| 摘要 |
In order to efficiently use processing and transmission bandwidth and data storage of a computer system, video data is represented using integer and fractional values. The integer value has a precision defined by the precision of the data paths of the computer system. These integer and fractional components are packed into byte-oriented data packets in a manner that minimizes waste of storage space and transmission bandwidth. This packing of data also may be done in such a way so as to minimize processing for performing packing and unpacking of the data. Because the video data may be easily separated and combined into its integer and fractional components, these components may be processed or transported separately, in parallel or in series, and then later recombined. As a result, lower precision devices may be used in parallel to process or transport streams of higher precision data without having a high precision data processing or transport path. |
