Thursday, April 2, 2009

ATSC is considering Scalable Video Coding (SVC) as part of their Video System Characteristics

The ATSC is currently considering the adoption of Scalable Video Coding (SVC) for the ATSC Mobile DTV Candidate Standard. One can find the exact definition in Part 7 – Video System Characteristics which defines the following:
  • Video processing before AVC compression followed by the source coding specification for AVC
  • Video processing before SVC compression (informative) followed by the source coding specification for SVC
From the specification I'd like to quote something from p. 15:
"The SVC bit stream comprises two layers: the term “SVC base layer” is used to designate the layer that has dependency_id and quality_id values both equal to 0, and the term “SVC enhancement layer” to designate the layer that has dependency_id value higher than 0."
That is, ATSC defines one base layer and one enhancement layer. The base layer is per definition compliant to AVC as defined in AVC part of this candidate specification. The enhancement layer has some constraints defined which can be found in Table 9.3 on p. 17. Finally, it's interesting to see that an RTP packetization is defined for the SVC transport in two RTP sessions and a single RTP session.

No comments: