Playing

[Guest Roy Jensen]

All

 

I need VG (or any software) to broadcast a voice message to multiple persons on a community league using a regular D/F/V modem. My system:

WinXP sp1, lots of HD and RAM

Apache A56-HCF (PCI) D/F/V/S modem

latest download of VoiceGuide

 

From reading the forum, VG has much greater functionality than I need, but most people use dialogic telephony hardware. Sorry, I don't have 500$ for that.

 

I REALLY don't understand why everyone keeps saying that VG can't detect when the phone is answered when using a modem. There is evidently no hardware signal for 'answered'. Fine, USE SOFTWARE! Why can't VG (or any other software) _listen_ for silence before delivering the pre-recorded message? You can even make an educated guess based on the length of the time before silence is achieved. Can't be done? Bull. BroadcastByPhone does it and does it well. But they want 300$ for their software and add advertising to the beginning and end of the delivered message for the demo software.

 

I was hoping that I could use VG to call the number while I sat on another extension and pressed a key on the local telephone to get VG to deliver the message to the answering machines. This kinda works, but I have not figured out how to select the message that is played. I assume that I have to generate a script but I cannot figure out how to run the script after exiting the AcceptAutoCall.wav. Assistance would be appreciated.

 

Please excuse the frustration, I have been at this for a few weeks with over 10 different software packages...

 

If Katalina Technologies get the modem-silence detection to work and simplifies the setup, they would have a reasonable market for volunteer coaches, etc.

(A price drop wouldn't hurt, either.)

 

Thanks,

Roy Jensen

[SupportTeam]

Why can't VG (or any other software) _listen_ for silence before delivering the pre-recorded message?

The 'listen for silence' approach is actually not very reliable. To start off, what length of silence do you listen for? Many phone networks have a 4 seconds time between the successive 'ringbacks' that the modem hears after dialing the number - so the length of silence to listen for has to be more then 4 seconds. Now when a human answers the phone they say 'hello' - and then when they don’t hear anything for about 2-3 seconds they say 'hello' again - and then 3-4 seconds later they hang up... so you see how there is little margin to allow the 'listen for silence' approach to work.

 

There is a reason why Dialogic went through a lot of effort to figure out how to reliably detect presence of Voice on the line - 'listen for silence' just does not work well enough for most applications.

 

You'll get much better results with the 'press 1 to accept call' approach that Voiceguide uses (if you must use a Voice modem).

 

BTW. A second hand Dialogic card from Ebay can be quite cheap...

[Guest Roy Jensen]

Thanks...

 

I am still hoping that it is possible to do the following,, but would appreciate assistance in setting it up.

I was hoping that I could use VG to call the number while I sat on another extension and pressed a key on the local telephone to get VG to deliver the message to the answering machines. This kinda works, but I have not figured out how to select the message that is played. I assume that I have to generate a script but I cannot figure out how to run the script after exiting the AcceptAutoCall.wav. Assistance would be appreciated.

 

Your points on 'listen for silence' are well taken. I did mention that BroadcastByPhone does software detection of pick-up. Would it be possible to detect pick-up by frequency analysis? Rings are a single frequency for about 1 second. Voice is _changing_ frequencies on the tens-of-millisecond timescale. Maybe BroadcastByPhone waits for 1 second of silence after observing changing frequencies above a certain threshold. I am a chemist, not a programmer, but this should be very easy to code. Your thoughts?

 

Roy Jensen

[SupportTeam]

Would it be possible to detect pick-up by frequency analysis?

Yes. This is how Dialogic does this on the analog lines.

Frequency analysis could be done but for now I know that there are a number of other high priority projects that we have here so I'd not expect work on this to happen any time soon...