>> What 'application-specific SDP attributes' do you have in mind?
> We are targetting professional live audio for studio. For precise sync and
> low delay we have to carry information giving the relation between RTP
> timestamp and PTP timestamp (NTP timestamp carried in RTCP is not precise
> enough).
The following work - currently underway in the IETF - might be relevant here.
(If so, you should adopt it and/or contribute to this effort.)
http://tools.ietf.org/html/draft-ietf-avtcore-clksrc-01
(If/when this work becomes an IETF standard, I will update the LIVE555 code to
support this new "ts-refclk" SDP attribute, if requested.)
> You can also think to other extensions, like "a=orient:portrait" and the ones
> given in Section 6 of RFC456, or "a=cliprect:..." (Quicktime) or "a=crypto"
> (for RTP/SAVP RFC4568), or de-facto standard extensions, like
> "a=x-dimensions" that you've chosen to implement...
Yes, however - although I have (reluctantly) implemented some non-standard
attributes that are already in widespread use, I am disinclined to support
*new* non-standard SDP attributes; only those SDP attributes that get defined
as parts of new IETF standards. Sorry.
> Ideally, it would be cumfortable to be able to extend "Media Level" and
> "Session Level" attributes through virtual methods.
That would be for your RTSP client. What about your RTSP server? It would
also need to generate these new SDP attribute(s). (Is your RTSP server not
also implemented using our code? If not, why not? :-)
But anyway, as noted above, modifying our code solely to make it easier for
people to use non-standard SDP attributes is not a priority for me.
Ross Finlayson
Live Networks, Inc.
http://www.live555.com/
_______________________________________________
live-devel mailing list
live-devel@lists.live555.com
http://lists.live555.com/mailman/listinfo/live-devel
