| 摘要 |
PROBLEM TO BE SOLVED: To provide technical means for generating a sound production instruction of a voice easily, at an intended timing.SOLUTION: A voice synthesis system 1 is configured so that: a control part 11 of a musical performance support device 10 acquires a breath detection signal which indicates breath, via an ADC 12 from a breath detection sensor 14 disposed on a breath detection part 13 which simulates a mouth piece of a wind instrument. The control part 11 extracts a feature amount of a breath waveform before a modality of the breath waveform which is expressed by the breath detection signal changes, then refers to a phoneme conversion table 162, then creates and outputs phoneme information which indicates a consonant allocated to a pattern of the feature amount of the breath waveform. Then the control part 11 outputs phoneme information which indicates a vowel of voice which is a synthesis object, with a momentum that, change of the modality of the breath waveform expressed by the breath detection signal, is detected.SELECTED DRAWING: Figure 1 |
