Wireshark-dev: Re: [Wireshark-dev] [Fwd: Re: help-help! Question about roundtrip delay?]
From: "Andreina Toro" <[email protected]>
Date: Tue, 7 Nov 2006 08:50:55 -0400
Hi Mr. Martin, I have a question about the bug in wireshark you were able to identify. Is there a way to download the lines in order tu run Wireshark without this problem?, or do I have to change more than simple lines in the source code?, or do I have to wait until the developers create a new full version?... In that case I´ll create a function in java that identifies that mistake and discards that specific data..
 
Best Regards,
Andreina


 
On 11/3/06, Martin Mathieson <[email protected]> wrote:
[ISP SMTP server is still broken]

I don't know very much about PESQ.  We used it at a previous company as
follows:
- make a call out to an endpoint
- play a standard speech WAV file down the RTP stream
- the other end echoes it straight back
- record the RTP stream that comes back
- run the PESQ software that compares the 2 (linear) WAV files (original +
recorded) pand produces a PESQ score.  The score could decrease due to:
       - jitter meaning there are gaps in the stream (a jitter buffer can
alleviate this, but increases latency)
       - transcoding and echo cancellation through gateways can degrade the
quality of the samples
       - I don't know if delay is something that reduces the score.

This takes a bit of setting up, and I don't think the PESQ part was cheap.
People charge a lot of money for automatic tools to do this kind of QoS
testing.  I don't know if its appropriate for you to pursue this, you'd best
talk to your real supervisor :)

I sincerely wish you the best of luck for your project.  Thanks for sending
the trace files, it was useful to fix the bug in wireshark we identified.
Best regards,
Martin

_________________________________________________________________
Windows Live™ Messenger has arrived. Click here to download it for free!
http://imagine-msn.com/messenger/launch80/?locale=en-gb