search

laurilo / draft-ilola-avtcore-rtp-v3c

Other
0 stars 2 forks source link

Reconsider the need for the three transmission modes (SRST, MRST and MRMT) #20

Closed laurilo closed 6 months ago

laurilo commented 7 months ago

Are the complexity of three transmission modes (SRST, MRST, and MRMT) actually needed or useful.

In practice, as I understand it, for the 2-D video codecs MRST and MRMT have not been used. This is the reason why the MRST and MRMT modes were removed for the VVC/H.266 and EVC payload formats (RFC 9328 and draft-ietf-avtcore-rtp-evc). Unless you have specific use cases that would require these other modes, I’d strongly suggest you simplify the RTP payload format by removing them, and modeling the V3C payload on VVC or EVC’s payload specifically. I believe this would also let you remove the grouping framework.

laurilo commented 7 months ago

The grouping of media lines is somewhat orthogonal to the proposal of removing different tranmisison modes.

I agree on the spec simplification by dropping SRST, MRMT and MRMT.

laurilo commented 6 months ago

Addressed in commit https://github.com/laurilo/draft-ilola-avtcore-rtp-v3c/commit/14005a0785dc8b0acf461810427334cb6bc965e4