[This CL is a rebase of an original CL by solenberg@:
https://codereview.webrtc.org/2948763002/ which in turn was a
rebase of an original CL by peah@:
https://chromium-review.googlesource.com/c/527032/]
Allow an external audio processing module to be used in WebRTC
This CL adds support for optionally using an externally created audio
processing module in a peerconnection. The ownership is shared
between the peerconnection and the external creator of the module.
As part of this the internal ownership of the audio processing module
is moved from VoiceEngine to WebRtcVoiceEngine.
BUG=webrtc:7775
Review-Url: https://codereview.webrtc.org/2961723004
Cr-Commit-Position: refs/heads/master@{#18837}
const int16_t* data() const;
int16_t* mutable_data();
- data() returns a zeroed static buffer on muted frames (to avoid unnecessary zeroing of the member buffer) and directly returns AudioFrame::data_ on unmuted frames.
- mutable_data(), lazily zeroes AudioFrame::data_ if the frame is currently muted, sets muted=false, and returns AudioFrame::data_.
These accessors serve to "force" callers to be aware of the mute state field, i.e. lazy zeroing is not the primary motivation.
This change only optimizes handling of muted frames where it is somewhat trivial to do so. Other improvements requiring more significant structural changes will come later.
BUG=webrtc:7343
TBR=henrika
Review-Url: https://codereview.webrtc.org/2750783004
Cr-Commit-Position: refs/heads/master@{#18543}
There's no longer any need to make the two arguments have the same
signedness, so we can remove a bunch of superfluous (and sometimes
dangerous) casts.
It turned out I also had to fix the safe_cmp functions to properly handle
enums that are implicitly convertible to integers.
NOPRESUBMIT=true
BUG=webrtc:6645
Review-Url: https://codereview.webrtc.org/2534683002
Cr-Commit-Position: refs/heads/master@{#15281}
There's no longer any need to make the two arguments have the same
signedness, so we can drop the "u" suffix on literal integer
arguments.
NOPRESUBMIT=true
BUG=webrtc:6645
Review-Url: https://codereview.webrtc.org/2535593002
Cr-Commit-Position: refs/heads/master@{#15280}
This CL re-routes audio through AudioMixer instead of AudioConferenceMixer.
This is done without any modifications to VoiceEngine.
Previously, output audio was polled by an AudioDevice through an AudioTransport
pointer, which was an instance of VoEBaseImpl. VoiceEngineImpl sent the
request for data on to OutputMixer and further to AudioConferenceMixer.
This CL changes the audio flow to an AudioDevice. We reconfigure the AudioDevice
to have another AudioTransport pointer, which points to an AudioTransportProxy.
The AudioTransportProxy is responsible for feeding mixed data to the
AudioProcessing component for echo cancellation, and to resample the audio data
after AudioProcessing and before it is sent to the AudioDevice.
The set up of the audio path was previously done during VoiceEngine
initialization. Now it is changed in the AudioState constructor.
This list shows where audio-path-related VoiceEngine functionality has been
moved:
OutputMixer --> AudioTransportProxy
VoiceEngineImpl --> AudioState, AudioTransportProxy
SharedData --> AudioState
Channel --> AudioReceiveStream, ChannelProxy, Channel
AudioState owns the new mixer and connects it to AudioTransport and
AudioDevice on initialization.
The audio input source is AudioReceiveStream, which registers itself with the
mixer (which it gets from AudioState) on Start and Stop.
# Since the AudioTransport interface contains non-const references.
NOPRESUBMIT=True
BUG=webrtc:6346
Review-Url: https://codereview.webrtc.org/2436033002
Cr-Commit-Position: refs/heads/master@{#15193}
All audio in calls is now routed through AudioTransportProxy. The
AudioTransport implemented by VoEBaseImpl is disconnected from
AudioDevice and replaced by an empty proxy layer that forwards calls
to the old Transport. This is a refactoring CL in preparation for
landing https://codereview.webrtc.org/2436033002/, which will connect
the new AudioMixer.
In the planned configuration, the currently empty AudioTransportProxy
will query the new mixer for audio instead of polling data from the
old Transport. Mixed audio will be passed to an AudioProcessing
interface. AudioTransportProxy is initialized with an AudioProcessing*,
which is currently unused.
No presubmit since we implement an interface with non-const references.
NOPRESUBMIT=True
BUG=webrtc:6346
Review-Url: https://codereview.webrtc.org/2454373002
Cr-Commit-Position: refs/heads/master@{#15133}
