A packet reassembler extracts meaningful chunks of data, such as video frames, from streams of RTP packets. A reassembler can be specific to a particular media type and compression format, such as a reassembler for Sorenson video, or it can be more generalized, such as a reassembler for any uncompressed audio, or even a reassembler for any QuickTime media.
Streaming media over RTP generally involves some packet loss. It is the responsibility of the reassembler to perform any loss recovery that goes beyond discarding data chunks that contain lost packets.
QuickTime includes a base reassembler which performs most of the routine work of packet reassembly. Your packet reassembler sets flags that control the behavior of the base reassembler. Your reassembler can also implement several functions to override the base reassembler, essentially taking over from it at almost any point.
The reassembler component must implement several functions, and must also provide a public component resource that describes the type of media, compression, and track characteristics that the reassembler supports. This resource also provides information on the reassembler's relative speed and the format's ability to handle loss.
Function definitions and additional discussion of all packet reassembler functions can be found in the Packet Reassembler Functions section. The QuickTime Streaming Reference section also defines the RTPRssmPacket and RTPRssmInitParams data structures.
| Previous | Chapter Contents | Chapter Top | Next |