摘要 |
PROBLEM TO BE SOLVED: To enable a user to select arbitrary and various synthetic voice generation by generating voice dictionary based on human voice, storing the dictionary in a hard recording medium and supplying it to the user. SOLUTION: The result of the reading of a voice provider 1 is recorded in a voice recording device 2. Based on the result, a voice dectionary is generated. Then, the result is stored in a computer 3 and the dictionary is stored in a CD-ROM 4. The CD-ROM 4 is supplied to the user, and the dictionary, which is stored in the CD-ROM 4, is installed in a user's personal computer 5. Moreover, the dictionary and an information table are stored in a server 6 and the user connects the computer 5 to the server 6 through a public telephone line and may install an arbitrary dictionary in the user's personal computer 5. Based on the text analysis obtained from a manuscript, the user refers to a voice element dictionary, connects various voice elements complements the connecting sections and generates phoneme sequence. The rhythm generated by referring to a phoneme sequence and a rhythm dictionary, are synthesized to produce synthesis data. |