Carsten,

Thank you for your reply.

Carsten Bock wrote:

Hi,

Did you try to post your problem on the RTP-Proxy-Users' List? http://lists.rtpproxy.org/mailman/listinfo/users Probably, you might get more help there....

Yes i have posted on rtpproxy users list.

Did you try to extract the Audio with Wireshark? If Wireshark can play the Audio (and i assume it is correctly implemented), then the recorded stream as such is correct. Then you can check, if the bug lies in rtpbreak (or as a next step: sox). Just to limit the possible sources of the problem...

I have not tried Wireshark to extract the audio. I have been using rtpbreak to generate the .raw file. Subsequent to which i use sox to convert the .raw to a .wav file. By importing the the .raw file into Hypersignal software, we found that the .raw file doesnt entirely seem to be composed of speech samples, so there could be a problem at the rtpbreak step.

This is the command i used to convert rtpproxy's capture file to .raw format

rtpbreak -W -r capturefile.rtp

Am i missing something here ?

Thanks and Regards, Vikram.

Carsten

2010/1/14 Vikram Ragukumar <vragukumar@signalogic.com mailto:vragukumar@signalogic.com>

Hello,

An update,

I tried using sox to convert the two .raw files into 2 mono channel
wave files. The command line i used is below :
sox -r 8k -b -c 1 -u rtp.0.0.raw rtp0.wav
sox -r 8k -b -c 1 -u rtp.1.0.raw rtp1.wav

When i listen to the .wav files, i hear speech but it is buried in a
lot of noise. During blank periods (periods of no speech) there is a
constant volume high pitched noise. Also during periods of speech,
there seems to be bursts of noise in the background.

The other engineer i work with and i, think that it is possibly
because non-speech data is being interpreted as speech.

What switch options should i change while invoking sox from the
command line to get rid of the noise?

Thanks and Regards,
Vikram.



Vikram Ragukumar wrote:

    Hello,

    I used Kamailio+rtpproxy to record a session and rtpproxy
    outputs the following files

    long_file_name.a.rtp, long_file_name.a.rtcp,
    long_file_name.o.rtp, long_file_name.o.rtcp

    http://www.rtpproxy.org/wiki/RTPproxy/FAQ
     From the Rtpproxy FAQ above, i tried to extract the audio using
    rtpbreak and sox.

    rtpbreak -W -r long_file_name.a.rtp
    rtpbreak -W -r long_file_name.o.rtp

    The above commands generate rtp.0.0.raw, rtp.1.0.raw.

    Then when i run sox using
    sox --combine merge -r 8k -A rtp.0.0.raw -r 8k -A rtp.1.0.raw -t
    wavpcm -s out.wav i get the following errors :

    sox: invalid option -- -
    sox: -c must be given a number

    Is there a switch/anything else that i am missing ?

    Thanks in advance,
    Regards,
    Vikram.






_______________________________________________
Kamailio (OpenSER) - Users mailing list
Users@lists.kamailio.org <mailto:Users@lists.kamailio.org>
http://lists.kamailio.org/cgi-bin/mailman/listinfo/users
http://lists.openser-project.org/cgi-bin/mailman/listinfo/users

-- Carsten Bock Schomburgstr. 80 22767 Hamburg Germany

Mobile +49 179 2021244 Home +49 40 34927217 Fax +49 40 34927218 mailto:carsten@bock.info mailto:carsten@bock.info

Mike Ditka http://www.brainyquote.com/quotes/authors/m/mike_ditka.html

• "If God had wanted man to play soccer, he wouldn't have given us arms."

I ran into a similar issue and while searching for a solution, I found a few threads describing the same problem but no solution.It turns out that the sox command must have the proper arguments in order to properly decode raw files.Feeding sox with an u-law stream and asking to be decoded as an a-law stream will create garbled sound. I posted a more descriptive procedure on how to extract audio from rtpproxy pcap files here:http://voipembedded.wordpress.com/extracting-audio-from-calls-recorded-with-... Regards,Ovidiu Sas On Thu, Jan 14, 2010 at 5:37 PM, Vikram Ragukumar vragukumar@signalogic.com wrote:

Hello,

I would also like to mention that the rtpproxy recorded .rtp files were generated using the following command

rtpproxy -l listen_ip_address -s udp:127.0.0.1:7722 -a -P -F -r record_directory_path -S spool_directory_path

where -P indicates that files would be recorded in the pcap format.

Thanks and Regards, Vikram.

Vikram Ragukumar wrote:

...

Carsten,

Thank you for your reply.

Carsten Bock wrote:

...

Hi,

Did you try to post your problem on the RTP-Proxy-Users' List? http://lists.rtpproxy.org/mailman/listinfo/users Probably, you might get more help there....

Yes i have posted on rtpproxy users list.

...

Did you try to extract the Audio with Wireshark? If Wireshark can play the Audio (and i assume it is correctly implemented), then the recorded stream as such is correct. Then you can check, if the bug lies in rtpbreak (or as a next step: sox). Just to limit the possible sources of the problem...

I have not tried Wireshark to extract the audio. I have been using rtpbreak to generate the .raw file. Subsequent to which i use sox to convert the .raw to a .wav file. By importing the the .raw file into Hypersignal software, we found that the .raw file doesnt entirely seem to be composed of speech samples, so there could be a problem at the rtpbreak step.

This is the command i used to convert rtpproxy's capture file to .raw format

rtpbreak -W -r capturefile.rtp

Am i missing something here ?

Thanks and Regards, Vikram.

...

Carsten

2010/1/14 Vikram Ragukumar <vragukumar@signalogic.com mailto:vragukumar@signalogic.com>

Hello,

An update,

I tried using sox to convert the two .raw files into 2 mono channel    wave files. The command line i used is below :    sox -r 8k -b -c 1 -u rtp.0.0.raw rtp0.wav    sox -r 8k -b -c 1 -u rtp.1.0.raw rtp1.wav

When i listen to the .wav files, i hear speech but it is buried in a    lot of noise. During blank periods (periods of no speech) there is a    constant volume high pitched noise. Also during periods of speech,    there seems to be bursts of noise in the background.

The other engineer i work with and i, think that it is possibly    because non-speech data is being interpreted as speech.

What switch options should i change while invoking sox from the    command line to get rid of the noise?

Thanks and Regards,    Vikram.

Vikram Ragukumar wrote:

Hello,

I used Kamailio+rtpproxy to record a session and rtpproxy        outputs the following files

long_file_name.a.rtp, long_file_name.a.rtcp,        long_file_name.o.rtp, long_file_name.o.rtcp

http://www.rtpproxy.org/wiki/RTPproxy/FAQ         From the Rtpproxy FAQ above, i tried to extract the audio using        rtpbreak and sox.

rtpbreak -W -r long_file_name.a.rtp        rtpbreak -W -r long_file_name.o.rtp

The above commands generate rtp.0.0.raw, rtp.1.0.raw.

Then when i run sox using        sox --combine merge -r 8k -A rtp.0.0.raw -r 8k -A rtp.1.0.raw -t        wavpcm -s out.wav i get the following errors :

sox: invalid option -- -        sox: -c must be given a number

Is there a switch/anything else that i am missing ?

Thanks in advance,        Regards,        Vikram.

_______________________________________________    Kamailio (OpenSER) - Users mailing list    Users@lists.kamailio.org mailto:Users@lists.kamailio.org    http://lists.kamailio.org/cgi-bin/mailman/listinfo/users    http://lists.openser-project.org/cgi-bin/mailman/listinfo/users

-- Carsten Bock Schomburgstr. 80 22767 Hamburg Germany

Mobile +49 179 2021244 Home +49 40 34927217 Fax +49 40 34927218 mailto:carsten@bock.info mailto:carsten@bock.info

Mike Ditka http://www.brainyquote.com/quotes/authors/m/mike_ditka.html  - "If God had wanted man to play soccer, he wouldn't have given us arms."

---

Kamailio (OpenSER) - Users mailing list Users@lists.kamailio.org http://lists.kamailio.org/cgi-bin/mailman/listinfo/users http://lists.openser-project.org/cgi-bin/mailman/listinfo/users

---

Kamailio (OpenSER) - Users mailing list Users@lists.kamailio.org http://lists.kamailio.org/cgi-bin/mailman/listinfo/users http://lists.openser-project.org/cgi-bin/mailman/listinfo/users