| 摘要 |
A virtual memory system for a video editor is disclosed which permits edited material from discrete takes to be electronically spliced in real time to form edited scenes. Typically, a plurality of cameras take film of originally acted scenes, either simultaneously from different angles or sequentially of differing takes of the same scene material replayed by skilled film actors. The film material is thereafter recorded from all uncut takes to a random access memory, preferably a DRAW laser video disk. An intelligent interface-preferably a microprocessor-connects at least two and preferably three laser disk heads for reading the video frames of the same uncut takes. Each disk head is provided with and reads the identical disk material containing identical unedited takes. An intelligent interface addresses a video disk head to randomly access the uncut takes and cause playing at controlled speeds forwardly, backwardly or still frame. A controlling computer addresses the intelligent interface and maintains and immediately updates a frame address list of edited takes including the first frame address of each take, and the last frame address of each take. All takes are maintained in the specified serial order of the edited scene. For immediate review at any time of an edited scene, the controlling computer commands the intellignet interface to play the current edited take and positions one player on the first frame of the next sequential take-this frame being the first frame of the next take in forward motion or the last frame of the previous take for reverse motion.
|
