Appeal No. 2002-1197 Application 09/131,167 should be kept within a predetermined range that does not cause distortion. However, Koski teaches not updating the filter coefficients based on the condition of double talk detection, not only updating the coefficients based upon the condition that the output signal of the microphone is within a predetermined range of possible output signal values, as claimed. Not updating the filter coefficients based on double talk detection will indirectly keep the output signal of the microphone within a predetermined range of output signal values, but this is not the same as updating based on the output signal value. We agree with appellant's argument "that when double-talk does not exist, the double-talk detector of Koski will not prevent the parameters of the adaptive filter from being updated, regardless of the prevailing value of the output signal of the microphone because Koski discloses freezing the parameters of the adaptive filter only when double-talk exists" (RBr3). Thus, the condition under which Koski freezes the coefficients of the adaptive filter are not based on whether the output signal is within a predetermined range. The examiner has not presented any reasoning that since distortion is caused when the magnitude of the superposed near- end and echo signals is too great, due to double talk, it would have been obvious to one skilled in the art to freeze the coefficients based on the magnitude of the output signal rather than based on the presence of double talk. - 11 -Page: Previous 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 NextLast modified: November 3, 2007