| 摘要 |
<p>PURPOSE:To attain fine sound editing by performing the editing job while confirming sentences, paragraphs, etc. or editing the sound data on the basis of a block unit after calculating the feature quantity from the sound data and displaying the time series of feature quantity. CONSTITUTION:For instance, a sentence ''Please give me your answer at once about this matter.'' including ETO (well) and GOHON (hack) is fed. A keyboard is operated together with use of a cursor on a display screen to ''indicate'' both start and end parts of undesired sounds ETO and GOHON (waveforms b and d) and to indicate ''deletion'' of these undesired sounds. Thus the waveforms (b) and (d) are deleted. Then the part of SHIKYU (at once) shown by a waveform (f) is indicated, and then the part between TSUKIMASHITEWA (about) and HENJIO (answer) is indicated. Then ''insertion'' is indicated through the keyboard. In other words, the waveform (f) is put between waveforms (a) and (c). Then the head and the end of a sound part is indicated and then ''output'' is indicated. Thus the edited sounds ''Please give me your answer at once about this matter.'' are delivered through a speaker.</p> |
