Audio hardware

Turn off your PC speakers for a day and you'll realise just how crucial sound is to the fun things you do with your PC. It therefore stands to reason that a great set of speakers will provide a shot of adrenalin straight to the heart of a PC that's built for entertainment. It's akin to switching from a 15in CRT to a 24in widescreen LCD - games spring to life, films become truly cinematic and music fires on all cylinders. Add surround sound to the equation and it's as though your monitor just popped into three dimensions - and all for considerably less than the price of a Zalman Trimon 3D screen.

This Labs test is divided into two sections, the first of which is devoted to headsets. High-octane warfare might be music to your ears, but your cohabitees, neighbours and the Environmental Protection Agency may have other ideas. Headsets keep it all under wraps and allow your ears to enjoy some seriously good sound without you spending a fortune. Meanwhile, their integrated microphones are perfect for VoIP, creating podcasts and taunting your opponent just before you remove their upper torso with a rocket launcher.

In the second half of this month's Labs test, we review seven surround speaker systems with prices ranging from £60 to £258. The best systems aren't cheap, but they're no more expensive than a high-end monitor or graphics card, and are just as effective at boosting your gaming pleasure.

It's also worth bearing in mind that the speaker market moves much more slowly than monitors and graphics cards, so this is one upgrade that won't seem dated in six months.

In some respects, testing speakers and headsets is a cinch. Once we had wrestled with miles of cables to plug them all in, it was quickly clear which sets sounded great and which didn't. However, making precise evaluations is harder, as a piece of music could sound best on one set at a particular level, but better on another after we turned them all up. Switch to a different track, and a third set might sound better. Below is a description of how we went about avoiding temporary insanity and arriving at some definitive conclusions.

The headsets can be thought of as two discrete products: headphones and a microphone. Fortunately, these particular microphones aren't expected to capture distant birdsong or a grand piano - all they have to do is record a human voice from a couple of inches away. Still, we looked out for a number of different traits during our tests. A clear, full-bodied tone was our number one priority. Upper frequencies needed to be bright and clean to aid intelligibility, but we wanted to hear richness and depth in the lower frequencies too. Excessively low frequencies can make vocals sound boomy, but all the microphones on test roll off low frequencies below 80Hz, so this wasn't an issue.

Another potential problem with vocal microphones is that plosive sounds - the letters B and P - cause blasts of air that hit the microphone diaphragm and result in loud pops. Foam windshields help to solve this problem, but the best solution for microphones is to adjust their position so they're not in direct firing range of your breath. Once again, after doing this for each microphone, plosives ceased to be a problem. We also listened out for background noise and hum, but this was much more dependent on the sound card's microphone input than the microphone. However, this remained relevant in the case of USB headsets, on which the sound card is in effect built into the unit.

Assessing the quality of headphones and speakers is a largely similar process. In both cases, we tested using a variety of music, games and DVDs, including 'Spider-Man', at various volumes. We concentrated predominantly on music tests - like any music fan, we know how certain tracks should normally sound, which makes it easier to detect flaws. Music tests also made it easy to play the same section of audio on a loop, and quickly switch between different systems using an 8-way audio switching box. This device doesn't support surround sound, but a looped playback of a DVD action sequence and some enthusiastic speed-rewiring provided the most accurate measure of each system's surround-sound capabilities.

In theory, we were looking for an even response throughout the 20Hz to 2KHz audible frequency range, plenty of volume and no distortion. However, our ears aren't calibrated measuring devices, so we had to judge by what sounded accurate and pleasing. The problem is that some inaccuracies - such as over-accentuated bass - can sound pretty good in some tests but less so in others. To minimise this as much as possible, we auditioned music from a wide variety of genres including pop, rock, jazz, classical, dance and various shades in between. If a supposed fault sounds good across the board, then it wouldn't be fair to regard it as a fault. Indeed, many headphone and speaker designers deliberately tune their products to sound flattering rather than accurate, and it's a matter of personal taste as to whether flattery or honesty is more important.