| 摘要 |
A recorder for storage and retrieval of digitized signals on a single track of a suitable record medium is disclosed. The recorder includes networks for partitioning digitized audio signals into successive frames, each of which preferably contains a plurality of alternate data words and parity words, followed, at the end of each frame by an error check code word and a sync word. The parity words of a given frame are generated from data words of other frames sufficiently spatially remote from the given frame to minimize the probability that a single drop out or other defect in a record medium will, upon playback, cause the loss of data associated with both the given frame and the other frames. During playback, each frame is checked for errors, and if found, corrected data words are reconstructed from the parity words and data words of other frames located both ahead and behind the erroneous frame, and are inserted in the data stream in place of the erroneous data words. |
