| 摘要 |
PURPOSE:To decide interval data with high accuracy and high efficiency by regularly synthesizing plural test clauses with various timing point intervals between syllables, temporarily hearing the synthesized clause and deciding the optimum hearing timing point interval. CONSTITUTION:S1-S5 are sound waveforms for the syllable NA, KA, TA, DE, SU (the square form of Japanese syllabary) of a regularly synthesized sound, and the timing point intervals of S1-S2, S3-S4 and S4-S5 are respectively fixed to a prescribed value. Then, the several number of the test clause sounds 'NA, KA, TA, DE, SU' stretching the interval of S2-S3 to the prescribed value are prepared. Afterwards, the timing point interval of S2-S3 is decided by selecting the test clause in which the interval between 'NA' and 'KA' is naturally heard. This decided timing point interval is filed as optimum data and used for really synthesizing the sound. Even concerning the other clause, the optimum timing point interval is similarly filed. Thus, since the timing point interval at every following clause can be easily decided, rhythm control can be executed by data with high accuracy. |
