[Asterisk-Dev] SIP RTP JitterBuffer in Asterisk

[Matt Hess]

This milliwatt is on the pstn.. it is not the asterisk based milliwatt() 
application. 303-458-0009 is a qwest milliwatt that I do a lot of 
testing to as it is very local to me.. I do have a milliwatt on another 
  asterisk system I can also call.. which yielded the same results as 
reported.

I have also tried dialing into the head system that has the jitter 
buffer on it.. called a milliwatt app on the system and the same choppy 
horrid audio was heard when the jitter buffer was forced.. however, with 
the forcejb=no setting I notice that most of my calls are not enabled 
with a jitterbuffer to begin with..
-- ***[RTP JB LOG]*** Detected bridged channel that can accept jitter. 
Disabling the jitterbuffer.

whether it is on a incoming call to a milliwatt app or a call 
transcoding through the patched head system does not matter.. without 
the forcejb variable set to 'yes' the jitterbuffer never seems to 
enable.. maybe it's just me.


Zoa wrote:
> 
> Can you try it with a playback of some file ? the milliwatt only gives
> audio packets when it receives audio, which might give strange results,
> we will reproduce your tests in the lab tomorrow.
> 
> Zoa
> 
> Matt Hess wrote:
> 
>> # gcc -v
>> Reading specs from /usr/lib/gcc-lib/i386-unknown-openbsd3.6/2.95.3/specs
>> gcc version 2.95.3 20010125 (prerelease, propolice)
>>
>> my results:
>> An unpatched head system calling an external milliwatt number is nice
>> and smooth. (good benchmark call)
>>
>> A head system with the latest patch on the same call is horrible. I've
>> tried using both the default variables/settings and modifying them to
>> no avail. Turning the jitter buffer off in sip.conf yields the same
>> smooth call.
>>
>> in tinkering further with the settings in sip.conf:
>>
>> usejb=yes
>> jbsize=300
>> forcejb=no
>> jblog=yes
>>
>> yields a good call but I note that I get the following log message:
>> -- ***[RTP JB LOG]*** Detected bridged channel that can accept jitter.
>> Disabling the jitterbuffer.
>>
>> with:
>> usejb=yes
>> jbsize=300
>> forcejb=yes
>> jblog=yes
>>
>> message log:
>> -- ***[RTP JB LOG]*** Jitterbuffer created and started. jb_alloc=1.
>> -- ***[RTP JB LOG]*** Jitterbuffer created and started. jb_alloc=2.
>>
>> and again the call is complete crap..
>>
>>
>> The call itself is a gsm codec call transcoded to ulaw.. gsm on the
>> external side and ulaw going to a bt-100 sip phone. The asterisk
>> server has 2 network interfaces.. one public and one private. The
>> bt-100 is on private ip space.
>>
>>
>> Slav Klenov wrote:
>>
>>> What version of gcc you are using? With 3.4.4 this compile pretty
>>> well. Actualy its a C++ style and some more strict compilers
>>> (including some gcc versions) generate an error.
>>>
>>>
>>> Matt Hess wrote:
>>>
>>>> I get a compile error.. did a fresh cvs get of asterisk and applied
>>>> patch cleanly.
>>>>
>>>> gcc -pipe  -Wall -Wstrict-prototypes -Wmissing-prototypes
>>>> -Wmissing-declarations -g  -Iinclude -I../include -D_REENTRANT
>>>> -D_GNU_SOURCE  -O6 -march=i386  -pthread
>>>> -fomit-frame-pointer    -c -o rtp.o rtp.c
>>>> rtp.c: In function `rtp_jb_destroy':
>>>> rtp.c:877: syntax error before `int'
>>>> rtp.c:882: `len' undeclared (first use in this function)
>>>> rtp.c:882: (Each undeclared identifier is reported only once
>>>> rtp.c:882: for each function it appears in.)
>>>> gmake: *** [rtp.o] Error 1
>>>>
>>>>
>>>> Slav Klenov wrote:
>>>>
>>>>> New SIP jitterbuffer patch against today cvs-head is available on
>>>>> mantis:
>>>>>
>>>>> http://bugs.digium.com/view.php?id=3854
>>>>>
>>>>>
>>>>> Slav
>>>>>
>>>>> Olle E. Johansson wrote:
>>>>>
>>>>>> Matt wrote:
>>>>>>
>>>>>>
>>>>>>> Hi,
>>>>>>> I heard talk that there was a SIP RTP JitterBuffer which was
>>>>>>> either in
>>>>>>> asterisk CVS, or was being made as a patch here on the dev
>>>>>>> list.   Can
>>>>>>> anyone confirm this or deny it?  And if it exists, what is the
>>>>>>> current
>>>>>>> status of it?
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> At this point, a lot of work is going on to fix this. We do not
>>>>>> know if
>>>>>> we can fix it in due time for 1.2. There are non-working patches
>>>>>> in the
>>>>>> bug tracker, but rumour (a less than one hour old rumour) tells me
>>>>>> that
>>>>>> something is working out there. Nothing is included in CVS at this
>>>>>> point.
>>>>>>
>>>>>> A bit longer answer than "no" this second time :-)
>>>>>>
>>>>>> /O
>>>>>> _______________________________________________
>>>>>> Asterisk-Dev mailing list
>>>>>> Asterisk-Dev at lists.digium.com
>>>>>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>>>> To UNSUBSCRIBE or update options visit:
>>>>>>   http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>>>>
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Asterisk-Dev mailing list
>>>>> Asterisk-Dev at lists.digium.com
>>>>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>>> To UNSUBSCRIBE or update options visit:
>>>>>   http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>>>
>>>>>
>>>> _______________________________________________
>>>> Asterisk-Dev mailing list
>>>> Asterisk-Dev at lists.digium.com
>>>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>> To UNSUBSCRIBE or update options visit:
>>>>   http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>>
>>>
>>> _______________________________________________
>>> Asterisk-Dev mailing list
>>> Asterisk-Dev at lists.digium.com
>>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>>> To UNSUBSCRIBE or update options visit:
>>>   http://lists.digium.com/mailman/listinfo/asterisk-dev
>>>
>>>
>> _______________________________________________
>> Asterisk-Dev mailing list
>> Asterisk-Dev at lists.digium.com
>> http://lists.digium.com/mailman/listinfo/asterisk-dev
>> To UNSUBSCRIBE or update options visit:
>>   http://lists.digium.com/mailman/listinfo/asterisk-dev
>>
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Asterisk-Dev mailing list
> Asterisk-Dev at lists.digium.com
> http://lists.digium.com/mailman/listinfo/asterisk-dev
> To UNSUBSCRIBE or update options visit:
>    http://lists.digium.com/mailman/listinfo/asterisk-dev
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mhess.vcf
Type: text/x-vcard
Size: 288 bytes
Desc: not available
Url : http://lists.digium.com/pipermail/asterisk-dev/attachments/20050825/3f79c2ad/mhess.vcf