Trick mode VTR which generates trick play data from a stream of images containing intra-pictures and predictive pictures and selects specific DCT coefficients for intra-pictures

摘要

Tape formatting, tracking and trick mode data generation are disclosed for providing trick modes for compressed video stored on a video tape. The tape is formatted so that trick mode data is recorded at normal speed within certain areas delineated by predetermined boundaries that are within a maintainable range of a predetermined ideal scan path of the heads. These boundaries are oblique to the normal speed tracks and define trick mode data areas in which trick mode data may be recorded. To facilitate tracking, sync words and trick mode identification codes are interleaved within the trick mode data areas. During tracking, the heads reproduce the sync words and identification codes, and the identification codes that match the currently utilized trick mode are counted. The count during a current interval is compared to a count during a previous interval or a threshold to determine if the heads have locked onto the ideal scan path or deviated therefrom. Trick mode data is generated by replicating a limited portion of the compressed video data of selected pictures, which trick mode data can be independently reproduced, decoded and presented during trick mode playback.