t38FaxMaxDatagram (H.245)
-
- Posts: 4
- Joined: Tue Aug 25, 2009 7:34 am
- Location: Ekaterinburg, Russia
t38FaxMaxDatagram (H.245)
Paul, please help me to understand the correct meaning of t38FaxMaxDatagram. When using UDP TL does this parameter define the maximum UDP data size which receiver CAN process? If receiver sets t38FaxMaxDatagram = 72 should sender transmit UDP packets with no more than 72 bytes of data?
- paulej
- Posts: 606
- Joined: Sun Aug 23, 2009 7:32 pm
- Location: Research Triangle Park, NC, USA
- Contact:
Re: t38FaxMaxDatagram (H.245)
This parameter represents the largest size of the datagram that can be accepted by the receiving device. This does not count RTP headers, for example. There is debate as to whether this counts redundancy, but I think the general consensus is that it does not consider redundancy. So, if a device that advertises 72 octets also supports redundancy can actually receive larger UDP packets (containing the primary and redundant data), but the intent is that the primary fax data would not consume more than 72 octets.
This and many of the T.38-related parameters have been a source of confusion and, for that reason, there are currently three efforts underway to work to clarify how T.38 is supposed to work. It is expected that a new revision of T.38 will be made available next year with better definitions. And, you're definitely encouraged to contribute. Part of the work is going on in the SIP Forum (where I'm engaged). Some of the work is going on in the ITU (in which I am indirectly involved). I just recently learned that i3 Forum is also doing work in this area.
This and many of the T.38-related parameters have been a source of confusion and, for that reason, there are currently three efforts underway to work to clarify how T.38 is supposed to work. It is expected that a new revision of T.38 will be made available next year with better definitions. And, you're definitely encouraged to contribute. Part of the work is going on in the SIP Forum (where I'm engaged). Some of the work is going on in the ITU (in which I am indirectly involved). I just recently learned that i3 Forum is also doing work in this area.