I am experiencing this issue simliar to other posters and to the TAC Case Collection mentioned in this forum previously.
TAC is requesting I get the disconnect tone from my provider. I am having a tough time trying to find this information from Korea Telecom. Does anyone know it? Will creating a custom cptone have better results than just setting cptone KR?
It is a FXO port that is having the issue:
connection plar opx xxxxxx
I also tried the supervisory stuff and that didn't seem to help.
Setting the cptone to KR will use the default settings for Korea, but we find that in many countries they do not use consistent tones across the entire nation, so even through cptone KR might work in one location, it may fail in another site. The best thing to do here is to get an ethereal/wireshark log of a call that is disconnected. If you use G711 as the codec, we can easily save the audio payload out of ethereal as an AU file and can analyse the frequencies and the cadence (duration) of the tone. You should be able to plug a PC into the phone's PC port and capture the voice traffic.
We do this all the time to create custom tones for supervisory disconnect.
If you can get the trace file, attach it to this thread of forward it to the TAC engineer, either way once we have it it is simple to create the custom tone class.
Great tip. I wish I could get my monitor session to work right. I tried what you said but I wasn't seeing any voice traffic coming out of the PC port. Is there something special I need to configure?
I also had a similar problem- after you hand up, the phone would ring again. It ended up being that the circuit was actually loop start and we had it configured for ground start (told be telco it was ground start).
this tone appears to be 480Hz+620Hz, 0.3 sec on, 0.3 sec off, same as Korea congestion as indicated on this useful site:
freq-pair 1 480 620
should be enough. Also please check which IOS are you using because with plar opx, the FXO does not answer the line before the ip phone answers, consequently tone detection does not apply, rather after the FXO detects ringing has stopped, it must also stop ringing the ip phone.
Hope this helps, please rate post if it does!
Sorry for the delay in getting back to you about this question.
Try the following config on your router (change the voice ports to suit) -
voice class custom-cptone Korea_customtone
frequency 480 620
cadence 300 200
We set up a custom disconnect class called 'Korea_customtone'. This will look for two tones of 480 and 620 hz, with a repeating on time of 300msec and an off time of 200msec
supervisory disconnect dualtone mid-call
supervisory custom-cptone Korea_customtone
timeeouts call-disconnect 3
timeouts wait-release 3
In case anyone wants to know how to do this themselves ...
Search on the internet for an audio application called 'cooledit96'(which is now Adobe Audition)
Load the audio file into the application.
Play back the audio to identify the disconnect tones
Use the cursors to highlight the start of the tone period, and then measure the on time duration. Do the same for the off time. This gives the cadence of the tones.
Move the cursor over the tone , then select Analyse-->Frequency Analysis
This will open a new window that shows a frequency domain view, and the two component tones are shown a individual peaks. Move the cursor over each tone peak and carefully read the displayed frequency (round it to the closest 5hz)
Thanks to everyone who spent time trying to help me on this. I was able to get the problem resolved with TAC. The following configuration resolved the issue:
supervisory disconnect dualtone mid-call
timeouts call-disconnect 5
timeouts wait-release 5
connection plar opx 70400
It sounds like the custom cptone and the KR cptone are the same. The supervisory disconnect mid-call dual-tone statement I believe is the one that fixed the problem.
I learned a lot though. Good information presented by all.
I just downloaded the software you recommended. Great suggestion. I was looking for something like this. It definitely takes some interpretation but the two peaks are visible. The second (620) seems to vary a bit more between 615 - 625. Just thought that was interesting.
We should relay thanks to Fourier and all the derived mathematician work. I wish I was able to understand a 10% of it.
Thanks for the nice rating and good luck!
this file cannot be analyzed because of low volume. In fact, just listening to it, you cannot hear anything. Can you try another recording ?
Hello Fender and welcome on this forum.
I often refer people to your database
and as many others here, find it one the best resource on the internet on telephony.
Keep the good work and again be welcome here for any issue you may want to discuss.