I think the underlying requirement is to be able to decode as much as possible 
of a picture in the event that part of the data for that picture is lost. MTU 
targetting is one way to do that, but I'd be reluctant to mandate that specific 
mechanism, as it's a means to that end. 

It might also be technically hard to do for a modern codec, because of 
interactions of packetisation with tiles for parallelism and large blocks for 
coding efficiency, for example. The natural granularity of a bitstream with 
tiles is the tile, and that might be quite coarse for exact MTU targetting, but 
fine enough for reasonable error concealment. I'd also want to allow the codec 
to route around any IP issues that could turn up in trying to solve any of 
these problems.  

My suggestion would be something like: "the codec should support effective 
mechanisms for allowing decoding and reconstruction of significant parts of 
pictures in the event that parts of the picture data are lost in transmission". 

best regards

Thomas

-----Original Message-----
From: video-codec [mailto:[email protected]] On Behalf Of Alexandros 
Eleftheriadis
Sent: 27 August 2016 09:59
To: Timothy B. Terriberry <[email protected]>
Cc: [email protected]
Subject: Re: [video-codec] WGLC: draft-ietf-netvc-requirements

Don’t forget the MTU targeting capability. 

What you get is ensuring your ability to decode the media in individual packets 
when other media packets may be lost, and also minimizing packetization 
overhead by sending as much media data as possible in a single packet. 


> On Aug 27, 2016, at 1:43 AM, Timothy B. Terriberry <[email protected]> wrote:
> 
> Jonathan Lennox wrote:
>> So perhaps" “The codec should support mechanisms that allow individual 
>> packets to be sent in common network protocols without bitstream dependency 
>> between the packets.”  Thoughts?
> 
> That feels like it's moving in the right direction. Perhaps the better 
> question is, "What breaks if there is a bitstream dependency?" That seems 
> like it could tell you what the actual requirement is.
> 
> _______________________________________________
> video-codec mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/video-codec

_______________________________________________
video-codec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/video-codec
_______________________________________________
video-codec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/video-codec

Reply via email to