Multichannel AEC is an acoustic echo cancellation solution which interfaces with multiple loudspeakers and/or multiple microphone signals. There are two main use cases of multichannel AEC which have fundamental differences in the AEC algorithm structure.
The first use case is for large conference room applications. In this scenario, there are multiple loudspeakers in the space, but they are driven by a mono output signal. The microphones are distributed throughout the conferencing space.
The second use case is for multichannel home audio systems. A soundbar capable of 5.1 channel audio is receiving 5 independent loudspeaker signals, while there is only a single microphone signal of interest.
Both scenarios require multiple adaptive filters. In the multi-mic scenario, the loudspeaker signal is shared among each channel of AEC. The filter update is performed independently on each microphone signal. Every adaptive filter has their own doubletalk detector (DTD) and variable stepsize control based on the signals observed by their respective microphone signal.
In the soundbar scenario, the predicted echo signal from each adaptive filter is subtracted from the same microphone signal. Therefore, the DTD and the error signals are shared for the filter update.
VOCAL’s Multichannel AEC software can support both multichannel architectures effectively and efficiently.
VOCAL’s Multichannel AEC software can support both multichannel architectures effectively and efficiently. Please contact us for specific Echo Cancellation supported platforms.