[AVT] Uncertain about timestamp usage in RFC 2833 named events

["Steve Lubbs" <s.lubbs@ieee.org>]

Hi Folks,
I am in a bit of a quandry related to how the timestamp for RFC 2833 named
events should be implemented. My interpretation of the RFC is that the
timestamp marks the start of the event and that this same timestamp value is
used in all subsequent packets associated with the event. Further, the
duration value is cumulative and increases with each subsequent packet
associated with the event.
In order to confirm my interpretation I investigated the behavior of some
existing implementations. Some, such as MS Messenger and OpenPhone, appear
to agree with my interpretation. Others, notably a product from Cisco,
don't. The Cisco implementation appears to increment the timestamp for each
subsequent packet and keeps the duration constant. This is the behavior I
would have expected for RFC 2833 tones.

My question is simple. Which interpretation is correct?

Regards,
Steve Lubbs


_______________________________________________
Audio/Video Transport Working Group
avt@ietf.org
https://www1.ietf.org/mailman/listinfo/avt