p.s.

You didn't respond to the part about milliwatt.... doing that gives you a visual feedback on the recording as well.



On Tue, Nov 12, 2013 at 9:03 PM, Anthony Minessale <anthony.minessale@gmail.com> wrote:
In fact,

Can you give us all the info to reproduce your problem so we can see if we can create the same situation in our lab.
file a JIRA http://jira.freeswitch.org and attach the stuff that brian has asked for and preferably the instructions to reproduce the calls (We can do originate from cli so we don't need the event socket stuff) and the PSEQ app you are using.

If we can see what you mean on another sever maybe we can prove its environmental or a persistent issue.


On Tue, Nov 12, 2013 at 8:37 PM, Scott Ross <slross16@gmail.com> wrote:

I wasn't advised to use ubuntu, just including it as another data point (its my dev machine).

I've tested against two physical servers. 
       RedHat 6, Xeon 6 core 2.6ghz, 32g ram
       The dev machine running Ubuntu 13.04, decent Core i7, 12g ram 
I've tested on several XenServer instances running RedHat 6.
All devises/hosts are on the same internal network.

Seeing the same results regardless. 

The device on the other end of the calls is an external conference bridge (physical host). There could be some device specific issues - but I see the same results when testing against another freeswitch instance on the same network set up with mod_conference.
Using g711 and recording to .wav.
The call sounds unsatisfactory in real time and in the recordings
Call quality is near perfect when the two calls aren't bridged into a freeswtich conference or our external conference bridge. But even setting up a basic freeswitch conference to run the tests against resulted in the lower quality. 

So "failure" case appears to be when two separately originated calls being bridged into any single conferenced environment. 

Just to be clear - the quality is ok, but not great. For reference, I'm expecting a PESQ score of 3.8 or greater - which is met with asterisk (seeing 3.8 - 4.3), but Freeswitch in its current setup consistently results in 0.5 - 1 point lower at around 2.9 - 3.7.

I'm hoping there is some environment/configuration tweaking I can do to fix this. I have tried running FS with the -hp flag. Do I need to make configuration changes on my OS/network to account for that? 

Does this feel like an OS level issue? Do I need to try Debian 6?

Thank you for the reply. Again, let me know what other info I can provide.  


_________________________________________________________________________
Professional FreeSWITCH Consulting Services:
consulting@freeswitch.org
http://www.freeswitchsolutions.com

FreeSWITCH-powered IP PBX: The CudaTel Communication Server
http://www.cudatel.com

Official FreeSWITCH Sites
http://www.freeswitch.org
http://wiki.freeswitch.org
http://www.cluecon.com

FreeSWITCH-users mailing list
FreeSWITCH-users@lists.freeswitch.org
http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
http://www.freeswitch.org






--
Anthony Minessale II

FreeSWITCH http://www.freeswitch.org/
ClueCon http://www.cluecon.com/
Twitter: http://twitter.com/FreeSWITCH_wire

AIM: anthm
MSN:anthony_minessale@hotmail.com
GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com
IRC: irc.freenode.net #freeswitch

FreeSWITCH Developer Conference
sip:888@conference.freeswitch.org
googletalk:conf+888@conference.freeswitch.org
pstn:+19193869900