The issue which occurs is that the audio from the SIP client to the IAX client will spend most of it's time sounded very robotic, and garbled. It is possible, although very difficult to understand someone who is on the SIP phone. I have asterisk 1.2.10 configured with realtime with both IAX and SIP clients. The SIP clients include a Grandstream gxp2000 hard phone, and Counterpath's X-Lite 3 (for windows) softphone. The IAX clients tested include idefisk (both windows & mac), JakenIAX, and LoudHush. GSM is the preferred codec of both IAX & SIP clients, and is indeed the codec being used in all tests. Audio from the IAX to the SIP client does not experience any issues. SIP to SIP (and presumably, although untested, IAX to IAX) communication does not experience any issues. We also have a T1 card through which many calls have been placed, both from the IAX and SIP phones, without any audio issues occurring, in either case. If it weren't for that there have been multiple clients tested to verify this robotic sound, I would cough it up to it being a incompatability between the particular clients, but this occurs on all SIP-IAX communication that has been tried. I'm running out of options as SIP-IAX intercommunication is kinda expected (and necessary for me), and out of good softphones for the mac, as most of the mac-compatible softphones are IAX2-based. Please let me know what additional information is needed to help me debug this problem. Thanks, -Joe
Joseph Love wrote:> The issue which occurs is that the audio from the SIP client to the IAX > client will spend most of it's time sounded very robotic, and garbled. > It is possible, although very difficult to understand someone who is on > the SIP phone. > > I have asterisk 1.2.10 configured with realtime with both IAX and SIP > clients. > The SIP clients include a Grandstream gxp2000 hard phone, and > Counterpath's X-Lite 3 (for windows) softphone. > The IAX clients tested include idefisk (both windows & mac), JakenIAX, > and LoudHush. > GSM is the preferred codec of both IAX & SIP clients, and is indeed the > codec being used in all tests. > > Audio from the IAX to the SIP client does not experience any issues. > SIP to SIP (and presumably, although untested, IAX to IAX) communication > does not experience any issues. > > We also have a T1 card through which many calls have been placed, both > from the IAX and SIP phones, without any audio issues occurring, in > either case. > > If it weren't for that there have been multiple clients tested to verify > this robotic sound, I would cough it up to it being a incompatability > between the particular clients, but this occurs on all SIP-IAX > communication that has been tried. > > I'm running out of options as SIP-IAX intercommunication is kinda > expected (and necessary for me), and out of good softphones for the mac, > as most of the mac-compatible softphones are IAX2-based. > > Please let me know what additional information is needed to help me > debug this problem.Can you try different codecs just to rule out any issues with that? E.g., if both devices use ulaw, do you still have the same problem? I've used both iaxcomm and x-lite to communicate with cisco, polycom, grandsteam, etc, without that type of problem. Is it possible to obtain an ethereal trace of both the iax and sip/rtp streams in the same trace? If so, a couple hundred packets should be more then enough to see what's going on.
On 26 Jul 2006, at 03:04, Joseph Love wrote:> The issue which occurs is that the audio from the SIP client to the > IAX client will spend most of it's time sounded very robotic, and > garbled. It is possible, although very difficult to understand > someone who is on the SIP phone. > > I have asterisk 1.2.10 configured with realtime with both IAX and > SIP clients. > The SIP clients include a Grandstream gxp2000 hard phone, and > Counterpath's X-Lite 3 (for windows) softphone. > The IAX clients tested include idefisk (both windows & mac), > JakenIAX, and LoudHush. > GSM is the preferred codec of both IAX & SIP clients, and is indeed > the codec being used in all tests. > > Audio from the IAX to the SIP client does not experience any > issues. SIP to SIP (and presumably, although untested, IAX to IAX) > communication does not experience any issues. > > We also have a T1 card through which many calls have been placed, > both from the IAX and SIP phones, without any audio issues > occurring, in either case. > > If it weren't for that there have been multiple clients tested to > verify this robotic sound, I would cough it up to it being a > incompatability between the particular clients, but this occurs on > all SIP-IAX communication that has been tried. > > I'm running out of options as SIP-IAX intercommunication is kinda > expected (and necessary for me), and out of good softphones for the > mac, as most of the mac-compatible softphones are IAX2-based. > > Please let me know what additional information is needed to help me > debug this problem.We have had reports like this, and it is looking like the iax jitterbuffer is the culprit. Try adding jitterbuffer=no to the general section of iax.conf and see if that helps. Tim Panton www.mexuar.com