Rick's Articles

Audio Technology - Volume 1 : Issue 2
The Right of Passage I've just found the secret to making great records, but my head hurts. I think it's a 24-bit hangover. The jitter resolution is so clear, it's astounding... Alesis are shipping their new 20-bit ADAT XT20’s and they sound much better than the ADAT XT’s which sounded much better than the original ADAT’s which sounded pretty good the day everybody bought them.

And Apogee, well, they have some new eight channel converters which sound much better than the Digidesign eight channel converters which were apparently pretty damn impressive the day everybody bought them. Hmm...

What I'm curious about is how the hell would anybody know? Perhaps they conduct some kind of test, a special test that is a scientific fact, indisputable and beyond reproach. Everybody who tries it gets the same results - they have to upgrade to the new stuff because their old stuff somehow turned horrible sounding overnight. In case you're wondering, I'm talking about the A/B test. It's the 'see if you can pick it, before/after' thing. Now I'm not going to describe the A/B test, because if you don't know what it is then I won't have to convert you away from the path of evil. If you do know what it is, stick around because things are going to get ugly.
A/B tests suck.

The A/B test is something I know a lot about. When I'm not producing records I spend most of my time mastering them, and mastering is nothing if it is not the king of all A/B sports. The kind of judgments made in a mastering suite need to be critically informed. Mastering engineers and technicians spend countless hours aligning, calibrating and generally checking everything, every day, so that when they switch between two things, any differences they hear really are differences. It is incredibly hard to get an A/B switching/monitoring system that is really accurate, and when I say 'accurate' I mean to within 0.ldB from 4Hz to 35kHz. I know it's hard because I've had to get one custom built. As far as I'm concerned, a mastering suite is one of the best and only places to perform A/B tests reliably. It's one of the only places where real money gets spent on the subject. I've spent at least $25,000 on switchers and converters and I find it insulting when somebody tells me that some gadget sounds horrible because "we did an A/B test”.

What I'm rattling on about here is that everybody demands the latest digital box because of its great sounding converters, or they swear by their favourite analogue box because of its warm sound. They'll all tell you that they A/B'd them and, because they're your mates, you want to believe them. But the thing is, right now, there are just a few guys in every town saying "yeah, but how did they A/B them?". Most of the time, in most places, an A/B test is a load of crap.

Let's do a quick reality check, here's how the A/B test does not work. It does not work if you are A/B switching using the source selector switches on your home stereo. The chances of the level and frequency response of any of the source inputs being even remotely matched is, well, remote. It does not work if you are A/B switching using the channel strips on your console. The line starts to the left for all those people who think any of their analogue console channels are exactly matched. You'd be horrified if you tested your channels for calibration tolerances. Trust me, unless you're particularly fanatical and have wads of cash to fix things nobody else bothers with, just don't go there. And if you think you can use the bypass switch on your box for a before/after A/B thing, forget it. Very few bypass switches are actually true bypasses, because most true bypass methods either cause clicks or are really expensive to make. If you're not sure, turn the power switch off while the box is in bypass. If you can't hear the signal properly, you don't have a true bypass.

A/B tests are almost always biased towards one product over another, whether intentionally or accidentally. Be especially aware of an A/B test conducted in a studio where the product getting tested is up against something the studio owners have had for years but haven't finished paying off. Leave the room and go to the pub. And please don't bother doing an A/B test on the showroom floor of a dealer trying to sell you something, or in a booth at a trade show.

Leave that one at the circus with the smoke and mirrors and other side show tricks. Even with all the equipment aligned and calibrated, there's no accounting for peoples' egos. Nobody wants to be the only person who can't hear a difference - especially the subtle differences due to new converters, low jitter clocks and the like. Most of these sounds are 'acquired learning' sounds, the kind you tune in to it after using the different products a lot. I own a lot of recording junk, but the stuff I really use and love has its own sound and its own distinct operations. I learnt to record by the test I call the 'rite of passage'. This is the method where you use a particular box in a particular way until it falls over and hurts you, and because it hurt so much the first time, you're very careful when you go that way again.

You soon learn where you can and can't use that box, and before you know it everything sounds cool, nothing burns you and your scars are war stories. You don't care about the grief the box gave you in the past, because you figure you're much better off for having known it. It's a kind of Zen thing. This method, this 'rite of passage', keeps the good stuff in the game and the stuff that's not so good disappears. Things come and go for the right reasons.

But that's not the way the A/B test works. The A/B test makes you put something away forever without letting it find its own flavour. We can't buy everything and test it fairly in the hope that it will all find a place in audio history, but we can edit what we buy and why we buy it. Just about all the new gear is incredibly useable, but that doesn't mean you should run out and buy it right away. Wait until some bright spark starts telling you about the A/B test he did before buying it. If you don't care about his A/B test, then it's time to buy. But if you find yourself intently listening to him prattle on, best hold off a little longer - you're not ready for the burn. You are not ready for the rite of passage.

Oh yeah, you wanna know the secret to making great records? Well, it's a bit like that rite of passage. Just make it sound good and then stop...

Back to article index