W3C home > Mailing lists > Public > public-webrtc@w3.org > May 2014

RE: RTPSender/Receiver - The case for Constraints

From: Bernard Aboba <Bernard.Aboba@microsoft.com>
Date: Sat, 17 May 2014 00:23:33 +0000
To: "Suhas Nandakumar (snandaku)" <snandaku@cisco.com>, Stefan Håkansson LK <stefan.lk.hakansson@ericsson.com>, "public-webrtc@w3.org" <public-webrtc@w3.org>
Message-ID: <797269dad1214dcfba77e633a48d943a@SN2PR03MB031.namprd03.prod.outlook.com>
Suhas said: 

"3. I am aware we might not be interested in solving Simulcast/Layered coding like Multi-streams
per track at this point in time, but enabling RTPStream level constraints is quite useful to achieve
such a functionality."

[BA] I am skeptical about the value of complex constraints for layered coding. 

Because there is often a fixed relationship between the base layer and other layers, attempting to constrain multiple layers simultaneously in SVC will most likely lead to an overconstrained condition.  If we assume constraints have already been applied in getUserMedia, then something as simple as specifying the number of desired layers of each kind (temporal, spatial, quality) can be sufficient to derive the properties of each layer from the characteristics of the mediaStreamTrack to be encoded. 
Received on Saturday, 17 May 2014 00:24:02 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 15:19:40 UTC