摘要 |
<p>PURPOSE:To obtain high synthesized voice while securing real time performance by measuring the processing time of voice synthesis and varying the order of the synthesis filtering in accordance with the processing time. CONSTITUTION:Voice output completion information of a voice output section 7 is monitored by a processing time measuring section 8. Then, the remaining time or the delay time of the voice synthesis process of a voice synthesis section 6 is measured. In accordance with the measured processing time, the order of phoneme parameters is decided by a mode control section 10. Based on the phoneme parameters generated by a synthesis parameter generation section 5 and rhythm parameters, the synthesis filtering having the decided order is executed by the section 6 and synthesized voice is generated.</p> |