Telephone conferencing vs Skype (VoIP)

Phone vs Voice over IP introduction

In this article we explore the differences between audio conferences held using traditional telephone networks and the breed of Voice over IP services (VoIP), the most well known of which is probably Skype.

Some of the material covered is necessarily a bit technical but we have kept it understandable and interesting for the lay-person. With apologies to more technical readers, we have deliberately oversimplified and generalised some points in order to achieve this objective.

Potted history of the telephone

As any school kid will be able to tell you, the telephone was invented by Alexander Graham Bell back in 1847 with the specific objective of carrying VOICE traffic. Until very recently all telephone networks were implemented so as to carry voice clearly and reliably.

The technology didn’t change much for many years and was even known as POTS which stands for “Plain Old Telephone System”.

Large companies and Governments have their own private POTS systems but the one we all use is known as the PSTN which stands for Public Switched Telephone Network.

The important word here is ‘switched’. This means that if someone calls you, your phone and their phone will be directly connected together through a series of switches at telephone exchanges. Back in the early days this switching was done by human operators. Nowadays, of course, computers do the switching.

In summary, phone systems were designed to carry voice not data. To carry data over the phone it must first be converted to sound (using a modem).

Potted history of the internet

The internet had its genesis back in the late 1950s as a way of connecting computers together to transfer DATA as opposed to voice as in the case of the telephone above. The distinction is important as we will see later.

The set of conventions or ‘protocol’ used to pass data over the internet is called, “Internet Protocol” which is usually abbreviated to IP. IP works by first chopping up the data to be sent into small chunks called ‘packets’. The packets are then sent one by one over the internet and then reassembled at the receiver.

The packets do not all necessarily take the same path over the internet and can arrive at the receiver at different times and in different orders to which they were sent. Some may get lost in the network and never arrive at all!

In the case of data this doesn’t really matter as the receiver can work out what’s happened and ask the sender to resend the missing or corrupted data packets again. You have probably heard of the mechanism used to add reliability to the transfer, it’s called TCP which stands for Transmission Control Protocol.

Using a combination of TCP and IP it is possible to send data reliably across the internet. Critically however, it is NOT possible to say how long it is going to take. It could take a fraction of a second or all day. So long as it arrives, the protocol has not been breached.

In summary, the internet was designed to carry data not voice. To carry voice, the sounds must first be converted to data (using a codec).

How does Voice over IP (VoIP) work?

Very simply, VoIP works by converting sound into data and sending it over the internet. The internet does not care what the data it carries represents. For instance, it makes no distinction between the data needed to read this web page and the data needed to carry a phone call. Data is just data.

So why does all this matter?

It matters because now that you understand the way VoIP works you can appreciate that as the internet makes no guarantee about how quickly data is delivered, there are no guarantees that there won’t be delays on the line.

The variable nature of the system explains what you have probably experienced in practice. Sometimes VoIP systems work well, sometimes they don’t. It depends on the load on the internet.

It is actually a bit worse than it seems, here’s why.

We mentioned Transmission Control Protocol earlier as a method of ensuring data is delivered reliably. The problem with using TCP is that it works by requesting that lost packets of data be retransmitted before data is presented to the receiver. This takes time. With most data transfer, a short delay is preferable to receiving corrupted data.

Why VoIP lines are often unclear and noisy

The problem with sending voice is that delays DO matter and matter very much.

It would be completely unacceptable to have to wait for a few seconds for your voice to be heard at the other end of the line. VoIP systems get around this problem by simply ignoring missing or corrupted data packets. When data is lost you experience it as ‘pops’ and noise and echos on the line.

Traditional telephone (POTS) systems do not suffer from that problem.

Why this is particularly disruptive on audio conferences

When using VoIP for a one to one conversation between just two parties you can put up with the odd noise on the line. However as soon as you get more than one or two callers on the line, the noise from each line is cumulative and can become very disruptive.

Audio conferences held using traditional telephone systems do not suffer from this problem and therefore it’s possible to hold crystal clear conferences with tens or even hundreds of participants.