Hi, Eduardo,
I had some tests and here are my comments FYI:
- From the view of testing, I suggest to add commands like "qos_mon start", "qos_stop stop", or so to let the qos logging more flexible to measure only part of a conversation instead of the whole conversation. The reason is that sometimes, a phone may reach AutoAttendent or IVR first, and then enter the main conversation...
- If the other party enabled VAD (let's say in a linphone-to-hard_phone case), the measured score would get worse. I know that results from this model which considers majorly packet loss and packet delay. Is there any way to deal with such scenarios to exclude the VAD periods?
I will try to do a patch for #1.
Thanks,
Clement.
On 9/8/05, Eduardo Bezerra <[EMAIL PROTECTED]> wrote:
Suggestions are welcomed.
2005/9/8, Eduardo Bezerra <[EMAIL PROTECTED] >:
> Hi all,
>
>
> Regeane Aguiar ([EMAIL PROTECTED]), Alexandre Passito ([EMAIL PROTECTED])
> and I developed a patch to linphone 1.0.1 to measure speech quality using
> the tool available in http://sourceforge.net/projects/medqos.
> The patch-qosfile-050721 file makes modifications on oRTP to
> generate the trace
> file and the patch-linphonec-qosfile-050721 modifies linphonec to
> accept aditional
> options about the speech measurement.
>
> --
> Eduardo Bezerra Valentin
> Bacharelado em Ciência da Computação - DCC - UFAM
> Laboratório de VoIP - GRSD - UFAM
>
>
>
--
Eduardo Bezerra Valentin
Bacharelado em Ciência da Computação - DCC - UFAM
Laboratório de VoIP - GRSD - UFAM
_______________________________________________
Linphone-developers mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/linphone-developers
_______________________________________________ Linphone-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/linphone-users
