Paul,
Table B.2/H.225.0v3 lists the sampling rate for both laws of G.711 at 8kHz, and H.323 references H.225.0.
I haven't read the SIP/H.323 ID, but remember that support for Fast Connect is optional in H.323.
A GK is not a callable entity, so it cannot terminate an H.323 call, whereas a GW can. One could co-locate a GW with a GK so that the system would both control the call and transcode the protocols and/or media, but they are still two functionally distinct entities.
Paul Long Smith Micro Software, Inc.
-----Original Message----- From: Edgar Martinez [1] [mailto:martinze@CIG.MOT.COM] Sent: Friday, January 14, 2000 7:41 AM To: ITU-SG16@MAILBAG.INTEL.COM Subject: [Fwd: Re: FW: I-D ACTION:draft-singh-sip-h323-00.txt,.ps]
FYI
-------- Original Message -------- Subject: Re: FW: I-D ACTION:draft-singh-sip-h323-00.txt,.ps Date: Fri, 14 Jan 2000 11:52:20 +0100 From: Paul Sijben sijben@LUCENT.COM Reply-To: Paul Sijben sijben@LUCENT.COM Organization: Lucent technologies, The Netherlands To: TIPHON@LIST.ETSI.FR References: E39024226822D311BC880008C77318A1816106@oteis01nok
I have been reading this draft and have been both puzzled and pleasantly surprised.
I have been puzzled over some of the statements that are made on H.323, I do not recall that there is a limit on the sample frequency on G711 in H.323, the draft claims its fixed to 8000. (Could someone resolve this?)
The authors also seem to forget that there is H.323v4 now as well as the FastStart option... so the resulting call flow uses very many message exchanges, this is not necessary is it???
I also have been puzzled by the authors proposing to make a SIP-H.323 gateway whereby the H.323 side looks like an H.323 terminal, in some cases an H.323 gatekeepr seems more appropriate.
I have been pleasantly surprised in that the authos basically use the same method of interworking as we have been looking at in the DTS02003 information flows. So either we are all wrong or all right :-)
Paul