In practice, we have been doing this since time immemorial, but have
relied on the user to do the downmixing (first voice engine then
Chromium). It's more logical for this burden to fall on AudioProcessing,
however, who can be expected to know that this is a reasonable approach
for AEC. Permitting two render channels results in running two AECs
serially.
Critically, in my recent change to have Chromium adopt the float
interface:
https://codereview.chromium.org/420603004
I removed the downmixing by Chromium, forgetting that we hadn't yet
enabled this feature in AudioProcessing. This corrects that oversight.
The change in paths hit by production users is very minor. As commented
it required adding downmixing to the int16_t path to satisfy
bit-exactness tests.
For reference, find the ApmTest.Process errors here:
https://paste.googleplex.com/6372007910309888
BUG=webrtc:3853
TESTED=listened to the files output from the Process test, and verified
that they sound as expected: higher echo while the AEC is adapting, but
afterwards very close.
R=aluebs@webrtc.org, bjornv@webrtc.org, kwiberg@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/31459004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@7292 4adac7df-926f-26a2-2b94-8c16560cd09d
This attenuates the noise pumping generated from the NS adapting to the AEC comfort noise.
When there is echo present the AEC suppresses it and adds comfort noise. This is underestimated on purpose to avoid adding more than the original background noise. The NS has to be called after the AEC, because every non-linear processing before it can ruin its performance. Therefore the noise estimation can adapt to this comfort noise, making it less aggressive and generating noise pumping.
By putting the noise estimation analysis stage from the NS before the AEC, this effect can be avoided. This has been tested manually on recordings where noise pumping was present: Two long recordings done in our team by bjornv and kwiberg plus the most noisy (5) recordings in the QA set.
On the other hand, one risk of doing this is to not adapt to the comfort noise and therefore suppress too much. As verified in the tested files, this is not a problem in practice.
BUG=webrtc:3763
R=andrew@webrtc.org, bjornv@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/24679004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@7289 4adac7df-926f-26a2-2b94-8c16560cd09d
In r6591 a shift macro was removed affecting AECM. In addition to that change a bug was fixed. The fix added a few voice_counts in ApmTest.Process.
This CL updates the reference file, even though it is not used in practice since the test is currently turned off for Android (where AECM is used).
BUG=3672
TESTED=locally
R=andrew@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/14089004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@6868 4adac7df-926f-26a2-2b94-8c16560cd09d
Internally, it already worked on floats. This patch just changes the
signature of a bunch of functions so that floats can be passed
directly from the new and improved AudioBuffer without converting the
data to int and back again first.
(The reference data to the ApmTest.Process test had to be modified
slightly; this is because the noise suppressor comes immediately after
the echo canceller, which also works on floats. If I truncate to
integers between the two steps, ApmTest.Process doesn't complain, but
of course that's exactly the sort of thing the float conversion is
supposed to let us avoid...)
BUG=
R=aluebs@webrtc.org, bjornv@webrtc.org, tina.legrand@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/13519004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@6385 4adac7df-926f-26a2-2b94-8c16560cd09d
Each audio processing step is given a pointer to an AudioBuffer, where
it can read and write int data. This patch adds corresponding
AudioBuffer methods to read and write float data; the buffer will
automatically convert the stored data between int and float as
necessary.
This patch also modifies the echo cancellation step to make use of the
new methods (it was already using floats internally; now it doesn't
have to convert from and to ints anymore).
(The reference data to the ApmTest.Process test had to be modified
slightly; this is because the echo canceller no longer unnecessarily
converts float data to int and then immediately back to float for each
iteration in the loop in EchoCancellationImpl::ProcessCaptureAudio.)
BUG=
R=aluebs@webrtc.org, andrew@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/18399005
git-svn-id: http://webrtc.googlecode.com/svn/trunk@6138 4adac7df-926f-26a2-2b94-8c16560cd09d
Break out the computation to a separate class, and call directly into
this from channel.cc rather than going through AudioProcessing. This
circumvents AudioProcessing's sample rate limitations.
We now compute the RMS over all samples rather than downmixing to a
single channel. This makes the call point in channel.cc easier, is
more "correct" and should have similar (negligible) complexity.
This caused slight changes in the RMS output, so the ApmTest.Process
reference has been updated. Snippet of the failing output:
[ RUN ] ApmTest.Process
Running test 4 of 12...
Value of: rms_level
Actual: 27
Expected: test->rms_level()
Which is: 28
Running test 5 of 12...
Value of: rms_level
Actual: 26
Expected: test->rms_level()
Which is: 27
Running test 6 of 12...
Value of: rms_level
Actual: 26
Expected: test->rms_level()
Which is: 27
Running test 10 of 12...
Value of: rms_level
Actual: 27
Expected: test->rms_level()
Which is: 28
Running test 11 of 12...
Value of: rms_level
Actual: 26
Expected: test->rms_level()
Which is: 27
Running test 12 of 12...
Value of: rms_level
Actual: 26
Expected: test->rms_level()
Which is: 27
BUG=3290
TESTED=Chrome assert is avoided and both voe_cmd_test and apprtc
produce reasonable printed out results from RMS().
R=bjornv@webrtc.org
Review URL: https://webrtc-codereview.appspot.com/16459004
git-svn-id: http://webrtc.googlecode.com/svn/trunk@6056 4adac7df-926f-26a2-2b94-8c16560cd09d
