Re: SDP attribute "a=imageattr:": (2) revised SDP O/A [3GPP CT4 "SIS-CT"]

Thanks Christian for you evaluation and further insights!

Concerning 3GPP Iq, I guess that both wildcarding levels need to be described:

"PT level" wildcard: expected not to be used at IMS call control level (due to single video codec); thus the IMS-ALG shall not use "PT wildcarding" as well at gateway control level

"attr level" wildcard: guess "$" and "-" both are applicable, dependent on the use case (IMS-AGW either agnostic to video processing or the IMS-AGW is requested for some kind of "video transcoding" (which might be "video transsizing" here)).

Nevenka, could you please indicate your specific points?

I think its not so simply just to check out H.248.1 wildcarding usage and command.reply rules. It was observed that wildcarding at the level of H.248-embedded SDP needs carefull consideration, that's H.248.39 was established to describe wildcarding at SDP level in more detail.

Further: I would appreciate if you could transform the 26.114 SIP/SDP example on an H.248 command request.
Just considering a single (H.248) Stream Endpoint/Termination might not be sufficient because the IMS-AGW must compare the two Stream Descriptors of each SEP/Termination (see above ... agnostic vs video specific support ...).

Regards,
Albrecht 


-----Original Message-----
From: Christian Groves [mailto:[email protected]] 
Sent: Donnerstag, 8. August 2013 10:55
To: Schwarz, Albrecht (Albrecht)
Cc: ext Nevenka Biondic; [email protected]; [email protected]
Subject: Re: [Megaco] SDP attribute "a=imageattr:": (2) revised SDP O/A [3GPP CT4 "SIS-CT"]

Hello Albrecht,

I wasn't part of the discussions so I'm not 100% what was being proposed.

I'm not sure using "*" for the PT would require the use of reserve value/group. My understanding is that it is simple shorthand if you have a list of payload. A MGC could provide an over specified list of payloads and utilise a=imageattr:*. The MG could response with a single payload and still have a=imageattr:* .

With respect to using "*" in the recv attr-list I'm not sure what this would indicate from a H.248 perspective. There seems to be a clash between the RFC6236 and the H.248 usage. RFC6236 indicates that "*" 
indicates that the user sees no reason to use the image attribute. 
Whereas the H.248 wildcard usage it would mean use "all" the values. I think this interaction is useful to specify. From a H.248 wildcarding perspective it would seem that "-" truly means the MGC doesn't care and "$" would mean provide a value.

There's also the issue of mapping the SDP to the local and remote descriptors. I take it that for H.248 you only specify a=imageattr:PT send attr-list or a=imageattr:PT recv attr-list depending on the descriptor.

Regards, Christian

On 7/08/2013 8:21 PM, Schwarz, Albrecht (Albrecht) wrote:
>
> The 2^nd discussion subject was related to revised SDP O/A elements:
>
> RFC 6236 explicitly indicates their (= RFC 5939 & RFC 6871) usage (see 
> clause 3.2.6/RFC 6236).
>
> ...
>
> a=mcap:1 video H264/90000
>
> a=acap:1 imageattr:%1% send [x=720,y=576,sar=[0.91,1.0,1.09,1.45]]
>
> ...
>
> Also 26.114 indicates (RFC 5939) usage in the example:
>
> ...
>
> a=tcap:1
>
> The general comment made in the CT4 meeting: the mapping of these SDP 
> elements from SIP to H.248 would need to take the work of H.248.80 
> into account (due to the different resource reservation principles at 
> SIP and H.248 interfaces).
>
> -Albrecht
>
> *From:*[email protected] [mailto:[email protected]] *On 
> Behalf Of *Schwarz, Albrecht (Albrecht)
> *Sent:* Mittwoch, 7. August 2013 12:07
> *To:* ext Nevenka Biondic; [email protected]; 
> [email protected]
> *Subject:* [Megaco] SDP attribute "a=imageattr:": (1) wildcarding 
> [3GPP CT4 "SIS-CT"]
>
> Hi Nevenka, all,
>
> you proposed to continue the technical discussion at next CT4 meeting; 
> like to suggest to start in parallel the clarification of raised H.248 
> aspects .
>
> The first topic is related to SDP wildcarding:
>
> RFC 6236 allows wildcard usage for multiple attribute parameters.
>
> I think we may just focus on the high level SDP format:
>
> a=imageattr:PT send attr-list recv attr-list
>
> The wildcard "*" (ALL) may be applied principally for multiple 
> attribute elements (in red), right?
>
> a=imageattr:PT send attr-list recv attr-list
>
> e.g., PT:
>
> A wild card (*) can be specified for the payload type number to 
> indicate that it applies
>
> to all payload types in the media description.
>
> You mentioned (in the CT4 discussion) wildcard usage at the level of
>
> 1) PT
>
> 2) recv attr-list
>
> as far as I remember, and the PT type wildcard is the difficult one, 
> because affecting possibly multiple media configurations, right?
>
> My initial comments and understanding would be:
>
> 1.A "PT level" wildcard would imply multiple (reserved) media 
> configurations, i.e., usage of ReserveGroup.
>
> 2.A "recv attr-list" level wildcard would enforce the MG to reply the 
> choosen/supported attribute values
>
> 3.Like to suggest you specify the SDP wildcarding of this attribute in 
> H.248.39, because
>
> a.the wildcard "*" is applicable for SIP signalling, whereas the SDP 
> mapping to H.248 allows principally also to use additional wildcards 
> (here e.g., "$" and "-"); and
>
> b.possible interactions with ReserveValue/ReserveGroup.
>
> 4.Example from 26.114 ("I've copied one below"):
> 26.116 is not using the "PT level" wildcard at all, because there's 
> just a single video codec (H.264) considered by 26.114, - right?
>
> Regards,
>
> Albrecht
>
> _______
>
> SIP/SDP example (26.114):
>
> *Table A.4.10a: Example SDP offer for H.264/AVC with image size
> negotiation*
>
> *SDP offer*
>
> a=tcap:1 RTP/AVPF
>
> m=video 49154 RTP/AVP 99
>
> a=pcfg:1 t=1
>
> b=AS:315
>
> b=RS:0
>
> b=RR:2500
>
> a=rtpmap:99 H264/90000
>
> a=fmtp:99 packetization-mode=0; profile-level-id=42e00c; \
>
> sprop-parameter-sets=J0LgDJWgUH6Af1A=,KM46gA==
>
> a=imageattr:99 send [x=176,y=144] [x=224,y=176] [x=272,y=224] 
> [x=320,y=240] recv [x=176,y=144] [x=224,y=176] [x=272,y=224,q=0.6] 
> [x=320,y=240]
>
> a=rtcp-fb:* trr-int 5000
>
> a=rtcp-fb:* nack
>
> a=rtcp-fb:* nack pli
>
> a=rtcp-fb:* ccm fir
>
> a=rtcp-fb:* ccm tmmbr
>
> a=extmap:4 urn:3gpp:video-orientation
>
>
>
> _______________________________________________
> Megaco mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/megaco