Gordon Laing

Three years ago, 40 experienced female clerical workers were randomly assigned to one of two offices: the first quiet, the second with background noise typical of the modern workplace.

Three hours later the environmental psychologists conducting the experiment discovered that the workers in the noisier office were more stressed, less motivated, and less likely to make ergonomic adjustments to their desks and chairs.

The paper, published in October 2000's Journal of Applied Psychology, concluded that low-intensity office noise was bad.

Any office employee will tell you that noisy workplaces result in lower concentration and reduced productivity.

Subconsciously suppressing background noise during an entire working day, coupled with the inability to escape from its incessant hum, might increase stress levels, but could the whirr of PCs really be bad for our health?

Whether you believe this to be hogwash or otherwise, there's no denying the enormous interest in quieter PCs.

As more computers appear in homes, their noise is becoming a serious issue. The latest IT conquest for the living room is almost entirely dependent on silent or at least very quiet PC systems.

An increasing number of buyers are requesting quiet options and a whole new market of considerate components has sprung up. So why is there so little information or regulation on equipment noise?

Unlike other emissions which enjoy a raft of certification and labelling, it's nigh-on impossible to find anything relating to noise.

PC adverts list every specification imaginable, apart from the one thing which is more likely to annoy you long-term.

Even on the rare occasions when noise is mentioned, the device is often just described as being 'whisper-quiet', which is like abandoning the measure of gigabytes and simply saying a hard disk is 'plenty big'.

Admittedly some manufacturers quote a decibel figure but, unless you know how far away the measurement was taken and with what weighting, it's virtually meaningless.

Measuring noise can be difficult, but people want quiet systems and some means of knowing what they're buying.

So how tough a problem is measuring and labelling noise? Those with an insatiable appetite for information should check out www.silent.se, home of Tomas Risberg's appropriately named Silent PC Website.

Risberg explains the pitfalls of measuring noise, makes a case for certain techniques and describes the ISO standards in place.

Yes, there's already an ISO standard (9296) which applies exclusively to the noise levels of IT equipment, and another (7779) which describes how measurements should be taken.

Noise certification also exists, from the likes of TCO. So, if standards and certifications exist, along with consumer demand, how come we don't see badges on PC systems boasting low emissions?

The first answer is cost, as measurements conforming to ISO 7779 understandably involve more than just pointing a cheap sound pressure level meter at a box. Making proper measurements under controlled conditions is expensive.

Second, and ultimately more problematic, is consistency. It's easy to certify monitors and justify the cost because they're mass-produced. One certification could apply to a model produced many thousands of times over a long period.

PCs, on the other hand, are commonly custom-built. Rarely a month goes by without new configurations using faster processors or graphics cards, while varying supplies could also see different hard disks or other components fitted.

Consequently it's virtually impossible to certify an entire PC system for noise unless it's a mass-produced box whose specification remains constant when sold over a long period.

To be fair, some of the largest PC companies, such as Dell, HP and IBM, do state audible emissions, but often hide them away under environmental specification sections.

The most prominent statement of system noise I've seen is for Via's Epia motherboards where noise emission ratings are listed second from the top of the spec sheet. But Via can't control what power supply and drives are hooked up to its motherboards.

A description of overall noise emissions should be high on the list of commonly advertised PC specs. Several people I know who've recently bought a PC checked the usual list of specs, only to be shocked by the racket it made when powered up.

I bet they'd have traded a few megahertz for the benefit of lower-noise components had they known or been given the choice.

Ultimately it boils down to consumer demand and how much we value low noise compared to other aspects of an entire system. We're in a catch-22 of constantly wanting higher performance at lower prices, while finding the resulting noise increasingly intolerable.

For lower noise levels, we're going to have to accept a compromise of lower performance or higher prices, but either way I'd like to be presented with comparative figures and a low-noise option rather than a nasty surprise when I first power up a new PC.

The jury may still out on whether PC noise is bad for your health, but there's no denying that it can be a royal pain in the neck.

Like this story? Spread the news by clicking below:

 del.icio.us     Digg this     reddit!