New User Help Required To Build Voice Recorder With Asterisk

Home » Asterisk Users » New User Help Required To Build Voice Recorder With Asterisk
Asterisk Users 9 Comments

Hi,

I am new to asterisk, i want to know that is it possible to use asterisk for build voice recording system.

Scenario :
ISDN PRI line (30 line)
I want every incoming & outgoing call has to recorded, but without manual action. system has to auto receive the call.

Please suggest, how should i start and with which hardware / cards it is possible.

9 thoughts on - New User Help Required To Build Voice Recorder With Asterisk

  • You should start getting a PRI card. I have good result with both Sangoma and Digium one. After having configured the card in the system (libpri, dahdi and asterisk part), it is a matter of few asterisk configuration row to save all calls to a wav file.

    For example, if your incoming calls are put in the “incoming” context and your PRI card is identified with the g1 group, the dialplan can be as easy as the following:

    context incoming {
    _X. => {
    MixMonitor(${UNIQUEID}.wav);
    Dial(DAHDI/g1/${EXTEN});
    }
    }

    Leandro

    2012/12/31 Vinod Nadiadwala

  • 🙂 Thought you need just voice recorder without human intervention for Inbound. If you want to record the calls which are getting connected with human, look at Monitor or Mixmonitor. https://wiki.asterisk.org/wiki/display/AST/Application_MixMonitor https://wiki.asterisk.org/wiki/display/AST/Application_Monitor

    Before you get going, start reading on https://wiki.asterisk.org, http://www.voip-info.org/wiki/view/Asterisk(though it is bit outdated) for more information

  • With just one PRI card this should not be an issue, but for larger systems you may consider using something like Oreka to offload the I/O from the Asterisk server…. l.

    2012/12/31 Vinod Nadiadwala

  • Top post for the New Year.

    Yes, if you might scale up to 60 or more simultaneous calls, definitely look at OrecX or RTPTap because you will run into I/O
    issues. Not sure what current hardware can accommodate but it is best not to find out.

    Considering the very low cost of hardware these days compared with the cost of possible downtime, poor audio, lost recordings or whatever else you can assign a monetary value, I always suggest a separate machine for “Passive” recording when dealing with more than a handful of simultaneous calls.

    Thanks, Steve Totaro

  • I don’t know how many I/O can be achieved on a modern hardware, but I don’t think 60 concurrent calls will be a problem. 60 calls are just 4 Mbit/s of data. However can be a good idea to start loading a server and be prepared to share the load on another server.

    Leandro

    2013/1/2 Steve Totaro

  • It depends on what you do with them.

    Years ago, 60 calls would start to crap out audio on live calls and I
    learned that the hard way on a production call center. There was the I/O of just SLIN, then converting to MP3, then transferring to a not too forgiving SAMBA share. Scheduling things for a slower times and moving the MP3 conversion to the mass storage significantly helped while scrambling to find the permanent solution.

    People could increase those numbers with RAMDisk and other tricks but just moving it off the “Phone System” makes more sense.

    Why not engineer something to scale and last without knowing that you will have to revisit it and quite possibly at the most inopportune time, like when you just spent a good deal of money on an advertising spot?

    Thanks, Steve T

  • Mixmonitor also muxes the two sides of the conversation after hangup. That is quite a bit of I/O for 60 simultaneous calls lasting an average of 5-15mins

  • I don’t think this should be an issue, but we have seen a lot of sites going live and discovering too late that they had recording problems. Maybe you won’t need to implement an external recorder, but it’s better to plan in advance, not when you are in production! 🙂
    l.

    2013/1/2 Leandro Dardini