July 2011 - FreeSWITCH-users - FreeSWITCH Mailing Lists

txfax and rxfax X headers
by Spencer Thomason 11 Jul '11

11 Jul '11

Hello all, Is there a way to add the fax result text as an X- header to the BYE after a fax attempt has been made. It seems that this would be useful in debugging a fax call. Just a thought :-) Spencer

2 2

play wav files, or speak prompts, into recording?
by Wes 11 Jul '11

11 Jul '11

Not sure how to phrase this correctly, which is why I'm having trouble searching for help on the internet. I'd like to "speak" some audio into the recording that I am making of the user's voice. For example, say I have the caller press some digits and then detect the digits that have been pressed, then, I can easily speak those digits back to them for confirmation, but what I'd really like to do is begin recording the call, and then speak those digits, or any other introductory message into the recording before they start speaking their message. Say they type in the digits 123, then I'd read into the recording: "The following is the dictation for user number 123, taken on July 3, 2011", and then the person would start speaking and the recording would continue. Perhaps this would have to be done by some sort of post processing, where I append the introductory message onto the user's recording using sox, but I'd still like to use the voice system in freeswitch because it is good at reading numbers, speaking time, etc. Thanks.

2 3

First few seconds of call are silent
by Ian Yates 11 Jul '11

11 Jul '11

Hi, We have a Microsoft Lync environment for our internal phone system and are using FreeSwitch to let Lync talk with our VOIP provider. Whilst the VOIP provider properly support Lync (one of the few to do so) our NAT firewall ruins the TCP-based SIP packets as it doesn't forward through the SIP OPTIONS packet and Lync eventually thinks the VOIP provider isn't there. Thus we're using FreeSwitch to support out NAT firewall (I've give it our static public IP address) and have disabled the firewall device's SIP ALG so that it doesn't ruin the SIP packets. Anyway, I have this working very well except that most of the time (not all the time - maybe 90%) we just have silence for the first 3-6 seconds after each call connects. This usually means that if we call someone we miss the "Hi, thanks for calling company XYZ. You're speaking with ABC" and confusion results. The call counter in Lync starts counting as soon as the call connects (we've tested with fixed landlines in one hand and our Lync-based phone in the other) but there's just silence. Lync to Lync calls don't have this problem. I've tried various options such as forcing codecs (as best as I could make it anyway) to G711a since that's all Lync supports anyway (our at least it's happy to use this). I've also tried <param name="rtp-autoflush-during-bridge" value="false"/> in both legs of the FreeSwitch connections (ie to our SIP trunk provider and to Lync) without any success. I was hoping that if it was some sort of lag then at least I would see it lag a lot in the call (and then have a definite problem to solve) but that wasn't the case as nothing changed. Should I do something with the proxy media settings? Reading the Wiki page at http://wiki.freeswitch.org/wiki/Proxy_Media seems to indicate that it really won't make much of a difference (our CPU isn't all that busy in the virtual machine running FreeSwitch and we only have a handful of phones anyway). Also, I think I'd like to take advantage of some of FreeSwitch's very powerful features in the future to take care of some of what seems trickier to do in Lync such as having our support hunt group fall back to mobile phones when we're out of the office (Lync does simultaneous ringing of a user's mobile phone for direct calls but not, as far as I can tell, for hunt groups) I previously had the 3CX software SIP phone talking to FreeSwitch when I was first testing out the system. I've since gutted the FreeSwitch XML config (the "internal" / "external" SIP profiles are gone) so I can't try the software phone at the moment but hope to get it going again soon. This way I can better see if it's Lync or FreeSwitch causing the problem. Before I fumble around getting that going I was hoping someone might just know the obvious thing I'm overlooking! :) Thanks very much for your help. Ian

2 1

482 Merged Request
by Spencer Thomason 11 Jul '11

11 Jul '11

Hello all, Sorry to dig up an old thread but we are having the same issue and I couldn't seem to find a solution. Our scenario is similar to this: http://lists.freeswitch.org/pipermail/freeswitch-users/2009-September/04670… Basically we are using FreeSWITCH as a basic B2BUA for topology hiding, CNAM lookups and number normalizing in our SBC setup. A call looks like this: Client UA -> Kamailio -> FS --back to--> Kamailo -> ISTPs I have three profiles configured: Outbound Inbound Media If we respond with a 486 or 404 to our providers gateways, they simply drop the call (which is totally fine). This doesn't give the person calling any real indication because the just hear the call drop. I have another profile setup to provide progress tones as early media and then issue the appropriate final response (404, 486, etc.) The problem is this. If a call passes thru a FS profile, back to the proxy and generates an error code that we redirect that call to the FS media profile to generate the tones, we get a 482 Merged Request back from FS because the Call-ID is the same and the call that just went thru, event though the transaction has ended. Is there any way to do this without using another server just to generate the tones? Thanks, Spencer

1 0

SIP OPTIONS
by Michael B 11 Jul '11

11 Jul '11

Hello, Our carrier by default sends us a SIP OPTIONS message at certain intervals to ensure we are alive. Is there anyway I can configure Freeswitch to respond to this with 200OK or similar? I have read that FS is able to do this itself via a SIP PING, but can't find anything that talks about the other way, i.e. FS being the one to receive the OPTIONS. Any advice would be greatly appreciated. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/SIP-OPTIONS-tp6570046p6570046… Sent from the freeswitch-users mailing list archive at Nabble.com.

2 2

domain_name variable
by Tomasz Kopacki 11 Jul '11

11 Jul '11

Hi, I've got a problem with device variable - user_context. In my scenario, context name equals domain name, so i want to use variable to set user_context. I've tried 'user_context=${domain_name}' but it doesnt work, every call goes to public context. How to do this properly ?

3 4

live messenger RevEng
by Madovsky 11 Jul '11

11 Jul '11

for the geeks who d like to create a mod_live_messenger http://msnwcrec.arrozcru.org/vc_1_1.pdf http://www.codeproject.com/KB/macros/wlmplugin.aspx http://msnwcrec.arrozcru.org/ http://imfreedom.org/wiki/MSN

1 0

No voice channels to mobile.
by Henk Oegema 10 Jul '11

10 Jul '11

I'm facing the following problem: External calls come in via an ITSP (Localphone) and then routed to internal extension 1000. That works OK. If there's no answer, then the call is routed outbound via ITSP (Powervoip) to my mobile +316xxxxxxxxx) In this case there is NO voice communication at all. (a direct call from 1000 to a mobile is OK) <extension name="1000"> <condition field="destination_number" expression="^1000$"> .......................................................................................... .......................................................................................... <action application="set" data="continue_on_fail=true"/> <action application="set" data="hangup_after_bridge=true"/> <action application="bridge" data="{originate_timeout=20}sofia/internal/1000%$${domain}"/> <----------audio OK <action application="bridge" data="{originate_timeout=20}sofia/gateway/powervoip/003163086xxxx"/> <------------NO audio both ways <action application="voicemail" data="default $${domain} 1000"/> </condition> </extension> How do I approach this problem? Rgds, Henk

2 2

Speech Recognition
by Roman Gelfand 10 Jul '11

10 Jul '11

I am looking to script a speech recognition ivr system. Has anyone used cmusphinx to accomplish? If not has anyone use a good open source speech recognition engine with freeswitch? Thanks in advance

4 5

segmentation fault with mod_spandsp
by Chris Cureau 10 Jul '11

10 Jul '11

Hi there! This morning, I did a fresh git checkout and attempted to build freeswitch with mod_freetdm and mod_flite defined in the modules.conf. I am now getting a segfault when compiling: making all mod_spandsp Making all in src /bin/bash: line 1: 22453 Segmentation fault ./make_modem_filter -m V.17 -r > v17_v32bis_rx_floating_rrc.h make[7]: *** [v17_v32bis_rx_floating_rrc.h] Error 139 make[6]: *** [all] Error 2 make[5]: *** [all-recursive] Error 1 make[4]: *** [/usr/local/src/freeswitch/libs/spandsp/src/libspandsp.la] Error 2 make[3]: *** [mod_spandsp-all] Error 1 make[2]: *** [all-recursive] Error 1 make[1]: *** [all-recursive] Error 1 make: *** [all] Error 2 running make_modem_filter manually gets the same result. Twiddling with the commandline options doesn't help either. OS: Debian 6.0.1a amd64, also a fresh install from last night. Any idea what is happening? Thanks in advance!

3 8