[Freeswitch-users] DTMF delay when using FreeSWITCH

Thu Oct 4 22:54:11 MSD 2012

I might have gotten excited too quickly. I think it has improved significantly, but still not as fast as I need it to be.
On Oct 4, 2012, at 1:06 PM, Emrah <lists at kavun.ch> wrote:

> And you are fabulous!
> 
> It worked and it looks reliable, thanks a million!
> 
> I use it to send DTMFs via a SIP carrier, all PCMu.
> 
> On Oct 4, 2012, at 8:40 AM, grmt <garmt.noname at gmail.com> wrote:
> 
>> I’m not sure I fully understand what you are doing with DTMF, sending or receiving, inband or outband, RFC2833 or not, but I notice you are using “start_dtmf”.
>> If you want FS to recognize non RFC2833 inband DTMF, (typically when only PCMA or PCMU is supported by your old fashioned sip provider), than use mod_spandsp ‘s DTMF detector, it’s better than the standard dtfm detector. Load mod_spandsp and use <action application="spandsp_start_dtmf" /> instead of start_dtmf
>> 
>> See the wiki for more details on configuring spandsp’s DTMF detector
>> 
>> From: freeswitch-users-bounces at lists.freeswitch.org [mailto:freeswitch-users-bounces at lists.freeswitch.org] On Behalf OfAnthony Minessale
>> Sent: Wednesday, October 03, 2012 23:44
>> To: FreeSWITCH Users Help
>> Subject: Re: [Freeswitch-users] DTMF delay when using FreeSWITCH
>> 
>> Set the following variable either per leg in dialplan or globally in vars.xml
>> 
>> rtp_manual_rtp_bugs=IGNORE_DTMF_DURATION
>> 
>> This will pass the dtmf in 2833 on receipt of the first packet.
>> 
>> If that doesn't work, don't bother going further because we do not put emphasis on realtime dtmf.  Asterisk can focus on that and we'll focus on scalability.  The suggestions you have seen pretty much cover your options.
>> 
>> 
>> 
>> 
>> 
>> On Wed, Oct 3, 2012 at 2:39 PM, Emrah <lists at kavun.ch> wrote:
>> I am running out of ideas and would really appreciate some input on this.
>> 
>> Can I optimize this in any way?
>> 
>> Best,
>> Emrah
>> 
>> On Sep 30, 2012, at 12:05 AM, Emrah <lists at kavun.ch> wrote:
>> 
>>> When I try this, the delay disappears but FS detects multiple DTMFs where I only send one…
>>> <action application="export" data="dtmf_type=none" inline="true"/>
>>> <action application="start_dtmf" />
>>> 
>>> start_dtmf_generate adds the delay and so I'm stuck…
>>> 
>>> Any suggestion?
>>> On Sep 29, 2012, at 2:29 PM, Emrah <lists at kavun.ch> wrote:
>>> 
>>>> Hi all,
>>>> 
>>>> I am now trying to force inband DTMF on my PSTN peers alone. I tried the following with no luck:
>>>> <action application="start_dtmf" />
>>>> <action application="export" data="dtmf_type=inband" />
>>>> 
>>>> The delay is still there and I get the following output in my console for a single DTMF:
>>>> 
>>>> 2012-09-29 19:22:57.357913 [DEBUG] switch_rtp.c:3797 RTP RECV DTMF 2:1040
>>>> 2012-09-29 19:22:57.357913 [DEBUG] switch_ivr_bridge.c:393 Send signal sofia/external/1234567890 [BREAK]
>>>> 2012-09-29 19:22:57.377912 [DEBUG] switch_rtp.c:2736 Send start packet for [2] ts=748000 dur=160/160/1040 seq=27258 lw=748000
>>>> 2012-09-29 19:22:57.397908 [DEBUG] switch_rtp.c:2636 Send middle packet for [2] ts=748000 dur=320/320/1040 seq=27259 lw=748160
>>>> 2012-09-29 19:22:57.417913 [DEBUG] switch_rtp.c:2636 Send middle packet for [2] ts=748000 dur=480/480/1040 seq=27260 lw=748320
>>>> 2012-09-29 19:22:57.437960 [DEBUG] switch_rtp.c:2636 Send middle packet for [2] ts=748000 dur=640/640/1040 seq=27261 lw=748480
>>>> 2012-09-29 19:22:57.457912 [DEBUG] switch_rtp.c:2636 Send middle packet for [2] ts=748000 dur=800/800/1040 seq=27262 lw=748640
>>>> 2012-09-29 19:22:57.477904 [DEBUG] switch_rtp.c:2636 Send middle packet for [2] ts=748000 dur=960/960/1040 seq=27263 lw=748800
>>>> 2012-09-29 19:22:57.497915 [DEBUG] switch_rtp.c:2636 Send end packet for [2] ts=748000 dur=1120/1120/1040 seq=27264 lw=748800
>>>> 2012-09-29 19:22:57.497915 [DEBUG] switch_rtp.c:2636 Send end packet for [2] ts=748000 dur=1120/1120/1040 seq=27265 lw=748800
>>>> 2012-09-29 19:22:57.497915 [DEBUG] switch_rtp.c:2636 Send end packet for [2] ts=748000 dur=1120/1120/1040 seq=27266 lw=748800
>>>> 2012-09-29 19:22:57.497915 [DEBUG] switch_rtp.c:2589 Queue digit delay of 40ms
>>>> 
>>>> Any idea would be greatly appreciated.
>>>> 
>>>> All the best,
>>>> Emrah
>>>> On Sep 27, 2012, at 3:54 PM, Emrah <lists at kavun.ch> wrote:
>>>> 
>>>>> Hey Ken,
>>>>> 
>>>>> I tried pass_rfc2833 with no noticeable change in the delay. It seemed to have made it less accurate though, especially in fast speed sequences.
>>>>> 
>>>>> Can I debug this further and how?
>>>>> 
>>>>> Thanks!
>>>>> On Sep 27, 2012, at 3:00 PM, Ken Rice <krice at freeswitch.org> wrote:
>>>>> 
>>>>>> There can be a delay of DTMF in and DTMF out if you are sending long DTMFs
>>>>>> using 2833, FreeSWITCH gets the entire DMTF and duration then regenerates
>>>>>> it...
>>>>>> 
>>>>>> If you don't need to interpret the DTMF you can set a variable to make it
>>>>>> just pass the DTMF through untouched... But this has its own set of caveats
>>>>>> (ie: if whatever is sending you DTMF is broken it just pass broken 2833
>>>>>> DTMF)
>>>>>> 
>>>>>> See http://wiki.freeswitch.org/wiki/Variable_pass_rfc2833
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> On 9/27/12 1:49 PM, "Emrah" <lists at kavun.ch> wrote:
>>>>>> 
>>>>>>> MC, the issue does not happen with inband DTMF and there is no delay!
>>>>>>> 
>>>>>>> Any idea on how to debug this further? I can't use inband continuously.
>>>>>>> 
>>>>>>> Thanks!
>>>>>>> Emrah
>>>>>>> 
>>>>>>> On Sep 27, 2012, at 12:46 PM, Emrah <lists at kavun.ch> wrote:
>>>>>>> 
>>>>>>>> Never tried with inband DTMFs. Will check.
>>>>>>>> 
>>>>>>>> Thanks!
>>>>>>>> On Sep 27, 2012, at 12:34 PM, Michael Collins <msc at freeswitch.org> wrote:
>>>>>>>> 
>>>>>>>>> Does it happen whether you use RFC2833 or inband DTMFs? Just curious.
>>>>>>>>> -MC
>>>>>>>>> 
>>>>>>>>> On Wed, Sep 26, 2012 at 3:44 PM, Emrah <lists at kavun.ch> wrote:
>>>>>>>>> Yes I did.
>>>>>>>>> BTW, the example in the Wiki contradicts the inline documentation in
>>>>>>>>> switch.xml.
>>>>>>>>> <!--
>>>>>>>>>    The min-dtmf-duration specifies the minimum DTMF duration to use on
>>>>>>>>>    outgoing events. Events shorter than this will be increased in
>>>>>>>>> duration
>>>>>>>>>    to match min_dtmf_duration. You cannot configure a dtmf duration on
>>>>>>>>> a
>>>>>>>>>    profile that is less than this setting. You may increase this value,
>>>>>>>>>    but cannot set it lower than 400. This value cannot exceed
>>>>>>>>>    max-dtmf-duration. -->
>>>>>>>>> The Wiki shows an example with the value at 100.
>>>>>>>>> 
>>>>>>>>> I tried increasing and decreasing it to no avail, it does not seem to
>>>>>>>>> interfere with anything I can measure with my ear. :P
>>>>>>>>> On Sep 26, 2012, at 5:56 PM, Cesar Bermudez <cesar.bermudez at gmail.com>
>>>>>>>>> wrote:
>>>>>>>>> 
>>>>>>>>>> You tried this:
>>>>>>>>>> http://wiki.freeswitch.org/wiki/Sofia_Configuration_Files#DTMF
>>>>>>>>>> 
>>>>>>>>>> On Wed, Sep 26, 2012 at 3:19 PM, Emrah <lists at kavun.ch> wrote:
>>>>>>>>>> Hi guys,
>>>>>>>>>> 
>>>>>>>>>> I am comparing this with an Asterisk and FreeSWITCH installation, using the
>>>>>>>>>> same route, same codecs, same carrier, same phones and same serversŠ :P
>>>>>>>>>> I experience a delay when pressing DTMFs on the line that uses FreeSWITCH.
>>>>>>>>>> I am estimating the delay to be around 500 ms.
>>>>>>>>>> 
>>>>>>>>>> What are the settings I can fine tune to avoid this?
>>>>>>>>>> 
>>>>>>>>>> All the best,
>>>>>>>>>> Emrah
>> 
>> 
>> _________________________________________________________________________
>> Professional FreeSWITCH Consulting Services:
>> consulting at freeswitch.org
>> http://www.freeswitchsolutions.com
>> 
>> 
>> 
>> 
>> Official FreeSWITCH Sites
>> http://www.freeswitch.org
>> http://wiki.freeswitch.org
>> http://www.cluecon.com
>> 
>> FreeSWITCH-users mailing list
>> FreeSWITCH-users at lists.freeswitch.org
>> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
>> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
>> http://www.freeswitch.org
> 