| 摘要 |
PROBLEM TO BE SOLVED: To reduce degradation of subjective quality in voice frames after a lost frame when a voice frame is lost in the changeover of a voice mode. SOLUTION: In this voice decoding device, whether a current frame corresponds to a lost frame or not is determined (ST 1100), and when the current frame is not the lost frame, whether a frame in front of the lost frame corresponds to a lost frame or not is judged(ST1300). When the front frame is the lost frame, whether there is a frame class changeover from a noise-signal to a voiced regular signal or not is confirmed(ST1400). When there is the frame class changeover, a decoding processing C(Overlap Add processing ) is performed to this frame(ST1500). COPYRIGHT: (C)2006,JPO&NCIPI
|
