| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide a song synthesis system which synthesizes a song by applying the same melody to parts having similar word sequence structures in words created by a user. <P>SOLUTION: A CPU splits text data DTX<SB POS="POST">1-8</SB>into text data DTX<SB POS="POST">1-2</SB>, DTX<SB POS="POST">3-4</SB>, DTX<SB POS="POST">5-6</SB>and DTX<SB POS="POST">7-8</SB>. Among musical performance data, musical performance data as respective melodies for words represented by the text data DTX<SB POS="POST">1-2</SB>, DTX<SB POS="POST">3-4</SB>, DTX<SB POS="POST">5-6</SB>and DTX<SB POS="POST">7-8</SB>is selected to generate an audio signal by combining the text data DTX<SB POS="POST">1-2</SB>, DTX<SB POS="POST">3-4</SB>, DTX<SB POS="POST">5-6</SB>and DTX<SB POS="POST">7-8</SB>with the respective melodies. When the text data DTX<SB POS="POST">1-2</SB>, DTX<SB POS="POST">3-4</SB>, DTX<SB POS="POST">5-6</SB>and DTX<SB POS="POST">7-8</SB>contains similar word sequence structures, the same musical performance data is selected. <P>COPYRIGHT: (C)2013,JPO&INPIT |
