I don’t think there’s an API that you can use to tap on the RTP stream. You can always record your call in stereo and post-process the file, if that’s valid for what you’re trying to do.

If not, check the MRCP protocol. I guess you can write an MRCP server (or use an existing one), and stream the audio from FS using mod_unimrcp. That’s how you integrate 3rd party ASR systems (e.g. Nuance’s Recognizer) with FreeSWITCH.

From: <freeswitch-users-bounces@lists.freeswitch.org> on behalf of Prashant Choudhary <get.prashant.007@gmail.com>
Reply-To: "freeswitch-users@lists.freeswitch.org" <freeswitch-users@lists.freeswitch.org>
Date: Sunday, July 19, 2015 at 4:24 PM
To: "freeswitch-users@lists.freeswitch.org" <freeswitch-users@lists.freeswitch.org>
Subject: Re: [Freeswitch-users] mod_managed vs eventsocket

Hello,
Now I have even managed to do outbound calls. I have a real problem now.
How can I get the audio stream from the person on the other end. I want the audio stream to be sent to my API in real time. I want to detect silence as well, but we can do that by experiment. Only thing is that I am unable to get the audio stream. And this stream should only have the voice of person being called, not mine(i.e. the incoming audio). Is it possible??
Please help

On Sat, Jul 18, 2015 at 2:57 PM, Prashant Choudhary <get.prashant.007@gmail.com> wrote:
Sorry for my last mail, which was a little more generic. Thing is, I am able to receive and manage Inbound calls. But I am unable to figure out, how to make outbound calls through mod_managed. I was able to do the same through event socket API. One benefit of event socket was that I could run it as separate console application , but I am not sure whether it is possible through mod managed as well.
SO again my question remains the same. But it is more towards outbound calling.

On Sat, Jul 18, 2015 at 1:11 AM, Prashant Choudhary <get.prashant.007@gmail.com> wrote:
Hello All,

I have worked with LUA and freeswitch. Now I want to migrate to .Net API as I want to experiment with machine learning using sample calls.  I am trying to control around some 2-300 calls directly through my API. This may require the call audio stream to be forwarded in real time to .NET
 I a few questions that might be very silly.

1. Which one is better to achieve this, mod_managed or ESL.
2. I was able to find some examples related to eventsocket, but same for mod_Managed was rare. Can you suggest me any doc(of-course except the wiki link) or any email chain which explains this or anything else. This will be really a big help for me. As currently I am stuck. And I still find the LUA way easier.

Thanks