| 摘要 |
A method is proposed for eliminating or suppressing unwanted acoustic signals in audio programs to be edited for reproduction in studios, in which the perceived disrupted segments in the signal are determined with respect to their time positions, and the signal ends having signal flow/polarity changes adjacent to the disrupted signal location are bridged by a transfer function, preferably in the shape of a semi-sine. The semi-sine is stored in the form of sampling values in an addressable read-only memory and can be read therefrom. In order to generate the desired transfer function, the signals read from the read only memory are uniformly distributed over the transfer function in accordance with the plurality of sampling values to be replaced across the width of the disrupted segment, the read signals being weighted in accordance with the margin conditions of the signal ends adjacent to the disrupted segment and being inserted in the proper sequence into the signal in the area of the original disrupted segment upon consideration of the direction of signal change.
|
