Abstract: A computer-implemented method, apparatus, and article of manufacture provide the ability to store image data for use in a real-time compositing computer application. A storage tile size is determined based on disk input/output (I/O) hardware testing. A processing tile size is determined based on graphics I/O testing. An image is obtained and processing tiles (of the processing tile size) are mapped over the image. A center of the image is used as a point of origin for the processing tiles. The processing tiles are mapped to storage tiles. The storage tile point of origin is located at a lower left corner of the processing tiles. Each storage tile is configured to contain complete processing tiles that are stored in the storage tiles based on the storage tile size, processing tile size, and the mappings.

Abstract: A method, apparatus, and article of manufacture for traversing a process tree in a digital video editing system. A process tree is loaded into a digital video-editing system. The process tree has parent nodes that are dependent on output data from child nodes. Each parent node and each child node represents a digital video processing task to be performed during a traversal of the process tree. Each parent node declares a data definition and data default values. The data definition defines a structure of the data used by that parent node. For each parent node processed, a determination is made regarding whether the data definition for the output data received from a dependent child node is compatible with the data definition declared for that parent node. If the data definition is incompatible, the parent node utilizes the data default values.

Abstract: A method, apparatus, and article of manufacture for traversing a process tree in a digital video editing system. A process tree is loaded into a digital video-editing system. The process tree has parent nodes that are dependent on output data from child nodes. Each parent node and each child node represents a digital video processing task to be performed during a traversal of the process tree. Each parent node declares a data definition and data default values. The data definition defines a structure of the data used by that parent node. For each parent node processed, a determination is made regarding whether the data definition for the output data received from a dependent child node is compatible with the data definition declared for that parent node. If the data definition is incompatible, the parent node utilizes the data default values.

Abstract: A computer-implemented method, apparatus, and article of manufacture provide the ability to store image data for use in a real-time compositing computer application. A storage tile size is determined based on disk input/output (I/O) hardware testing. A processing tile size is determined based on graphics I/O testing. An image is obtained and processing tiles (of the processing tile size) are mapped over the image. A center of the image is used as a point of origin for the processing tiles. The processing tiles are mapped to storage tiles. The storage tile point of origin is located at a lower left corner of the processing tiles. Each storage tile is configured to contain complete processing tiles that are stored in the storage tiles based on the storage tile size, processing tile size, and the mappings.

Abstract: A method of processing data is provided, which includes a plurality of processes 502 to 505 equipped with dependencies 603 and a continuation manager 814, said processes 502 to 505 including a plurality of sub-processes 507. The continuation manager 814 recurrently selects (1301) each of said plurality of processes for execution in the order of their dependencies. The continuation manager further recurrently selects (1302) each of said plurality of sub-processes 507 within each of said plurality of processes 502 to 505 for execution. The output data generated at each of said selected sub-processes 507 is generated in the order of said dependencies 603.

Abstract: A method of caching data is provided, which includes a plurality of processes 1602 to 1605, a cache manager 813 and a data type register 805 including at least one data type 1901 and a corresponding data type bit 1903. Said data type bit 1903 is set (1904) within the register 805 on being accessed by each of said processes and subsequently reset (1905) within the register. The cache manager 813 restores (1501) each of said set data type bit and identifies its corresponding data type 1901. The cache manager writes the output data 1609, 1610, 1611 of each of said processes 1603, 1604, 1605 within a memory cache 2001 and said cache manager resets (1505) said memory cache 2001 when the data type bit set by the last of said processes 1602 is reset.

Abstract: A method of processing data is provided, which includes a plurality of processes (1602 to 1605). A first process 1603 supplies data 1609 of a predetermined type 1901 to a second process 1602, wherein said second process 1602 stores type data 606 identifying the type of data it expects to receive from said first process 1603. Said second process is configured to modify its output 1606 with processing default data 607 if the data 1609 which is received from said first process 1603 differs from said stored data type 1901.

Abstract: A method of caching data is provided, which includes a plurality of processes 1602 to 1605, a cache manager 813 and a data type register 805 including at least one data type 1901 and a corresponding data type bit 1903. Said data type bit 1903 is set (1904) within the register 805 on being accessed by each of said processes and subsequently reset (1905) within the register. The cache manager 813 restores (1501) each of said set data type bit and identifies its corresponding data type 1901. The cache manager writes the output data 1609, 1610, 1611 of each of said processes 1603, 1604, 1605 within a memory cache 2001 and said cache manager resets (1505) said memory cache 2001 when the data type bit set by the last of said processes 1602 is reset.

Abstract: A method of processing data is provided, which includes a plurality of processes 502 to 505 equipped with dependencies 603 and a continuation manager 814, said processes 502 to 505 including a plurality of sub-processes 507. The continuation manager 814 recurrently selects (1301) each of said plurality of processes for execution in the order of their dependencies. The continuation manager further recurrently selects (1302) each of said plurality of sub-processes 507 within each of said plurality of processes 502 to 505 for execution. The output data generated at each of said selected sub-processes 507 is generated in the order of said dependencies 603.