So, what's the real story?
Reading down to the second paragraph of the Brix release, we get a clue. It announces the availability of a downloadable desktop readout applet for Brix's www.TestYourVoIP.com facility. This is a website where visitors can click to run a test connection from their location to any one of seven citiesBoston, Helsinki, London, Montreal, San Jose, Sydney, and Vienna. The Java application running on the site then reports an MOS score for the specified connection. (At the time of this writing, only one city, Boston, shows an acceptable MOS rating of 3.6, the rest are in the unacceptable zone, ranging from 2.9 to 3.4.)
Brix's press release pronouncements are based on approximately one million TestYourVoIP.com ratings compiled between late 2004 and mid-2006. According to Minacom, however, Brix's test calls are all essentially PC-to-PC calls over the unmanaged Internet, calls you might make using a free VoIP service like Skype or Gizmo Project. (Brix spokesman John Ricciardone explained to VoIPplanet.com that this assessment wasn't quite accurate, but didn't press the point.)
The Minacom release points out that "PC-PC VoIP quality is subject to many diverse impairments, including firewall settings, computer performance, antivirus installations, high-compression codecs, and Internet bandwidth shared with gaming, file downloads, web surfing, and email."
By contrast, VoIP offered by service providers is switched using telecom grade equipment, uses lower-compression codecs, and is prioritized over regular Internet traffic using sophisticated, standards-based multimedia telephone adapters (MTAs), maintained and monitored by the operator. It is over such managed broadband and cable linesas well as the PSTNthat Minacom's testing is run.
Minacom regularly updates its free QoS Benchmark Report with results from hundreds of test calls each month to public destinations worldwide over the PSTN and a number of IP-based network types, including broadband, DSL VoIP, and cable VoIP. (For what it's worth, Brix's Ricciardone pointed out that this is a much smaller sample than the one on which Brix's conclusions were basedthousands of test calls, as compared to roughly a million.)
In any case, the latest Minacom results indicatecontrary to the Brix findingsonly 1 in 10 VoIP calls worldwide1 in 50 in North Americafall below the MOS acceptable threshold of 3.6, while 85 percent of VoIP calls exceeded the average MOS score for PSTN calls. In North America, PSTN calls still connect more quickly (4.3 seconds on average) than VoIP calls (5.7 seconds), but worldwide, VoIP calls connect nearly two seconds faster, on average, than PSTN calls. And that VoIP connection times are accelerating, having picked up two seconds over the past year.
All this should be heartening for those seriously considering adopting IP-based or hybrid phone systems. Minacom's findings certainly undercut one of the traditional barriers to VoIP adoption: the notion that VoIP call quality is not up to the telcos' "toll" standards.
So why the contradictory messages from two vendors in the same business?
They represent two sides of the same business-promotion coin: For VoIP to compete with traditional telephony, the IP networks carrying VoIP calls need to be accurately monitored and carefully managed. Both Brix and Minacom offer such services to service providers the world over.
This article was first published on VoIPPlanet.com.