| 摘要 |
PURPOSE:To reduce the required scale of a storage device by storing time information, an input part identifier, and a memory address as control information for each voiced section of voice data on input voices from plural voice input parts and storing only voiced section voice data as voice data. CONSTITUTION:Voiced section data are extracted from the inputted voice data by special voice data voiced sound extraction parts 110 and stored in a voice data storage part 1232; and storage memory address information of the voice section data and time information indicating the start time of voiced section data are stored in a voiced section control information storage pat 1231 for each the identifier of a voice data input part. Then the time information and storage memory address information of the input part identifier are read out and when they are put back into the contents of the time information, voiced section data is read out according to the storage memory address information and supplied to voice output parts 130, which output voices. Consequently, voice data of only a specific speaker can be outputted among input voice data from plural speakers without increasing the hardware quantity. |
