На 18.05.11 16:29, andu novac написа:
На 18.05.11 14:10, andu novac написа:
Hi, I tried to use Jitsi (OS X) with talkr.im (ejabberd) service and I
noticed there is no option for voice/video.
What application was your contact using in this case? We need clients to
advertise support for standard Jingle in order to display the call options.
That explains it then.
I was able to do
voice/video with iChat though and the performance was better then
Skype of which I want to get rid of.
Note that iChat for example does not support standard Jingle.
Shall I understand that the issue was with the other party's iChat and
that Jitsi actually supports XMPP/Jingle? That would be easy to fix
In a way, although it would be more accurate to say that the "problem"
doesn't lie with either party. Both applications simply use incompatible
mechanisms to communicate. Jitsi supports standard XMPP Jingle. I am not
sure exactly what protocol iChat uses but would assume some sort of a
proprietary Jingle variant.
Also I tried sip with iptel.org and there is a horrible echo which
made the call unusable, my conversation partner was using Jitsi on
Windows. Thanks for any help.
Depending on the exact configuration echo cancellation may take a while
to kick in (e.g. sometimes up to 10 seconds). The algorithm is adaptive
so the longer you speak, the better the results should get.
Note that in order to improve results you'd generally want to use the
same audio device for both playback and capture.
Well, I have an Intel iMac so I use the built-in devices.
If you were the one hearing the echo then it was your interlocutor's PC
that was failing to cancel it. You'd have to check with them. Not that
cancellation may also fail for other reasons like for example high CPU
load or varying capture latency.
The only reason I decided to use xmpp as opposed to sip was the echo.
Well ... echo cancellation has hardly anything to do with whether you
are using SIP or XMPP. You would use the same audio devices to capture
and play audio, the same echo canceller to remove it, and the same rtp
stack to and codecs to transport it.
On Wed, May 18, 2011 at 5:13 PM, Emil Ivov <firstname.lastname@example.org> wrote: