atcomsystems.ca/forum Forums General General Telephone & Systems Information Acceptance of the decline of voice quality

Interesting... Along the same lines I saw this article in VoIP news today

______________________
Hello? Hello? Survey says VoIP Call Quality Down -
https://www.voip-news.com/news/voip-call-survey-072806/

July 28th, 2006 by Marin Perez

BRIX networks says VoIP call quality has
steadily declined over the last 18 months.

Nearly 20 percent of VoIP test callers experienced unacceptable call quality over the last 18 months, according to a new survey by Brix Networks. The results come from almost one million VoIP tests through the company's free VoIP-quality testing portal, TestYourVoIP.com.

Brix said their tests have shown a consistent decrease in overall voice quality from late 2004 though mid-2006. Overall call quality was calculated via Mean Opinion Score (MOS), which rates calls on a scale from one (bad) to five (excellent). Test calls with an MOS of 3.6 or better are typically considered satisfactory, and 81 percent of VoIP calls tested on the site were rated 3.6 or better.

Brix's survey comes on the heels of a recent Telephia survey which found that more than 27 percent of subscribers that were likely to switch VoIP providers cited call quality as the main reason.

"Over the last few years, the global market for consumer VoIP services has grown to nearly 20 million subscribers. These results from TestYourVoIP.com indicate that during this same period Internet call quality has declined," said Kaynam Hedayat, vice president, engineering, and chief technology officer at Brix. "For long-term sustainability, providers of Internet phone services will need to concentrate on the root causes of call quality degradation, including late packet discards, lost packets, and round-trip voice latency."

It could be argued that because of the open nature of Brix's test, the results might be skewed towards disgruntled VoIP users. John Burnham, vice president of marketing for Brix, discounts this notion. Burnham said the “majority” of people using TestYourVoIP.com were doing so to prequalify for VoIP services, not to troubleshoot problems. This, Burnham said, makes the results more like a randomized Internet sample.

Burnham also said despite the fact that Brix sells services that can help fix VoIP quality issues, there was no conflict of interest with the company's findings. “We're in the information business,” he said. “We're essentially neutral.”

--------------

Along these lines does QoS not have a lot to do with the current call quality of VoIP? Many customers not having proper brandwidth and hardware in place yet blaming the service provider?

I checked out the talkswitch.com... interesting.

I'm guessing most of the voice quality problems are outside the IP phone system??? That is, assuming.... that a LAN is in good working order, would it be fair or accurate to say that most of the degradation of would be outside. Beyond that I would think that it would work better over some dedicated circuits than the internet. ??

Richard

You are correct, Richard. Much of VoIP's bad reputation comes from poorly-designed LANs or through the use of the public Internet for connectivity. In a perfect world with proper LANs and a congestion-free Intenet, there would be no question about using VoIP. It is the outside influences that make it so hard to install, manage and support. I am sure that these issues will be resolved in time.

That's the rub. In order to make VOIP viable it takes quite a bit of money and structure to make it work properly. Running two seperate LANS kind of defeats the purpose IMHO.

Much of VoIP's bad reputation comes from poorly-designed LANs or through the use of the public Internet for connectivity... I am sure that these issues will be resolved in time.

Stepping back and looking at the big picture, why do we even want or need to use the internet, LANs, and the rest of the BS for voice when we already have a perfectly functioning system that's been in place for years?

Only reason I can see is because there are some cost savings due to taxes and surcharges but that is sure to go away around the same time quality issues improve, if they ever do.

-Hal

I'm in total agreement with that Hal. Once the tax revenues go down because of enough people switching to VOiP than it will also be taxed at the same rate as other telecommunications. Heck they just got rid of the tax put on in 1898 to fund the Spanish-American War, that tax was a luxury tax on long distance.

The system that has been in place for years is a circuit switched network. Every call is carried on it's own 64Kbps circuit.

LANs, the Internet, and VoIP are running on the newer packet switched network. Since a voice signal is mostly idle time, VoIP uses an average of far less than 64Kbps. The same backbone can carry over three times as many VoIP calls as circuit switched calls -- they could charge half as much for a VoIP call as for a circuit switched call and make > 50% more profit.

Exactly Mike. It uses less actual time and costs less. I dont see VoIP going anywhere but up.

Neither does the industry from the surveys I've seen. Interesting to read others opnions on it though.

The same backbone can carry over three times as many VoIP calls as circuit switched calls

Ummm, doesn't the existing network work that way now? Matter of fact isn't data traffic carried on the same network?

-Hal

AFAICT, no. There are still the two separate networks. While both may be digital, the technology is different and incompatable.

The digital bitstream on a circuit switched network is divided by time into a multitude of channels each 1/8000 of a second in length. Each timeslot is dedicated to the one point to point circuit. While we can select alternate routes on call setup, once the circuit is up end to end an interruption will drop the call.

On the packet switched network the packets are addressed and sent out willy-nilly to find their way to the other end. If a link goes down in the middle of a call the packets simply redirect or resend.

Data traffic can be carried on a dedicated circuit. But, IFAIK, we have not yet packetized the switch to switch circuits. That would involve building a new network to ride the old until the old is removed. I'm also unsure how the SS7 network would work on any such migration.

When I try to learn more on this subject my head aches. I gotta get out of this business.