Learn Your Stuff
by Jonathan Wyner
There's an age-old debate between classicism and exploration, improvisation and through composed music, cooking by the book and making up a new recipe based on an idea. The argument in support of the classicists would say something like, “You've got know the rules before you try and break them." The opposing point of view might say, “The old ideas will keep you from discovering truly new or original."
Mapping this to recordings, there are some good examples of “rules" that the classicists would point to as being pillars of good engineering practice. For instance:
The 3-1 rule states that if a microphone is pointing at an instrument, the next-closest microphone in proximity of that instrument needs to be at least three times the distance in order to avoid the worst instances of phase cancellation. This relates primarily to close miking or spot miking. It makes sense, considering the physics of sound. Phase cancellation (or comb filtering, more specifically) is most likely to occur when the amplitude of two signals is reasonably similar along with the spectral content. The 3-1 rule minimizes the problem.
There are instances where it's necessary to violate this rule. Take drum overheads, for example, where the mics might be fairly distant from the cymbals and drums. The 3-1 rule might mean separating the mics by 9 feet if they were 3 feet above the drums. An unlikely mic placement, to say the least. One way around it is to roll the low end out of the overhead mics to avoid phase cancellation in the bass and midrange, where it's most troublesome.
But let's say we wanted to religiously adhere to the 3-1 rule. How in the world would you ever achieve a convincing stereo pickup of an acoustic sound? Blumlein, M-S and spaced omni pairs of mics rely on phase cancellation in order to be effective. In this case it's creative use of phase cancellation in order to achieve an effect.
OK, so none of this is groundbreaking stuff, but you can bet that there are instances where breaking the 3-1 rule will create an unusual and otherwise perfect effect. How important is it to know the 3-1 rule so you can understand when to apply it or not? More to the point, how important is it to understand phase cancellation in order to be able to use it to good effect?
I suppose you could say that worrying about phase might limit your willingness to try certain things, but I tend to come down on the side that says that knowing the theory gives you the freedom to break the rules. It also gives you the tools to fix problems when you need to.
Here's a bit of a mindbender (actually it's not, when you think about it). Take a stereo recording of an instrument such as an acoustic piano. If you run it through an analog mixer, mult the two channels into four, flip the phase on one of the pairs of channels and just add a tiny bit of the phase-flipped channels to the full-on channels, you end up with a subtly bigger image. Would that just phase cancel the signal? Yes, but that's the point. Two analog versions of the same sound will not be 100 percent identical, and it's that very subtle difference that allows you to use this effect. You cannot do the same thing with a digital mixer, by the way, since there you have two absolutely identical signals.
So many engineers are making recordings without knowing the rudiments of signal flow, the physics of acoustics, microphone technology and digital audio that they are often flying blind when trying to make good recordings. If they are sensitive listeners, they might occasionally get lucky and come up with a great recorded sound and mix, but if that's the case it's not often for them to repeat the results, and harder if they need to make a recording in a new environment with new gear.
The moral: Learn your stuff. There are a few audio textbooks out there that are like the bible of audio engineering, and many of them are written using plain English that anyone can understand. Get one, read it, play with the ideas and find out what the tools and techniques are that you have at your fingertips. You never know when you might need to ignore them … or come up with your own new recipe.
Jonathan Wyner (www.m-works.com) has recorded, mixed and mastered more than 5000 CDs during the last 23 years, spanning every musical idiom (and some nonmusical idioms as well!). He is a professor at the Berklee College of Music in Boston. His credits range from the extremely well known (James Taylor, David Bowie, Aerosmith, Bruce Springsteen, Kiri Te Kanawa) to the idiosyncratic and independent artists/labels. A 2007 Grammy nominee, his most recent production has been appearing on PBS since March 2007.
