| 摘要 |
The replacement values for a set of samples to be corrected, occurring in a sequence of samples of, for example, an audio signal, are generated with the aid of valid samples in a remote pre-environment of the relevant set of samples to be restored. In such a remote pre-environment of the relevant set of samples to be restored a set of valid samples is namely chosen whose environment, but for a scaling factor, approximates as closely as possible the environment of the set of samples to be restored. Such a restoration technique has the practical advantage that the sequence of valid samples subsequent to the set of samples to be restored, which sequence is necessary to determine the environment thereof, can be relatively short. |
