I like maths. There, I said it. I think it a beautiful thing and, sadly, often underused in the recording world. I am, however, often in a rush and I don’t want to be fiddling around for 10 minutes with pen, paper and a calculator to work out how to record something. Below I’m going to describe a pain free way of getting the best from your main pair.
Main pairs: we need them and we love them but do we actually know how to get the most out of them? Those of you who regularly record drums might have experimented with using a stereo XY pair rather than the standard spaced overheads at one time or another, or those of the classical music persuasion may well regularly use such standards as ORTF (Office de Radiodiffusion Télévision Française) and NOS (Nederlandse Omroep Stichting) as a stereo pair in all manner of recordings. But what’s the reason behind these choices? Do we really know what situation calls for which setup? I’m mainly going to talk about main pairs in the classical recording domain but all of this can be extremely relevant for pop recordings (particularly drums and horn sections etc.).
To understand what we are doing and why, it is important to have a basic concept of how stereo sound works. As you know, even if you had your eyes closed, if somebody was talking on your left hand side you would perceive them to be at your left. This is due to two main factors: level differences and timing differences. For the sake of argument, let’s say that there is a clear path from the person on your left to your left ear. For that same sound to get to your right ear it will have to get through some of your head, or diffract around and have to travel through a greater amount of air. Both of these cause a lower level signal to reach your right ear. Similarly, if the sound from the left has to travel further to get to you right ear than to your left ear, it will take longer to get there. This means that there will be a time difference, albeit very small, between your brain receiving the two signals from different ears. When you’re listening to recorded music the speakers or headphones are reproducing these level and timing differences to give you a stereo image.
Microphones can pick up these two stereo elements quite easily. If two omni-
XY pairs rely predominantly on level differences and spaced omni microphones mostly use time of arrival differences to generate a stereo signal. Each method has its pros and cons (time of arrival sounds rubbish when played back in mono, however, the omni mics you use in a spaced pair sound awesome etc.) but a combination of the two sounds like a pretty good idea.
Near coincident cardioids are the most common type of stereo microphone technique that uses both level and timing differences to capture a stereo image. ORTF and NOS both fall in this category. The idea is that you use two cardioid microphones, a set distance apart with a certain angle between them.
I am a firm believer that, in classical recording, the two most important things are level balance and stereo image; everything else is secondary. Making sure that you have an appropriate stereo microphone technique is very important. If you put up a main pair without really thinking about what you’re doing you’re likely to squash the stereo image either in the middle or at the edges. For instance, if recording four instruments equally spaced in a line, putting up the wrong pair could cause instruments 1 and 2 to appear to be fully left hand side and the 3 and 4 to be fully right leaving a big ol’ gap in the middle. Or, equally embarrassingly, 1 and 4 appear to be in the right place but 2 and 3 are sitting on each other’s laps in the middle. It’s important to get it right!
ORTF is great; one of France’s finest exports, in my mind. It consists of two cardioid microphones 17cm apart with an angle between them of 110°. This works well for an acceptance angle of about 100° (the total angle from your main pair that what ever you’re recording takes up). This is because it will have a very evenly spread image within that range, thus giving an accurate result.
Obviously, if you’re trying to record a group who fill a 100° sound stage from where you are putting up your main pair up, then ORTF is the one for you! The difficulties arise when people just use these standards no matter what acceptance angle they are looking for.
We could try and come up with (and then remember) a load of standards from which we can choose one that will be roughly right. Thankfully Michael Williams (a very clever chap) has come up with some stunning maths and physics to produce a main pair cheat sheet – The Williams Curves. These curves plot the angle between microphones vs. the spacing required to record a very accurate stereo image for different acceptance angles.
Each curve shows the relationship of the angle and distance between microphones required to get the best stereo image for a particular acceptance angle (described as angle either side of on-
If you follow the ±50° curve and find the point with an angle between the mics of 110° you will find that it correlates with a distance between the capsules of pretty close to 17cm, thus showing that ORTF, for an acceptance angle of 100°, is as good as anything else I can suggest!
One of the beautiful things about these curves is that it offers you so much flexibility depending on what you’ve got to work with. If you have particularly long microphones you might need them to be further away from each other so you can actually get the XLR in! Not a problem: just look at the curve you’re going for and find the angle that corresponds to the mics being further apart. Conversely, you can do the opposite if you have a particularly short stereo bar: just look for the correct angle for a pair that are closer together.
I have found these curves extremely useful and my recordings have dramatically improved since I started using them. I’d advise staying clear of using extremes of angle or spacing between the mics. As I said, level and timing differences have their pros and cons and you can particularly emphasize the cons by using one way more than the other. Stay away from the shaded areas, stick to a happy medium and you can’t go wrong!
If you want to check out Michael Williams’ work in full head to his website at http://www.mmad.info/MAD/2%20Ch/2ch.htm. There you’ll find more information about what he’s done and curves for different directivity patterns.