Re: [mediacapture-main] Add a mechanism to inform a video track about its type of content

I think @stefhak has raised good points, i.e. UA should be able to get
 enough info from the capture device (or device driver) as long as the
 device is connected to the system.  Otherwise, it'd be a limitation 
in the driver design.  It won't make sense for all apps (incl. 
webpages) to figure out properties of the specific capture device.  
Assuming we have that, the encoder optimization could be treated as UA
 internal operation or implementation issue, there is no need for apps
 to pass in additional hint.

In addition, encoder optimization, especially for real-time encoding, 
can be a fairly complicated process, and can vary based on the 
specific scenarios, network bandwidth, and CPU load during runtime, 
etc.  I would expect each consumer interface should expose the right 
encoding control knobs based on the key scenarios targeted by the 
specific consumer interface. The degradationPreference in WebRTC is a 
good example to me, apps can adjust that based on the content they are
 sending.  I think this matches what @stefhak mentioned earlier in the
 discussion as well.

So, I agree with @alvestrand and @stefhak to position this as LATER.

-- 
GitHub Notification of comment by ShijunS
Please view or discuss this issue at 
https://github.com/w3c/mediacapture-main/issues/391#issuecomment-250518577
 using your GitHub account

Received on Thursday, 29 September 2016 16:26:58 UTC