Why are manufacturers constantly opening up new fronts in the loudness war?
Louder is not better. The message has got home at last. We've listened to Bob Katz and the gurus at TC Electronic, and we're keeping an eye on our crest factors. We're asking the mastering engineer to back off the limiter and leave our transients intact. We've noticed that ruthlessly squashed mixes don't sound much cop on the radio. We exchange knowing smiles when Metallica's latest single comes on the jukebox. Hey, it's not like any of us is actually selling any CDs these days, so they might as well sound how we want them to sound. Perhaps we even feel confident enough to hope that our recordings will stand up against the competition regardless of their RMS levels.
So why don't the manufacturers of audio equipment share this confidence? Take software instruments, for example. While I appreciate that your crack team of pop scientists are proud of the months they've spent sampling that tambourine, I'm afraid I do need to leave a little room in the mix for minor details like the lead vocal. And why bother going to the trouble of emulating every transistor in a classic analogue synthesizer if I can't play a note on the resulting plug-in without clipping the master bus? Now that we're all recording at 24-bit, we know we should leave plenty of headroom for signals coming into our computers from outside. So why do manufacturers expect us to be impressed by soft synths that go straight into the red?
Nor are programmers the only guilty parties. When I've managed to turn my soft synths down enough for a vocal to be heard, I often record one using a Shure SM7B microphone. The specs list its sensitivity as 1.1mV/Pa, which works out very well for close-miked sources. Vocals, bass drums, guitar cabinets: they can all be handled by pretty much any preamp. (In fact, most budget IC-based preamp designs actually give better noise performance in the top half of their gain range than when they're doing very little.)
The latest and greatest vocal mic to pass through these doors, by contrast, is Audio-Technica's AT5040. With its innovative quad-capsule design, clever magnetic shockmount and amazing 137dB dynamic range, it could well represent the next leap forward in microphone design. But does it really need to be quite so... loud? The AT5040 has a sensitivity quoted at over 56mV/Pa. In other words, given the same acoustic source, its output will be some 34dB higher than you'd get from an SM7B. Or, to put it another way, stick Captain Beefheart in front of one, and the output signal would stun a horse.
This is far from an isolated example. The sensitivity of capacitor mics has been creeping up for years. When Neumann replaced the U87 with the U87A, for instance, they "improved” its sensitivity by 10dB. Up to a point, that's no bad thing, at least for mics designed to be used in the diffuse field; with quiet sources and long cable runs, it makes sense to try to minimise interference. But we're talking about mics intended for close-up use in the studio, where the risk of interference pickup is negligible — and where the potential for overloading equipment further down the chain is considerable. Really, what's the point in having a mic with an enormous dynamic range, if you can't use it for everyday recording without clipping something else? If this goes on, the cork-sniffers on Web forums will have to forget about comparing preamps and start arguing about the relative merits of in-line pads instead. (Which passive attenuator for a death metal vocalist? What resistors sound creamy with detail?)
The people who design audio equipment are not stupid, so I dare say they have their reasons for doing this. Perhaps someone out there even has hard sales figures which show that there's a real relationship between the number of units a product shifts and the number of electrons it pushes around — but I doubt it. At any rate, I'm sure I'm not the only one who wishes this inexorable drive towards ever-hotter output levels was at an end. Perhaps if we all get together we can make it stop, if we raise our voices and shout loud enough?
Oh, wait. .
Sam Inglis is disappointed to find that there is no longer honey for tea, and that the village clock now stands at 15:00 hours. He is Features Editor for Sound On Sound.