The "(((audio_sample << 1) | 1) << frame->scale_factor[ch][sb])"
part of expression
"frame->sb_sample[blk][ch][sb] =
(((audio_sample << 1) | 1) << frame->scale_factor[ch][sb]) /
levels[ch][sb] - (1 << frame->scale_factor[ch][sb])"
in "sbc_unpack_frame" function can sometimes overflow 32-bit signed int.
This problem can be reproduced by first using bitpool 128 and encoding
some random noise data, and then feeding it to sbc decoder. The obvious
thing to do would be to change "audio_sample" variable type to uint32_t.
However the problem is a little bit more complicated. According
to the section "12.6.2 Scale Factors" of A2DP spec:
scalefactor[ch][sb] = pow(2.0, (scale_factor[ch][sb] + 1))
And according to "12.6.4 Reconstruction of the Subband Samples":
sb_sample[blk][ch][sb] = scalefactor[ch][sb] *
((audio_sample[blk][ch][sb]*2.0+1.0) / levels[ch][sb]-1.0);
Hence the current code for calculating "sb_sample[blk][ch][sb]" is
not quite correct, because it loses one least significant bit of
sample data and passes twice smaller sample values to the synthesis
filter (the filter also deviates from the spec to compensate this).
This all has quite a noticeable impact on audio quality. Moreover,
it makes sense to keep a few extra bits of precision here in order
to minimize rounding errors. So the proposed patch introduces a new
SBCDEC_FIXED_EXTRA_BITS constant and uses uint64_t data type
for intermediate calculations in order to safeguard against
overflows. This patch intentionally addresses only the quality
issue, but performance can be also improved later (like replacing
division with multiplication by reciprocal).
Test for the difference of sbc encoding/decoding roundtrip vs.
the original audio file for joint stereo, bitpool 128, 8 subbands
and http://media.xiph.org/sintel/sintel-master-st.flac sample
demonstrates some quality improvement:
=== before ===
--- comparing original / sbc_encoder.exe + sbcdec ---
stddev: 4.64 PSNR: 82.97 bytes:170495708/170496000
=== after ===
--- comparing original / sbc_encoder.exe + sbcdec ---
stddev: 1.95 PSNR: 90.50 bytes:170495708/170496000
The header files with constants and structures for audio specific
interaction with Pulseaudio are suppose to be under LGPL license.
For some odd reason a2dp-codecs.h ended up being under GPL license
which is against the intention of this being shared and re-used by
non-GPL programs. Fix this now to avoid any future confusion.
This adds a pa_str_in_list() to check for a given string in a
space-separated list of strings. For now, this is merely present to
avoid duplication of role matching code (intended roles can be a
space-separate list) across modules.
The documentation says we expect a comma-separate list of intended
roles, but the code splits the string on whitespaces, so this corrects
the documentation to match the implementation.
The error message for snd_pcm_hw_params_set_period_wakeup was
printing "ret", but "ret" wasn't being set.
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Allow a module argument to specify that we should act globally
rather than just within a given sink.
The default value is to not opporate globally thus retaining the
current behaviour.
Operate on a list of 'trigger roles' and 'cork roles'. i.e.
react to any stream with a role in the trigger list and apply a
cork to any stream with the a role in the cork list.
The trigger roles default to 'phone' and the cork roles default
to both 'music' and 'video' thus achieving the same functionality
as currently when called without any arguments.
When starting via a console login, PA will likely not have a session DBus
to play with. As there is no X11 environment, libdbus will be unable
to launch a session DBus for us and thus the module will fail to load
which in turn prevents PA from loading.
If the user subsequently logs into X11 this it will still not be possible
to load the module as the server will be ignorant of the X11 and DBus
environment variables so a longer term solution for handling this should
be found.
Some of the license wording was less than clear. Try to clarify the
different GPL 'downgrade' scenarios but also be generic to ensure that
those packagers where GPL is a problem check thoroughly before they ship.
Inspired by comments from Brian Cameron @ Oracle via fdo#41822
I'd rather not have to do this, as I don't really see the point in
duplicating what is done in pkgconfig, but this is likely the
easiest way to avoid nasty hacks.
pa_sink/source_used_by() ignores corked/monitor streams, but we need to
make sure there aren't any of these while updating rate (at least for
now -- this is a restriction that would be nice to get rid of).
Basically adds code to handle passthrough sources. This isn't a tested
path at the moment, but in the future, when we do wish to support these,
it'll save us the trouble of having to sync all the code again.
This is where the actual changes happen.
Some additional checks would be required to make sure the
rate is actually supported
Tested with both PCM and passthrough streams
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Avoid resampling or use integer resampling when supported by the
sinks/sources
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This adds the WebRTC echo canceller as another module-echo-cancel
backend. We're exposing both the full echo canceller as well as the
mobile echo control version as modargs.
Pending items:
1. The mobile canceller doesn't seem to work at the moment.
2. We still need to add bits to hook in drift compensation (to support
sink and source from different devices).
The most controversial part of this patch would probably be the
mandatory build-time dependency on a C++ compiler. If the optional
--enable-webrtc-aec is set, then there's also a dependency on libstdc++.
The new module argument can be used to provide a custom
directory for loading alsa path configuration files. This is
useful for testing: no need to be root to create test
configuration files.
module-stream-restore and modile-filter-apply can get into an infinite
loop if m-s-r is called before m-f-a (m-s-r rescues a stream and
attaches it to a sink/source, which then triggers m-f-a to move it back
to the filter sink/source, and so on). The purpose of the m-f-a hooks is
to beat m-s-r, so moving them to be run first.
This removes the nasty side-effect that a call to
pa_namereg_get_default_{source,sink}() will also *set* the default
source/sink.
This is a more complete fix for commit 766dbc68 ("conf: Make sure
module-dbus-protocol is loaded after module-default-device-restore")
https://bugs.freedesktop.org/show_bug.cgi?id=40897
This makes sure that we only perform any processing (resync or actual
cancellation) after the source provides enough data to actuall run the
canceller.
When a source-output isn't connected to our virtual source, we skip echo
cancellation altogether. This makes sense in general, and makes sure
that we don't end up adjusting for delay/drift when nothing is
connected. This should make convergence faster when the canceller
actually starts being used.
This increase the threshold for difference between the playback and
capture stream before samples are dropped from 1ms to 5ms (the
cancellers are generally robust to this much and higher). Also, we make
this a module parameter to allow easier experimentation with different
values.
