Summary:ASTERISK-00418: [patch] h.323 with G.729 for openh323 12.0
Reporter:adam (adam)Labels:
Date Opened:2003-10-22 20:22:31Date Closed:2008-01-15 14:37:56.000-0600
Versions:Frequency of
Environment:Attachments:( 0) ast_h323.cpp.patch
( 1) ast_h323.cpp.patch
( 2) ast_h323.h.patch
( 3) ast_h323.h.patch
Description:This patch addresses two issues I've come across. Since OpenH323 1.12 was released, the G729 capability isn't defined (by default). I've simply added the capability class to ast_h323, as is already done with G.723.1.

Second issue I've had is with some H.323 gateways not understanding G.729A capability and only G.729. Hence, blocking my bridge to sip G.729 capable devices. To solve this, I've added G.729 capability when G.729A is on.
H323_G729Capability *g729Cap;
endPoint->SetCapability(0, 0, g729Cap = new H323_G729Capability);
I'm unsure how this will affect VoiceAge G.729 Codec. It's worked fine for myself. Can someone shed some light on G.729 vs G.729a in asterisk and why allowing G.729 would affect asterisk negatively? Sip has a mime type of G729 (not G729a), which would indicate allowing it should be fine
Comments:By: jerjer (jerjer) 2003-10-30 10:19:36.000-0600

How will this work if someone is still running the currently recomended versions of Open H.323?

By: John Todd (jtodd) 2003-10-30 12:47:34.000-0600

Just as importantly, does this prevent usage of the G729 codec from Voiceage?  Or is there no ill effect if you are transcoding?

By: adam (adam) 2003-10-30 16:30:16.000-0600

Very good point jerjer, a prefix of Ast_ should stop any class name clash

Jtodd, it's just an added capability that's advertised in H.245 negotiations. Like adding another mime type in sip's SDP.

I found i needed this with a H.323 call manager which only asked for G.729, yet the Voiceage G.729 codec handled it correctly.

By: adam (adam) 2003-11-11 16:49:33.000-0600

I've modified the patches for openh323 1.11 compatibility.

By: Isamar Maia (isamar) 2003-11-12 04:43:14.000-0600

After 6 months completely frozen with Asterisk+X100P+G729, finally a light
in the tunnel's end.
it solve my problem completely.
I had a G729 handshake problem with the following devices:

Now, with this patch, the problem is solved.


By: x martinp (martinp) 2003-11-12 09:54:39.000-0600

I heard that the diffrence between G729 and G729A and G729AB and G729B is the way the DSP routines are done in the codec itself. As far as the output as long as it's 8kbps it should be handled by the voiceage codec. So at least the patch that allows to use G729 is ok.

By: jerjer (jerjer) 2003-11-12 12:29:24.000-0600

For some reason i've been able to use G.729 pass-thru since day one on our 5300s. I'll apply this patch, but I no longer have the facilities to properly test it.

By: jerjer (jerjer) 2003-11-12 12:53:59.000-0600

All this patch really does is add G.729A.  The rest is just adding in the G.729 H323Capability.  

So in reality, one line of code has been stopping other people from using G.729????

By: adam (adam) 2003-11-12 16:25:13.000-0600

correct, just 1 line.

By: Isamar Maia (isamar) 2003-11-12 17:03:59.000-0600

Yeah... the dificulty, most of times, is the simplicity.

By: jerjer (jerjer) 2003-11-19 16:30:25.000-0600

applied with fixes to cvs

By: Digium Subversion (svnbot) 2008-01-15 14:37:56.000-0600

Repository: asterisk
Revision: 1767

U   trunk/channels/h323/Makefile
U   trunk/channels/h323/ast_h323.cpp
U   trunk/channels/h323/ast_h323.h

r1767 | jeremy | 2008-01-15 14:37:56 -0600 (Tue, 15 Jan 2008) | 2 lines

apply G.729 patches (with changes) from bug ASTERISK-418