| 摘要 |
The method has a precise and natural synthesizing voice by the overlay of formant contour. The method includes the 1st step which initializes the formant index, and reads the formant information from the formant base, the 2nd step which overlaps between the starting formant contour and the end formant contour, the 3rd step which searches the overlay of overlapped formant, the 4th step which switches the starting formant contour, the 5th step which smoothes 2 frame area of the switching part, the 6th step which forms the formant contour, and the 7th step which generates the one syllable formant contour.
|
