| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a singing data generating program capable of generating singing data with which a natural human voice can be generated. <P>SOLUTION: For example, when it is decided in a pitch generation step that there is no sound for a specified time before one piece of note data, the pitch of the note data is set to be lower than the original pitch of the note data by a specified quantity at its sounding start time and then increased in steps to the original pitch. Based upon the generated data, singing data can be sounded as a natural human voice. <P>COPYRIGHT: (C)2005,JPO&NCIPI |
