摘要 |
PURPOSE:To generate a sound synthesized waveform with a small number of discontinuous points by deciding a repetition section, and weighting respective phonemes pieces right before and after the section as required and superposing them. CONSTITUTION:A waveform element piece segmentation part 12 segments waveform element pieces of a sound in each section from digital sound data stored in sound data storage part 11 and when a repetition decision part 13 decides the repetition of the section according to the element pieces, the waveform element pieces in sections right before and after the repetition section are stored in corresponding buffers B and A of a pitch waveform element piece synthesized part 14. Then the element pieces in the buffer B are weighted with weight which decreases with time and the element pieces in the buffer A are weighted with weight increasing with time; and they are superposed on over the other and the sound composite element pieces are outputted from a synthesized part 14 to generate a sound synthesized waveform having a small number of discontinuous for each section. |