Time scale modification method and system using Dual-SOLA algorithm

摘要

PURPOSE: A time scale modification method and system using a dual SOLA algorithm are provided to generate less clicking and reverberation. CONSTITUTION: A voice signal is multiplied by a window to be segmented into frames each of which has a predetermined length(S510). A synthesis shift value of an output signal fixed to a predetermined value is selected according to rate variation rate of the frames, and an analysis shift value of an input signal is changed to obtain the next frame(S520). The frames are moved by a synchronization length obtained according to cross-correlation in the overlap section of the frames to rearrange the frames(S530). A weight is added to a frame sample overlapped in the overlap section, to output a time-scale-modified voice signal(S540).