摘要 |
PURPOSE:To obtain a natural smooth speech signal without requiring the interpolation calculation of a filter coefficient when a sound source encoded by using speech compression is decoded by using an LPC synthesizing filter. CONSTITUTION:The LPC synthesizing filter 2 consists of two parts 21 and 22 and when their outputs are denoted as yA and yB, the output of the LPC synthesizing filter 2 is set to (1-alpha)yA+alphayB (alpha is varied sequentially from 0 to 1 by an equal value in the time of the decoding of multiple pulses of one frame). |