Re: [webrtc-stats] scalabilityMode: Same as in getParameters() or not? (#731)

In my opinion, the scalability mode is that which the encoder is configured with. If the frame is dropped due to some rate limit then it does not affect the encoding structure, and if it is dropped in egress then it is also not affecting the encoding structure but rather a dropped frame in the network. 

So with the point-of-view of the encoder, if you reconfigure to drop layers, you are going from S spatial layers to S-1 spatial layers, then it will have a defined name (L{S}Tx => L{S-1}Tx). Same thing with temporal layers, T -> T-1 temporal layers has a defined name (LxT{T} -> LxT{T-1}). You can never drop the last layer, and there can't be more than 3 layers, so they are always named.

Technically a codec implementation could produce a structure which is not named, for exampling if scaling was neither 2 nor 1.5, or with some very custom spatial or temporal scaling. This is not a real problem in practice since there is no way to configure that to happen.

-- 
GitHub Notification of comment by eshrubs
Please view or discuss this issue at https://github.com/w3c/webrtc-stats/issues/731#issuecomment-1420908481 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Tuesday, 7 February 2023 14:54:57 UTC