Hi Massilimiano, we are also looking into Network Impairment emulators and PESQ tools. What are you using and could you make any recommendations? BR/Olle -----Original Message----- From: pjsip-bounces@xxxxxxxxxxxxxxx [mailto:pjsip-bounces at lists.pjsip.org] On Behalf Of Massimiliano Montevecchi Sent: den 25 september 2008 11:09 To: pjsip at lists.pjsip.org Subject: Re: Degradation of Speech Quality > > [Montevecchi] Hi, Benny. > Thank you for your answer. I can perform PESQ tests on each pjsip > version you are going to release. As I told you, I have an automatic > test environment based on windows xp hosts (cpu: intel Pentium M 1.8 Ghz). > > [Montevecchi] Hi Benny Hi Massimiliano, that's an awesome offer! And I think I will take that. :) I would not ask you to test each pre-release version though, I suppose that would be too much of a burden. But it would be great if you could report any degradation in voice quality every now and then. [Montevecchi] Ok, I will test audio quality for each new main stone. But for a start, Nanang has just made some changes in the latest SVN, could you fetch it and re-run the test? We think that that should improve the score, but haven't verified that. So it will be great if you could test this, since we're about to release rc3 very soon. [Montevecchi] This morning I am going to run a new test campaign with the new fix regarding enhancement #638. I apply the only this fix to pjsip.1.0.rc2 > A single call of my tests is about 1 minute long and each party plays > the same speech sample 4 times. PESQ evaluation is performed for each > played sample. I noticed (that confirm your opinion) that often the > worst PESQ score is the first one. But sometimes also some PESQ score > in the middle is > bad. > > I see. Are you using pjsua for the test, or a custom application? [Montevecchi] Yes, I do! I am using pjsua with very limited changes in pjsua_app.c > If it could be helpful I can provide you a lot of traces. > > Couple of traces/logs from the failed tests would probably help. [Montevecchi] Ok. How can I send you the logs? they are bigger than 100Kb. I have other 2 questions. The first one is about pjsua call statistics. I have put a network impairment emulator between the two hosts involved in the voip call, that's in order to test the network statistics reported by pjsua. I noticed something strange when I simulate a jitter with the impairment emulator: pjsua reports a lot of packet loss (see jitter.log). The jitter introduced by the impairment follows a uniform law for the distribution of the probability (please look at jpg in attach). Could you please explain these statistic values? The second one is about problem during the call establish session (INVITE). My test environment is based on OPENIMSCORE as proxy and registrar. So when I do a call it is mediated by OPENIMSCORE. Sometimes happens that the INVITE sent by OPENIMSCORE to the called host is bigger than 2000 bytes. In this case pjsua doesn't manage that pdu. How can I solve the problem? Do I have to change the #define PJSIP_MAX_PKT_LEN? Thnak you again Best Regards. Massimiliano
