TrueNet needs more WorldxChange volunteers

Forums › WorldxChange / XNet / VFX › TrueNet needs more WorldxChange volunteers

1 | 2 | 3

gchiu

397 posts

Ultimate Geek

Reply # 605463 5-Apr-2012 08:26
Do ISPs prioritize VOIP traffic originating on their networks even if they're not the ones providing the service?

maverick

3442 posts

Uber Geek

Trusted

WorldxChange

Reply # 605497 5-Apr-2012 08:58
Okay so I will go through the actual reason on why I think this methodology is flawed and why I have an issue with this. Test environment: BOF FTTH Connection 30/10 (Actual FTTH circuit using GPON) WRP400 Router SPA525 VoIP Phone Laptop (WxC Magic Box) Testing The BoF circuit was loaded with live Internet traffic which included ISky media streaming and HTTP downloads, a VoIP test was placed basically using the same methodology i.e running a MOS Test to a hosted box not within the WorldxChange network similar to what Truenet are proposing. At the same time a VoIP call was placed using the VFX service. These are the 2 calls one from the VFX service one by the test program. This is the Traffic Data from this connection, Note the RTP stream in blue at 07:39 this is the VoIP RTP component of the traffic at the time, it is very small amount of traffic in basically a maxed out FTTH 30Mb connection, Now this is the what the hosted MOS measurement saw when using a G711 codec. A MOS score of 3.5, this is an unacceptable call to me The reason is due to packet loss from the hosted Server (i.e this could be Truenets Media Sink) Now at the same time I had a call up using our VFX service, and here lies the difference, the Call does not drop below 4.2 or even close to it The stas form this call are pretty much perfect even with the circuit being hammered even when the router is a low cost RGW (WRP400) So basically what we have here is 2 vastly varying results using your methodology John a hosted media sink sitting outside a providers network, so you will report on what you see a MOS of 3.5 based on a very short duration test, yet my 2 minute call has so much better MOS statistics, how is this then a fair and accurate reflection of a providers VoIP service... I know which one I trust, yet you would be comfortable putting these type of stats in the public domain. Yes I am a employee of WxC (My Profile) ... but I do have my own opinions as well

dolsen

622 posts

Ultimate Geek

Subscriber

Reply # 605525 5-Apr-2012 09:30
JohnButt: As we have done in the fixed line broadband results, we compare service with service. VoIP is a voice (& video) connection service and it should be fairly measured by us, independent of how the service is built. The thing I see an issue with is when the way the service is built has a material impact on the results, I do not think it is fair on those who do go to the effort to do it correctly (whatever that means) when they are measured on something else.

maverick

3442 posts

Uber Geek

Trusted

WorldxChange

Reply # 607681 11-Apr-2012 09:37
John, when are you sending out these test boxes, I haven't received notification yet ? Yes I am a employee of WxC (My Profile) ... but I do have my own opinions as well

freitasm

BDFL
43688 posts

Uber Geek

Administrator

Trusted

Geekzone

Subscriber

Reply # 607684 11-Apr-2012 09:41
John, are you sending boxes out to VFX users? I use VFX and I am not on WorldxChange... Mauricio Freitas (Geekzone Admin, Devil's Advocate, my blog, technology disclosure, my tweets, Web Performance Optimization) Geekzone Price comparison \| Amazon Kindle order page \| How to get the best out of Geekzone \| New Zealand IT Jobs \| Geekzone News \| Jinglebugs

JohnButt

213 posts

Master Geek

Trusted

Subscriber

Reply # 607704 11-Apr-2012 10:29
Sorry to take so long to reply - been away In answer to Mavericks superb post, We do need to be sure we have the ability, technology and accuracy before we publish anything in the VoIP space - hence we have not yet published. Interesting comparison that you make between your "hosted MOS measurement saw when using a G711 codec" of 3.5 and your own "MOS Estimate" measurement chart of 4.2. I can say that our MoS measurement we are doing has never got as low as 3.5, it is nearly always like your chart, but at 4.38 - not a number I trust, so we still have more work to be sure of our answers - hopefully your input can help. Your box should be on the way this week - we have a large number to process, thanks to Geekzone Mauricio - we are not yet planning on sending out separate boxes for VFX or 2Talk customers :-) www.truenet.co.nz We are seeking more volunteers here :-)

gchiu

397 posts

Ultimate Geek

Reply # 611173 18-Apr-2012 14:45
Anyone got their test box yet? Or a confirmation letter?

maverick

3442 posts

Uber Geek

Trusted

WorldxChange

Reply # 611174 18-Apr-2012 14:46
yep we do Yes I am a employee of WxC (My Profile) ... but I do have my own opinions as well

JohnButt

213 posts

Master Geek

Trusted

Subscriber

Reply # 611321 18-Apr-2012 17:49
gchiu: Anyone got their test box yet? Or a confirmation letter? Hi, Your probe is awaiting the address verification. Have you received the letter yet? happy to send it again if not, but please check your address if that is the case. A probe will be on the way within days of your AV code being entered on the website Looking forward to having your probe up and running www.truenet.co.nz We are seeking more volunteers here :-)

Dratsab

1387 posts

Uber Geek

Trusted

Subscriber

Reply # 611335 18-Apr-2012 18:16
JohnButt: Looking forward to having your probe up and running Noooo.... Chuck Norris has abolished the periodic table of elements. The only element he recognises is the element of surprise!

1 | 2 | 3