SPEECH DECODER, SPEECH DECODING METHOD, PROGRAM, RECORDING MEDIUM

摘要

A code separation/decoding unit restores a vocal tract characteristic sp 1 and a vocal source signal r 1 . A vocal tract characteristic modification unit modifies the vocal tract characteristic sp 1 and outputs the modifiedvocal tract characteristic sp 2 . In this method, an emphasized vocal tract characteristic sp 2 is generated to output by applying formant emphasis directly to the vocal tract characteristic sp 1 for instance. A signal synthesis unit synthesizes the modified vocal tract characteristic sp 2 and the vocal source signal r 1 to generate and output an output voice, s.