| 摘要 |
An audio conferencing system in which each participant in the conference generates a monaural signal for transmission on an output channel (42) to the other participants, and receives a plurality of input channels (411, 412, 413) from the other participants, the input channels (411, 412, 413) being individually converted to stereophonic signal pairs by respective filter banks having selected predetermined transfer functions (81, 82, 83), and the stereo channels are combined (92) linearly to provide a stereophonic audio output (22L, 22R). The system has an echo cancellation system in which an adaptive filter process (62) is applied to the input channels (411, 412, 413) to generate a combined echo cancellation signal for applying to the output channel (42). A stereo adaptive filter (62) is applied to the combined input channel pair to generate an echo cancellation signal for applying to the output channel (42). Correlations between the stereophonic signal pair are allowed for by controlling the adaptation process based on knowledge of the predetermined transfer functions (81, 82, 83). A separate adaptive filter (621, 622, 623) may be provided for each input channel (411, 412, 413), (either before or after conversion to stereo pairs), the outputs of the filters being combined (91) to form the cancellation signal.
|
