3.1.5.3 Understanding the Sequence Header

  • Article

The sequence header carries enough information for the RTVideo receiver to decode a group of video frames. Because of this, this protocol requires that the sequence header MUST be present for the first video data packet of an RTVideo I-frame.

In addition to the semantics described in the SMPTE 421M standard [SMPTE-VC-1], the semantics of the MAX_CODED_WIDTH and MAX_CODED_HEIGHT fields in the sequence header is extended to represent the original picture aspect ratio of the video frames. Once a video frame has been received and decoded using the CODED_WIDTH and CODED_HEIGHT information in the entry point header, the display process in the receiving endpoint SHOULD<6> use the MAX_CODED_WIDTH and MAX_CODED_HEIGHT field values to reconstruct a video frame with horizontal and vertical dimensions agreeing with the original picture aspect ratio.

The sequence header present in the Codec Header Bytes field also contains the DISP_HORIZ_SIZE and DISP_VERT_SIZE fields. The receiving endpoint SHOULD<7> ignore these fields and continue to use the MAX_CODED_WIDTH and MAX_CODED_HEIGHT field values to determine the horizontal and vertical dimensions of the displayed video frame, as described in the preceding paragraph.