next up previous
Next: Implementation Up: About Adding Voices to Previous: About Adding Voices to

Background Information

By last year, there were 10700 English words which could be looked up in the Hyper Dictionary, and more words should be added to it. A new list of 12813 words was distributed to the members of our laboratory, but it had mistakes in spelling, and some words in it overlapped with words which had already been exhibited in the Hyper Dictionary. Since voices must be obtained from a native English speaker to obtain genuine pronunciation of English words, a list of words which would be inserted into the latest Hyper Dictionary needed to be made as soon as possible. Hence, the first step in conducting this research was to make the list. Then the list which was constructed by checking overlaps and mistakes between the new 12813-word list and the old one, and it was sent to an English teacher at the University of Aizu to record his native voice on computer. The voice was made as one WAV file, a music file format, and submitted to me by my supervisor. However, the file had to be split into several files so that each file had only one pronunciation since the WAV file included pronunciation of many English words. It was conducted using free software, spwave for Windows[*], My Wave Cutter[*], and a shareware ALLREC SOUND cutter[*]. In addition, they were converted to MP3 files with CDex[*], another music file format, and it is generally said that MP3 files have smaller volumes than WAV files, so they can be reproduced more efficiently.
next up previous
Next: Implementation Up: About Adding Voices to Previous: About Adding Voices to
root 2004-03-02