| 摘要 |
By data every a predetermined number of discrete frequencies obtained by implementing orthogonal transform processing to acoustic signal data by using the same window function for first and second frames in sequential frames of an acoustic signal having a predetermined fixed time length, phase information every respective discrete frequencies are obtained every the first and second frames. Then, quantities of changes of phase information every discrete frequencies corresponding to each other in the first and second frames are determined. On the assumption that the quantities of changes of phase information every individual discrete frequencies are fixed on the time base, individual phase information every a predetermined number of discrete frequencies in a frame at a third time position existing at a time position of a multiple of integer of a time difference between the first and second time positions are determined to predict phase information of the frame at the third time position. Thus, it is possible to efficiently encode an acoustic signal to facilitate it to satisfactorily record or transmit it, or to easily provide a sound source for an electronic musical instrument excellent in performance.
|
