome.org/show_bug.cgi?id=563251
2011-08-25 08:32:20 +02:00
Vincent Penquerc'h
53c8656248 ogg: use memory slices where appropriate
While there, avoid zeroing newly allocated memory where unnecessary

https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-25 08:26:49 +02:00
Vincent Penquerc'h
7b8b0fa1bb oggdemux: do not warn when reaching EOS while scanning for the end chain
After all, we were asking for it.

This gets rid of the last warning-about-expected-condition.

w00t.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:32:21 +02:00
Vincent Penquerc'h
df40ddf0aa oggdemux: add media type to chain information reports
One more little step in making logs a little less abstruse.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:31:58 +02:00
Vincent Penquerc'h
1e606c0456 oggstream: correctly identify skeleton EOS packet
It is 0 byte, and was triggering the "bad packet" logic.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:31:51 +02:00
Vincent Penquerc'h
68ed992e7e oggdemux: do not warn about expected occurences
In this case, finding a skeleton packet.
Once upon a time, it used to be rare indeed, but no more.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:31:30 +02:00
Vincent Penquerc'h
4fdb52871c oggdemux: do not warn when finding a non BOS page
After all, we do hope to find actual data for these streams.
However, warn if we could not set up a chain when we find a
non BOS page, as that means we don't have a valid Ogg stream.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:31:17 +02:00
Vincent Penquerc'h
564eedd214 oggdemux: rename local variable for clarity
While the casual reader might end up bewildered by just why this
change might increase clarity, it just happens than, in the libogg
and associated sources, op is the canonical name for an ogg_packet
whlie og is the canonical name for an ogg_page, and reading this
code confuses me.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:30:55 +02:00
Vincent Penquerc'h
5d18496a5b oggdemux: do not try to determine duration of header packets
Headers are inherently durationless.
Instead, set duration to 0 to avoid increasing tracked granpos,
and do not warn about it, since it is totally expected.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:30:45 +02:00
Vincent Penquerc'h
67a882afe7 oggstream: include stream type in warnings
It makes it easier to work out what's going on.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:30:30 +02:00
Vincent Penquerc'h
43cb76b1d8 oggstream: set skeleton stream media type to application/x-ogg-skeleton
This is to match the typefinder, and to make logs clearer.

https://bugzilla.gnome.org/show_bug.cgi?id=657151
2011-08-24 08:30:17 +02:00
Vincent Penquerc'h
2301f1806f oggmux: add skeleton write support
Version written is 3.0

Base times are left empty for now.

Content-Type should be the MIME type of the stream. It is set to
the GStreamer media type for now, which is probably the same for
the streams oggmux supports.

https://bugzilla.gnome.org/show_bug.cgi?id=563251
2011-08-24 08:21:34 +02:00
Vincent Penquerc'h
8a752e44e2 oggdemux: do not skip sparse streams when determining start times
This fixes demuxing of streams containing only sparse streams,
which would cause an infinite loop in _read_end_chain.

https://bugzilla.gnome.org/show_bug.cgi?id=657062
2011-08-23 10:36:18 +02:00
Vincent Penquerc'h
4e9508e2ec oggdemux: do not ignore sparse streams' start time
But do not wait for them either, if we don't have a packet for them.

https://bugzilla.gnome.org/show_bug.cgi?id=657062
2011-08-23 10:36:03 +02:00
Monty Montgomery
9cbe7c1403 vorbisenc: Relax overly-tight jitter tolerances in gstvobisenc
vorbisenc currently reacts in a rater draconian fashion if input
timestamps are more than 1/2 sample off what it considers ideal. If data
is 'too late' it truncates buffers, if it is 'too soon' it completely
shuts down encode and restarts it.  This is causingvorbisenc to produce
corrupt output when encoding data produced by sources with bugs that
produce a smple or two of jitter (eg, flacdec)
2011-08-23 10:11:18 +02:00
Vincent Penquerc'h
7d3858a14d textoverlay: fix text buffer leak
Make sure to always unref the input text buffer.

Reported by bcxa.sz@gmail.com.

https://bugzilla.gnome.org/show_bug.cgi?id=657049
2011-08-22 10:48:06 +01:00
Vincent Penquerc'h
b7bb1e5633 ogg: do not use 32 bit modifiers to print serial numbers
If ints are 64 bits, 32 bits should get promoted in varargs anyway,
and we don't care about 16 bit ints.
This makes the code a lot more readable, and still gets us nice
hexadecimal 32 bit serialnos.

https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 20:18:53 +01:00
Vincent Penquerc'h
1dadc5eac1 ogg: get the operator precedence right, even if only a doc
https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:17:40 +02:00
Vincent Penquerc'h
cf370e0219 oggstream: vorbis has a preroll of 2
https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:17:28 +02:00
Vincent Penquerc'h
96e7bf53e9 oggstream: new convenience function to get a stream's media type
This will make logging a lot clearer, both in code and in output.

https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:17:15 +02:00
Vincent Penquerc'h
8304b7b40d ogg: move the "always flush page" to oggstream
It avoids checking for specific media types in the muxer.

https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:16:50 +02:00
Vincent Penquerc'h
fd8434fc74 oggmux: use oggstream to decide which BOS packets to place first
Ogg recommends video BOS packets to be first.
Use the "is_video" flag in oggstream to select those, rather than
check for known mime types.

https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:16:19 +02:00
Vincent Penquerc'h
5bbf7109ec ogg: rationalize serialno type to guint32
It is a 32 bit unsigned number.
Sure, the libogg API uses a long, but that's an unfortunate oversight.

https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:15:51 +02:00
Vincent Penquerc'h
684b90ba74 oggmux: factor the header packet creation code
https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:13:59 +02:00
Vincent Penquerc'h
3fe42b8e7e oggmux: headers should always have granpos 0
https://bugzilla.gnome.org/show_bug.cgi?id=656775
2011-08-18 11:13:47 +02:00
Jonathan Liu
8bd3bdaf37 oggstream: Fix crashes with 0-byte vorbis packets
Fixes bug #655574.
2011-08-03 10:18:29 +02:00
Philip Jägenstedt
7d22d91fca theoradec: segfault on 0-byte ogg_packet in _chain_reverse 2011-08-03 10:03:00 +02:00
Stefan Sauer
2db389f775 textoverlay: improve the example
Mentioned that this is not ment to be used with subtitles and suggest alternatives.
2011-07-25 12:04:48 +02:00
Stefan Sauer
5d5ab90e51 textoverlay: add example for feeding from stdin 2011-07-23 14:22:03 +02:00
Stefan Sauer
0667b1adf5 textoverlay: keep untimestamped textbuffer until next one
Instead of discarding untimestamped text-buffers immeditely after rendering,
keep them until we receive the next text buffer.
Fixes #654959
2011-07-23 14:04:14 +02:00
David Schleef
a5323107d4 oggmux: check for EOS on both current and best pad
Oops, need both.  Fixes #654270.
2011-07-09 18:33:38 -07:00
David Schleef
2fa9bf2be5 oggmux: check for EOS on current pad, not best
Fixes #654270.
2011-07-09 18:24:26 -07:00
David Schleef
ea0d666d11 oggmux: refactor how EOS is determined
This decreases the number of buffers held on each pad by one,
eliminating next_buffer.  Simplifies the logic by relying solely
on CollectPads to let us know when a pad is in EOS.  As a side
benefit, the collect pads related code is structured more like
other CollectPad users.

The previous code would occasionally mark the wrong pad as EOS,
causing the code to get in a state where all the streams were
finished, but EOS hadn't been sent to the source pad.
2011-06-10 23:54:07 -07:00
Tim-Philipp Müller
c692191c33 GST_PLUGINS_BASE_LIBS is not defined in -base. 2011-06-08 12:21:43 +01:00
Mark Nauwelaerts
99944a5a81 oggmux: determine granulepos metadata using stream mapper whenever possible
... which unfortunately is not the case for all types, but at least so for
most common ones.
2011-06-06 13:02:49 +02:00
Mark Nauwelaerts
de1f593537 oggmux: convert incoming buffer timestamp to running time
... so all subsequent manipulation can take place in the proper timeline
without further ado.
2011-06-06 13:02:47 +02:00
Mark Nauwelaerts
2e7eddd35e oggmux: remove superfluous code
... since there is nothing in oggstream that cares (or even should)
about granulepos for what is being asked from it.
2011-06-06 13:02:43 +02:00