Re: [AVTCORE] Leap seconds

[Magnus Westerlund <[e-mail]>]

On 2011-09-14 21:46, Kevin Gross wrote:
> I'm getting the clear impression from this discussion and from the RFC
> that the NTP timestamp is not sacred. I'm free to us that for any
> timebase I wish and I may well do so.
> 
> With all these different timebases in play, how do receivers know which
> senders is running off which timebase?
> 
> For example, if a receiver receives 48 kHz audio from two different
> senders, how does the receiver know that those two sources can be
> directly mixed without requiring sample-rate conversion to first
> synchronize them? Is this an SDP thing?

It is beyond the assumption of RTP makes about clock bases. All SSRCs
that have a common CNAME can be assumed to come from a common
synchronziation context, and thus the same underlying wall clock.
Sources that are from different contexts needs to be dealt with as non
synchronized sources, with possible drifting clocks in relation to each
other.

Cheers

Magnus Westerlund

----------------------------------------------------------------------
Multimedia Technologies, Ericsson Research EAB/TVM
----------------------------------------------------------------------
Ericsson AB                | Phone  [phone]
Färögatan 6                | Mobile [phone]
SE-164 80 Stockholm, Sweden| mailto: [e-mail]
----------------------------------------------------------------------