In computerized cartoon manufacturing system, many researches have been focused on computer graphical themes. Noting,however that computerized cartoon also requires sound tracks, the synchronization of voices and motions pictures are important. In this thesis, a voice modification systems are proposed, which changes the time scale of voice signals so that it fits into the motion picture. Various different characteristics of voiced and unvoiced speech signal has been tested and applied for time-scale modification of speech signals.
This systems can also be applied to computer games. Many existing computer games exhibit different animation displaying performances on different computer systems due to the dissimilarity computers processing speed. By using the proposed system, speech segments of a character can be elongated or shrunk to synchronize them with its motion, and mouth movement. By using of the proposed technique, several sample animation file based on Windows™ have been generated and compared.