SndCardStream VoiceEngine Rendering Capturing NeedMorePlayData() [in] speech [in] samples [out] fs [out] stereo RecordedDataIsAvailable() [out] speech [out] delay [out] fs [out] skew [in] newMicGain SndCardStreamObserver SndCard expects the mixer output to be at the given sample rate and stereo mode. Conversion takes place in the mixer. SndCard always generates 10ms buffers at the specified sample rate. Conversion must take place before encoding. The data flow is driven by the sound card (thread(s)) using callback functions. An additional API is required if an external source shall act as driver. Current client audio implementation newMicGain fs stereo