Take into account the atoms at the end of the 'trak' atom when
recovering it. So that its size (already computed and added in the trak
size) isn't making offsets wrong.
https://bugzilla.gnome.org/show_bug.cgi?id=771478
Fix the check for whether the start time of the segment has
been reached when playing in reverse. Otherwise, playback
stops after reaching the start of any file part, instead of
continuing until all parts within the segment have played
We parse the next moof in advance of having pushed
all samples from the previous one in some cases, and
we'll still need the crypto info from the previous
fragment so keep around any unused crypto info entries
when adding new ones
qtdemux.c: In function ‘qtdemux_parse_samples’:
qtdemux.c:8450:39: error: ‘*’ in boolean context, suggest ‘&&’ instead [-Werror=int-in-bool-context]
if (stream->samples_per_frame * stream->bytes_per_frame) {
~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~
gstmpegaudioparse.c: In function ‘gst_mpeg_audio_parse_reset’:
gstmpegaudioparse.c:209:3: error: ‘memset’ used with length equal to number of elements without multiplication by element size [-Werror=memset-elt-size]
memset (mp3parse->xing_seek_table_inverse, 0, 256);
^~~~~~
gstmpegaudioparse.c: In function ‘gst_mpeg_audio_parse_handle_first_frame’:
gstmpegaudioparse.c:951:7: error: ‘memset’ used with length equal to number of elements without multiplication by element size [-Werror=memset-elt-size]
memset (mp3parse->xing_seek_table_inverse, 0, 256);
^~~~~~
This prevents storing an infinite amount of e.g. comment headers if they
come without a new initialization header in front of them. There can
only be one header of each type.
If we also replace all headers when receiving any possibly following
comments header, we would throw away the config header before being able
to make use of it.
A sparse stream's ending timestamp can be considerably smaller
than the ending timestamps of the other streams, which can lead
to skipping considerable time from the next part.
https://bugzilla.gnome.org/show_bug.cgi?id=761086
In function rtp_jitter_buffer_calculate_pts: If gap in incoming RTP
timestamps is more than (3 * jbuf->clock_rate) we call
rtp_jitter_buffer_reset_skew which resets pts to 0. So components down
the pipeline (playes, mixers) just skip frames/samples until pts becomes
equal to pts before gap.
In version 1.10.2 and before this checking was bypassed for packets with
"estimated dts", and gaps were handled correctly.
https://bugzilla.gnome.org/show_bug.cgi?id=778341
The payloader needs to reset and update the vorbis config data which is
pushed on the network if it receives new headers, or at least, it may
have to do so.
Without this, the stream configuration could change without the
payloader sending the new configuration to the other side.
This reverts commit 107902ec51.
This commit intended to ensure that keyframe seeks land at the
start timestamp of a keyframe, rather than in the middle of one,
but they cause trouble on files with sparse streams, or with
JPEG 'cover art' tracks that have only one or a few JPEG samples
with very long durations.
That's still desirable for doing seamless cutting of videos,
but needs a rethink for implementation.
https://bugzilla.gnome.org/show_bug.cgi?id=778690
Add a new boolean surround-delay property that makes
audioecho just apply a delay to certain channels to create
a surround effect, rather than an echo on all
channels. This is useful when upmixing from stereo - for example.
Add a surround-mask property to control which channels
are considered surround sound channels when adding a
delay with surround-delay = true
Original patch from Jochen Henneberg <jh@henneberg-systemdesign.com>
This goes around the inefficient control message based filtering and
does all the filtering kernel-side. Unfortunately this is Linux-only and
there is no IPv6 variant of it (yet).
Some radio streams uses StreamTitle='' to reset the title after a
track stopped playing, e.g. while the host talks between tracks or
during news segments.
This change forces an empty tag object to be distributed if
StreamTitle or StreamUrl is received with empty value, thus allowing
downstream elements to get notified about this.
https://bugzilla.gnome.org/show_bug.cgi?id=778437
Upstream elements like videoflip can transform caps, such as changing width and height.
When an imagefreeze downstream receives an ACCEPT_CAPS query it will NOW return
all caps that it can accept.
https://bugzilla.gnome.org/show_bug.cgi?id=778389
Used signed calculations when measuring the max_ts of an input
fragment, so as to calculate the correct duration and offset
when buffers have timestamps preceding their segment
The n_frames field (frames per second) should follow the nominal frame
rate for drop-frame timecodes.
Also, the trak's timescale (and duration, accordingly) should follow the
STSD entry's timescale and frame duration (fps_n and fps_d accordingly),
not the other way around.
https://bugzilla.gnome.org/show_bug.cgi?id=777832
In case wavparse receives a manually injected FLUSH_STOP event
while operating in pull mode we get criticals because we'd try
to clear a NULL adapter.
https://bugzilla.gnome.org/show_bug.cgi?id=777123
Insert VPS/SPS/PPS before the first NAL unit containing an I-frame in an
access unit only. If an access unit consists of several such NAL units
(tiles) VPS/SPS/PPS should only be inserted before the first of them so
that parameters are only updated between frames.
Do not insert VPS/SPS/PPS before P-frames when config-interval is -1.
https://bugzilla.gnome.org/show_bug.cgi?id=775817
qtdemux_handle_xmp_taglist() requires a writable taglist,
but qtdemux->tag_list can become non-writable, specifically
after sending global tags (qtdemux.c:958), which adds a
second reference. Ensure the list is made writable before
calling (make_writable will copy the list if necessary).
https://bugzilla.gnome.org/show_bug.cgi?id=766177
These are usually much bigger than icon size and required by
iTunes to be certain fairly large sizes. In qtmux it is also
the IMAGE tags which we write out as 'covr' atoms.
When reset, don't restart request pad numberings, as
request pads can survive across state changes. Only
restart at 0 if all request pads are handed back first.
https://bugzilla.gnome.org/show_bug.cgi?id=777174
'stream-format' and 'alignment' are defined in pad template caps so
there is no need to check them again here. Also remove bitrate parsing from
caps as bitrate in caps doesn't make sense but from tags, which is
actually the case.
https://bugzilla.gnome.org/show_bug.cgi?id=777181
Needed for QuickTime 7 to properly play files.
Also write the clap atom for MOV files always, not only when ProRes is
used as a video codec. It's mandatory for MOV.
https://bugzilla.gnome.org/show_bug.cgi?id=777100
The seqh buffer allocated in qtdemux_parse_svq3_stsd_data() needs to
be freed by the caller after use.
https://bugzilla.gnome.org/show_bug.cgi?id=777157
Signed-off-by: Andre McCurdy <armccurdy@gmail.com>
If a fragmented stream doesn't have a tfdt, don't
reset the output timestamps at each fragment boundary
by erroneously using the default value of 0. Introduced
by commit 69fc48
https://bugzilla.gnome.org/show_bug.cgi?id=754230
Majorly change the way that splitmuxsink collects
incoming data and sends it to the output, so that it
makes all decisions about when / where to split files
on the input side.
Use separate queues for each stream, so they can be
grown individually and kept as small as possible.
This removes raciness I observed where sometimes
some data would end up put in a different output file
over multiple runs with the same input.
Also fixes hangs with input queues getting full
and causing muxing to stall out.
Add a new signal for formatting the filename, which receives
a GstSample containing the first buffer from the reference
stream that will be muxed into that file.
Useful for creating filenames that are based on the
running time or other attributes of the buffer.
To make it work, opening of files and setting filenames is
now deferred until there is some data to write to it,
which also requires some changes to how async state changes
and gap events are handled.
When performing a key-unit seek, always snap to the start ts
of the keyframe buffer we landed on so that the keyframe is
entirely within the resulting outgoing segment. That seems
the most sensible result, since the user requested snapping
to the keyframe position.
Segments times and seek requests are stored and handled
in raw 'PTS' time, without the cslg_shift - which only applies
to outgoing samples. Omit the cslg_shift portion when
extracting PTS to compare for internal seek snaps.
If the cslg_shift is included, then keyframe+snap-before seeks
generate a segment start/stop time that already includes the
cslg_shift, and it's then added a 2nd time, causing the
first buffer(s) to have timestamps that are out of segment.
Remove an old check from atom_stsc_add_new_entry() that
extends the last entry in the STSC if the samples per chunk
matches, as the new interleave merging logic requires that
the final entry by updateable. There's already code
below which simply merges the final entry into the previous
one when needed, so rely on that instead.
Fixes asserts like:
ERROR:atoms.c:2940:atom_stsc_update_entry: assertion failed:
(atom_array_index (&stsc->entries, len - 1).first_chunk == first_chunk)
Make sure the state of the parser is set to
collecting streams before chaining up to the
parent change_state() method, to close a
small window that can cause playback to
never commence.
Use GQueue instead of a GSList so we don't have to traverse
the whole list to append something every time. And it also
keeps track of the number of items in it for us.
Add a function to add filenames to the list of old files and
use it in more places, so that memory doesn't build up in
other modes either if no max_files limit is specified.
https://bugzilla.gnome.org/show_bug.cgi?id=766991
Technically we weren't leaking the memory, just storing it internally
and never using it until the element is freed. But we'd still use more
and more memory over time, so this is not good over longer periods
of time. Only keep track of files if there's actually a limit set,
so that we will prune the list from time to time.
https://bugzilla.gnome.org/show_bug.cgi?id=766991
Previously, seeking to position y where y is (strictly) within a keyframe
would seek to that keyframe both with SNAP_BEFORE and SNAP_AFTER,
where the latter is now adjusted to really snap to the next keyframe.
Rather amazingly (and equally unnoticed), keyunit seeking resulted in segments
where start != time (which is bogus for simple avi timeline). So, properly
adjust the segment (start) rather than fiddling with segment time (only).
... by using the original seek event's flags rather than the corresponding
segment flags, which do not have such counterpart flags (and
do no longer have them covertly sneaking in nowadays).
With Xiph codecs the stream header buffers are both in the caps and are
usually also at the beginning of each input stream, but it's perfectly
possible that the input stream does not have the stream header buffers
inline in the data. Matroskamux would drop the first N buffers assuming
they're stream headers, but this meant it would drop actual payload data
when the stream didn't contain the stream headers inline. Fix this by
only dropping leading buffers if they're flagged as stream headers. This
fixes issues with streams that are being tapped into after streaming
has started.
https://bugzilla.gnome.org/show_bug.cgi?id=749098
That is, whenever we go through start/stop we have to ensure that on the
next opportunity the buffers are reallocated again. Otherwise the
buffers might be NULL because the element was reused with the same
configuration as before (i.e. set_caps() wouldn't have reinited the
buffers).
https://bugzilla.gnome.org/show_bug.cgi?id=775898
Redirect on PLAY wasn't doing the necessary session cleanup. Fixed by
removing code from gst_rtspsrc_send that changed the state varable upon
encountering a redirect. Better to let the redirect handlers in
gst_rtspsrc_retrieve_sdp and gst_rtspsrc_play do their own
state-dependent cleanup.
https://bugzilla.gnome.org/show_bug.cgi?id=775543
When providing items with a seqnum, there is a (very small) probability
that an element with the same seqnum already exists. Don't forget
to free that item if it wasn't inserted.
And avoid returning undefined values when dealing with duplicate items
We can't simply assume that the length of the tag value as given
inside the stream is correct but should also check against the amount of
data we have actually available.
https://bugzilla.gnome.org/show_bug.cgi?id=775451
qtdemux.c: In function ‘qtdemux_parse_trak’:
qtdemux.c:10184:38: error: format ‘%lu’ expects argument of type ‘long unsigned int’, but argument 9 has type ‘gint {aka const int}’ [-Werror=format=]
GST_DEBUG_OBJECT (qtdemux, "Found jpeg: len %u, need %lu", len,
^
If an element queries the number of retransmission buffers pushed
*while* the push is still taking place (and before the object lock
is taken just after) it would end up with the wrong statistic
being reported.
Increment it just before the push, avoids races when getting statistics
https://bugzilla.gnome.org/show_bug.cgi?id=768723
39f7e52266 was setting the buffer duration
to 0 if is not valid, under the assumption that this is "the last"
buffer and no others are coming next. This is wrong, last_buf is the
previous buffer and not the very last one.
4e3c13c87c was setting DTS to 0 if there
was none. This will set DTS to 0 for all e.g. audio streams, completely
messing up calculations if streams don't start at 0.
https://bugzilla.gnome.org/show_bug.cgi?id=774840
Solves overreading/writing the given arrays and will error out if the
streams asks to do that.
Also does more error checking that the stream is valid and won't
overrun any allocated arrays. Also mitigate integer overflow errors
calculating allocation sizes.
https://bugzilla.gnome.org/show_bug.cgi?id=774859
After finding a cluster id in the byte reader, we skip ahead the reader
position by one further byte to be able to continue searching from there
inside the same chunk if the cluster candidate was a false positive.
We have to accomodate for that additional byte when resuming the search,
otherwise all following pulls are off-by-one for every resume and we run
into an assertion.
bf43f44fcf was comparing an unsigned
expression to be < 0 which was always false.
gstflxdec.c: In function ‘flx_decode_brun’:
gstflxdec.c:322:33: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
if ((glong) row - count < 0) {
^
gstflxdec.c:332:33: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
if ((glong) row - count < 0) {
^
https://bugzilla.gnome.org/show_bug.cgi?id=774834
| ../../../git/gst/isomp4/qtdemux.c: In function 'qtdemux_parse_tree':
| ../../../git/gst/isomp4/qtdemux.c:10224:24: error: 'size' may be used uninitialized in this function [-Werror=maybe-uninitialized]
| offset += size;
| ^~
| ../../../git/gst/isomp4/qtdemux.c:10197:25: note: 'size' was declared here
| guint32 size, tag;
| ^~~~
https://bugzilla.gnome.org/show_bug.cgi?id=774747
Always write an edit list for the whole track. In general this is not
necessary except for the case of having a gap or DTS adjustment but
it allows to give the whole track's duration in the usually more
accurate media timescale.
https://bugzilla.gnome.org/show_bug.cgi?id=774403
splitmuxsink requests pad from element using pad template like "video_%u", "audio_%u" and "sink_%d". This is true for most of the muxers.
But splitmuxsink not able to request pad to flvmux as flvmux has "audio" and "video" as pad templates.
fix: splitmuxsink should fallback to "audio" and "video" when template not found.
https://bugzilla.gnome.org/show_bug.cgi?id=774507
A new signal named on-bundled-ssrc is provided and can be
used by the application to redirect a stream to a different
GstRtpSession or to keep the RTX stream grouped within the
GstRtpSession of the same media type.
https://bugzilla.gnome.org/show_bug.cgi?id=772740
aacparse resizes input buffer while converting ADTS stream to RAW,
During buffer resize buffer write permission is not checked.
This throws gst_buffer_is_writable assertion and leads to AV sync issue some times.
It is corrected by making buffer writeable using gst_buffer_make_writable
https://bugzilla.gnome.org/show_bug.cgi?id=774129
TIME segment implies that stream/running time is being handled by upstream.
So, we shouldn't override it without any clue.
This patch is for fixing seek in DASH streaming.
https://bugzilla.gnome.org/show_bug.cgi?id=774196
The accumulator is filled by intersecting with all the pad caps, as such
it must be initialized with ANY (like it is before the iteration is
started) and not to EMPTY.
Fixes the CAPS query always returning EMPTY caps when resyncing happened
during the query, e.g. because pads were added/removed.
The g_object_unref (saddr) before receiving message seems to be redundant as it
is done just before jumping to retry
Though not directly related, part of
https://bugzilla.gnome.org/show_bug.cgi?id=772841
Control messages are used only in multicast mode - to detect if the destination
address is not ours and possibly drop the packet. However in non-multicast
modes the messages are still allocated and freed even if not used. Therefore
request control messages from g_socket_receive_message() only in multicast
mode.
https://bugzilla.gnome.org/show_bug.cgi?id=772841
Do not use last buffer TS + buffer duration because buffer duration
might be inaccurate, especially for frame rates like 30fps where a
rounding error is observed.
https://bugzilla.gnome.org/show_bug.cgi?id=773785
When doing rtx, the jitterbuffer will always add an rtx-timer for the next
sequence number.
In the case of the packet corresponding to that sequence number arriving,
that same timer will be reused, and simply moved on to wait for the
following sequence number etc.
Once an rtx-timer expires (after all retries), it will be rescheduled as
a lost-timer instead for the same sequence number.
Now, if this particular sequence-number now arrives (after the timer has
become a lost-timer), the reuse mechanism *should* now set a new
rtx-timer for the next sequence number, but the bug is that it does
not change the timer-type, and hence schedules a lost-timer for that
following sequence number, with the result that you will have a very
early lost-event for a packet that might still arrive, and you will
never be able to send any rtx for this packet.
Found by Erlend Graff - erlend@pexip.comhttps://bugzilla.gnome.org/show_bug.cgi?id=773891
The lost-event was using a different time-domain (dts) than the outgoing
buffers (pts). Given certain network-conditions these two would become
sufficiently different and the lost-event contained timestamp/duration
that was really wrong. As an example GstAudioDecoder could produce
a stream that jumps back and forth in time after receiving a lost-event.
The previous behavior calculated the pts (based on the rtptime) inside the
rtp_jitter_buffer_insert function, but now this functionality has been
refactored into a new function rtp_jitter_buffer_calculate_pts that is
called much earlier in the _chain function to make pts available to
various calculations that wrongly used dts previously
(like the lost-event).
There are however two calculations where using dts is the right thing to
do: calculating the receive-jitter and the rtx-round-trip-time, where the
arrival time of the buffer from the network is the right metric
(and is what dts in fact is today).
The patch also adds two tests regarding B-frames or the
“rtptime-going-backwards”-scenario, as there were some concerns that this
patch might break this behavior (which the tests shows it does not).
The new timeout is always going to be (timeout + delay), however, the
old behavior compared the current timeout to just (timeout), basically
being (delay) off.
This would happen if rtx-delay == rtx-retry-timeout, with the result that
a second rtx attempt for any buffers would be scheduled immediately instead
of after rtx-delay ms.
Simply calculate (new_timeout = timeout + delay) and then use that instead.
https://bugzilla.gnome.org/show_bug.cgi?id=773905
Commit 83e718 added a pad template to splitmux request
pads, which means that GstElement now releases the pads on
dispose, but after having removed all elements in the bin
and unlinked them. Make sure we can handle cleanup in that case
without throwing assertions.
https://bugzilla.gnome.org/show_bug.cgi?id=773784
qtdemux.c: In function ‘qtdemux_parse_tree’:
qtdemux.c:10139:16: error: ‘color_table_id’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
if (color_table_id != 0) {
^
qtdemux.c:10121:19: note: ‘color_table_id’ was declared here
guint16 color_table_id;
^~~~~~~~~~~~~~
The ProRes guidelines suggest an interleave of 0.5s is common, but
specifies that for ProRes at most 2MB (for SD) and 4MB (for HD) should
be used per chunk.
It might also make sense to use similar numbers in general.
https://bugzilla.gnome.org/show_bug.cgi?id=773217
Previously we were switching from one chunk to another on every single
buffer. This wastes some space in the headers and, depending on the
software, might depend in more reads (e.g. if the software is reading
multiple samples in one go if they're in the same chunk).
The ProRes guidelines suggest an interleave of 0.5s is common, but
specifies that for ProRes at most 2MB (for SD) and 4MB (for HD) should
be used per chunk. This will be handled in a follow-up commit.
https://bugzilla.gnome.org/show_bug.cgi?id=773217
It's required for ProRes to work with other software.
It is also in the MP4 standard, but inventing values here seems a bit
tricky for the general case and it does not really give any extra
information.
https://bugzilla.gnome.org/show_bug.cgi?id=769048
Some buggy payloaders, e.g. rtph263pay, may use mode B for packets
that starts with a picture (or GOB) start code although it's not
allowed. Let's be nice and not drop these packets/frames.
https://bugzilla.gnome.org/show_bug.cgi?id=773516
Bump the bitstream parsing to TRACE log level so it doesn't flood the
output when trying to read the more useful DEBUG and LOG messages.
Also use GST_DEBUG_OBJECT instead of GST_DEBUG in various places
https://bugzilla.gnome.org/show_bug.cgi?id=773514
Altough commits 6a16be7, 64f9d08 and 0c7e3a8 fixed some issues they
introduced others. This patch fixes the leak of one macroblock for every
B fragment.
Macroblock structures must not be freed immediately after finding the
boundaries as they are stored and used later. However the inital dummy
structure (used for finding the first boundary) must be freed.
CID #1212156https://bugzilla.gnome.org/show_bug.cgi?id=773512
Instead of sending EOS when a source byes we have to wait for
all the sources to be gone, which means they already sent BYE and
were removed from the session. We now handle the EOS in the rtcp
loop checking the amount of sources in the session.
https://bugzilla.gnome.org/show_bug.cgi?id=773218
Improve RFC2326 - chapter C.3 compatibility:
In case just a single stream is specified in SDP and the control attribute
is missing do not drop the stream but rather assume "a=control:*"
https://bugzilla.gnome.org/show_bug.cgi?id=770568
Use the number of milliframes per second for integral and drop-frame
framerates, as suggested by the QT file format specification and other
places. We already did that for integral framerates before, but not for
drop-frame framerates. This now keeps precision better.
For all other framerates, check if it's close to a well-known framerate
and use that instead.
https://bugzilla.gnome.org/show_bug.cgi?id=769041
We consider there's a sifnificant difference when it's larger than on second
or than half the duration of the last processed fragment in case the latter is
larger.
https://bugzilla.gnome.org/show_bug.cgi?id=754230
Modify the caps string to allow width and height greater than 4096.
There is no need to restrict it since the matroska format allows the
width and height values to be up to eight bytes long.
https://bugzilla.gnome.org/show_bug.cgi?id=773582
This solves a hanging mainloop in following scenario:
* connect to source
* network/server drops
* pipeline set to NULL (and connection to flushing as part)
* pipeline set to PAUSED/PLAYING (connection to non-flushing, but not recorded)
* [connecting still not possible]
* pipeline set to NULL => mainloop hangs (since no actual flushing is done)
The pacing of the overall muxing is controlled
by the video GOPs arriving, so we can only handle
1 video stream, and the request pad is named accordingly.
Ignore a request for a 2nd video pad if there's already
an active one.
In file included from ../subprojects/gst-plugins-good/gst/monoscope/gstmonoscope.c:42:0:
../subprojects/gst-plugins-base/gst-libs/gst/audio/audio.h:26:39: fatal error: gst/audio/audio-enumtypes.h: No such file or directory
#include <gst/audio/audio-enumtypes.h>
^
compilation terminated.
https://ci.gstreamer.net/job/GStreamer-master-meson/271/console
Found via the Jenkins CI:
FAILED: subprojects/gst-plugins-good/gst/multifile/gstmultifile@sha/gstsplitmuxsink.c.o
[...]
In file included from ../subprojects/gst-plugins-good/gst/multifile/gstsplitmuxsink.h:24:0,
from ../subprojects/gst-plugins-good/gst/multifile/gstsplitmuxsink.c:59:
../subprojects/gst-plugins-base/gst-libs/gst/pbutils/pbutils.h:30:43: fatal error: gst/pbutils/pbutils-enumtypes.h: No such file or directory
#include <gst/pbutils/pbutils-enumtypes.h>
^
compilation terminated.
https://ci.gstreamer.net/job/GStreamer-master-meson/263/console
If the seek stop point (or start, during reverse play)
was within the segment we just finished, go EOS immediately
instead of proceeding through all other parts and sending
0 length seeks to them.
https://bugzilla.gnome.org/show_bug.cgi?id=772138
When one part moves ahead of the others - due to excessive
downstream queueing, or really small input files - then
we can end up activating parts more than once. That can lead to
effects like shutting down pad tasks prematurely.
https://bugzilla.gnome.org/show_bug.cgi?id=772138
This reverts commit f1ceaab02f.
This broke atomic file writes in "buffer" mode. It did make
sure that any streamheaders are prepended to each file in
buffer mode as well, but that's not really needed in practice,
whereas atomic file writes are, so let's restore the status
quo ante for now since this was primarily a code cleanup anyway,
and if anyone needs to streamheaders in buffer mode too they
can make a patch to implement that differently. Re-implementing
the atomic writes in the element also seems way too much work.
https://bugzilla.gnome.org/show_bug.cgi?id=766990
We were just picking the timestamp of the last buffer pushed into our
adapter before we had enough data to push out.
This fixes things to figure out how large each frame is and what
duration it covers, so we can set both the timestamp and duration
correctly.
Also adds some DISCONT handling.
The basic idea is this:
1. For *larger* rtx-rtt, weigh a new measurement as before
2. For *smaller* rtx-rtt, be a bit more conservative and weigh a bit less
3. For very large measurements, consider them "outliers"
and count them a lot less
The idea being that reducing the rtx-rtt is much more harmful then
increasing it, since we don't want to be underestimating the rtt of the
network, and when using this number to estimate the latency you need for
you jitterbuffer, you would rather want it to be a bit larger then a bit
smaller, potentially losing rtx-packets. The "outlier-detector" is there
to prevent a single skewed measurement to affect the outcome too much.
On wireless networks, these are surprisingly common.
https://bugzilla.gnome.org/show_bug.cgi?id=769768
Assuming equidistant packet spacing when that's not true leads to more
loss than necessary in the case of reordering and jitter. Typically this
is true for video where one frame often consists of multiple packets
with the same rtp timestamp. In this case it's better to assume that the
missing packets have the same timestamp as the last received packet, so
that the scheduled lost timer does not time out too early causing the
packets to be considered lost even though they may arrive in time.
https://bugzilla.gnome.org/show_bug.cgi?id=769768
There is no need to schedule another EXPECTED timer if we're already
past the retry period. Under normal operation this won't happen, but if
there are more timers than the jitterbuffer is able to process in
real-time, scheduling more timers will just make the situation worse.
Instead, consider this packet as lost and move on. This scenario can
occur with high loss rate, low rtt and high configured latency.
https://bugzilla.gnome.org/show_bug.cgi?id=769768
This patch fixes an issue with the estimated gap duration when there is
a gap immediately after a lost timer has been processed. Previously
there was a discrepancy beteen the gap in seqnum and gap in dts which
would cause wrong calculated duration. The issue would only be seen with
retranmission enabled since when it's disabled lost timers are only
created when a packet is received and the actual gap length and last dts
is known.
https://bugzilla.gnome.org/show_bug.cgi?id=769768
Stats should also be collected for unsuccessful packets.
rtx-rtt is very important for determining the necessary configured
latency on the jitterbuffer. It's especially important to be able to
increase the latency when retransmitted packets arrive too late and are
considered lost. This patch includes these late packets in the
calculation of the various rtx stats, making them more correct and
useful.
Also in the case where the original packet arrives after a NACK is sent,
the received RTX packet should update the stats since it provides useful
information about RTT.
The RTT is only updated if and only if all requested retranmissions are
received. That way the RTT is guaranteed to make sense. If not we don't
know which request the packet is a response to and the RTT may be bogus.
A consequence of this patch is that RTT is not updated for a request
when one of the RTX packets for that seqnum is lost, but that since
measured RTT will be more accurate.
The implementation store the RTX information from the timed out timers
and use this when the retransmitted packet arrives. For performance
these timers are stored separately from the "normal" timers in order to
not impact performance (see attached performance test).
https://bugzilla.gnome.org/show_bug.cgi?id=769768
When disabled we can save some iterations over timers.
There is probably an argument for rtx-delay-reorder to exist, but
for normal operations, handling jitter (reordering) is something a
jitterbuffer should do, and this variable feels like functionality that
is not "in-sync" with what the jitterbuffer is trying to achieve.
Example: You have 50ms jitter on your network, and are receiving
audio packets with 10ms durations. An audio packet should not be
considered late until its rtx-timeout has expired (and hence a rtx-event
is sent), but with rtx-delay-reorder, events will be sent pretty much
all the time due to the jitter on the network.
Point being: The jitterbuffer should adapt its size to the measured network
jitter, and then rtx-delay-reorder needs to adapt as well, or simply
get out of the way and let the other (better) rtx-mechanisms do their job.
Also change find_timer to only use seqnum as an argument, since there
will only ever be one timer per seqnum at any given time. In the
one case where the type matters, the caller simply checks the type.
https://bugzilla.gnome.org/show_bug.cgi?id=769768
