SPEECH SYNTHESIZER, SPEECH SYNTHESIZING METHOD, SPEECH SYNTHESIZING PROGRAM, AND COMPUTER READABLE RECORDING MEDIUM WITH SPEECH SYNTHESIZING PROGRAM STORED THEREIN

摘要

PPROBLEM TO BE SOLVED: To solve the problem that, when speech information such as electronic mail is output while a user listens to music etc. with the conventional device in which speech is main output information, text voice synthesizing corresponding to the melody of music under listening is not performed and synthesizing speech overlaps an important part of a song, and thereby music listening is disturbed. PSOLUTION: Melody information of " bright", "dark" and "quick" is extracted from music information, and a rhythm control parameter and a piece control parameter are calculated by a melody synthesizing control means referring to the melody information. The melody synthesizing speech signal of an input text is generated in tune with the melody by a speech synthesizing means on the basis of the rhythm information as the result of analysis of the input text obtained by the rhythm setting means from the rhythm control parameter and the piece control parameter, and synthesized speech with high quality corresponding to the melody in which music listening by the user is not disturbed, is generated. PCOPYRIGHT: (C)2007,JPO&INPIT