Mic placement and phase cancellation

Is three mics better than two?

I recently read an article on recording grand pianos that recommended the positioning of three microphones above the soundboard of the piano. Why wouldn’t such a configuration lead to problems with phase cancellation when the signals from the different mics are mixed or played together? Or is three mics better than two in part because it helps deal with phase cancellation?

The wave length for, say, a 1000 Hz tone is about 13.5 inches. If the signal from two mics positioned at one and one and one half wave-lengths from the sound source (13.5 amd 13.5 + 6.7 inches) is mixed, phase cancellation should severely attenuate the sound. If there were three mics positioned around a sound source (of whatever frequency), it seems unlikely that they could be positioned so that there could be extensive phase cancellation. Two could be receiving signals that were 180 degrees out of phase, but the third mic ought to pass its signal with little cancellation. Am I right in my thinking about this?

I guess that another way to avoid phase cancellation would be to use the x/y configuration, with mic capsules close together, that I have seen others use. No phase cancellation at audible frequencies. Sure took a long time to sink in why people might prefer the x/y configuration to the fairly wide spacing that I have typically used.

T

Quote (tspringer @ May 27 2006,18:20)
The wave length for, say, a 1000 Hz tone is about 13.5 inches. If the signal from two mics positioned at one and one and one half wave-lengths from the sound source (13.5 amd 13.5 + 6.7 inches) is mixed, phase cancellation should severely attenuate the sound. If there were three mics positioned around a sound source (of whatever frequency), it seems unlikely that they could be positioned so that there could be extensive phase cancellation. Two could be receiving signals that were 180 degrees out of phase, but the third mic ought to pass its signal with little cancellation. Am I right in my thinking about this?

I guess that another way to avoid phase cancellation would be to use the x/y configuration, with mic capsules close together, that I have seen others use. No phase cancellation at audible frequencies. Sure took a long time to sink in why people might prefer the x/y configuration to the fairly wide spacing that I have typically used.

T

Eyup!

Yes and no. Your post assumes that there will be no interference from sound reflected from other surfaces.
These may be an odd multiple of a half wavelength away, more than likely they will not.
Don't forget also, that for complete cancellation to take place, not only must the wave be in anti phase, but also of equal amplitude and the same shape.

I think this is why "listen for the sweet spot" is good advice!

Steve

They are referring to a variation on what is known as the Decca Tree stereo technique. Hit Google and you’ll get more insight on why this works.

Thanks guys. I’ve learned that I can always count on you for insight.

T

OK. I went off and did my homework. And after a re-read of the article, the focus of the question has shifted some.

I didn’t give a detailed enough description of the three mic setup. It is not a Decca tree. The Decca tree consists of three mics, arranged in a triangle, that are in front of a sound source. The apex of the triangle is nearest the sound source. The microphone arrangement that stimulated my question is ABOVE a grand piano. It is the configuration used by Terry Howard to record much of Ray Charles piano work.

One mic is placed facing down over the bass strings, a foot or so back from the hammers. The second mic is positioned similarly over the treble strings, but is oriented off-axis (away from the hammers) to reject hammer noise. The third (omnidirectional) mic is positioned near the back of the piano and centered between the other two mics. The main purpose of using three mics is to avoid an extreme left-right sound. But here is the part that I’m trying to understand now. Howard went on to say that his rule of thumb for avoiding phase problems was to keep all mics at least 9 inches from the strings and from each other, and that further apart is better. Why is 9 the magic number? And why is further apart better? I want to understand this because I’m sure that the principles will apply to lots of other recording situations.

T

Further is better because the difference in level will be greater which reduces the chance of phase cancellation. Where the magic number 9 comes from I don’t know. I also do not know how much the sounding board resembles a plane-wave source or how localized the radiation from each string is on the sound-board (or if direct radiation is significant). These will effect both the phase and magnitude of the signals. Without a bit of complicated research you will just have to try and see. If you do happen map out the detailed phase and magnitude radiation patterns of the piano please post it:)

Jim

I think Jimbob’s got it. A piano is a very complex instrument in theory and practice. My guess is that the magic 9" is heuristic – based on practice, it works & screw the theory. I would suggest trying different distances; some folks find that being this close causes too much mechanical noise (even when angled away from the hammers).

One key to recording a grand piano is to do it in a big room with a very high ceiling. If you don’t have that, then do anything possible to deaden the room and add ambience later using a plugin like SIR.

Thanks again guys. Your insights are thoroughly appreciated.

T

I knew you were referring to that article - I too read it - my thought is that the method was intended to capture a certain sound, not the “natural” sound of a piano from far away, but that special sound you get with close micing - an aesthetic that developed from the need to close mic the piano for recording when the whole band was playing, I would guess. They would have put a bit of reverb on it afterwards. The placement described in the article generally minimizes phase problems, and whatever phase problems are there, are part of the “sound.”

Last band I recorded I finally did an X/Y on the drums and they sounded amazing. I’ve been doing a wide spread and have had lots of trouble with the toms cutting through at all and everything sounding slightly blurry. When I did the x/y the toms came alive and the whole kit was nice and clean.

I wish I was positive it was solely from the micing but there were a few other contributing factors…

the other bands I’ve recorded had crappy drums, didn’t know how to tune, and we recorded in a basement. this last group was a Christain group so I recorded it in their sanctuary. It was quite large, had a center stage that was basically a large riser, the acoustics were neutral, no big echoes or anything, the closest wall was at least 30 feet away and the kit itself wasn’t too bad.

Next band I’ll do the x/y again and we’ll see if that really does help a bunch.

Guitar69

When you get around to using the xy configuration on drums again, do post what you find. I’m curious.

T

Live recording (microphone placement) techniques are near and dear to my heart. I have recorded literally hundreds of live concerts for various orchestras, choirs, bands and small groups. Every room is different and even in the same room, different groups require different micing techniques.

For many years I prefered the spaced microphone (A/B) setup. I would space the microphones to be at thirds to the width of the group being recorded. For instance, if the orchestra was 30 feet wide, I would place the microphones at 10 feet in from each side pointing directly at the orchestra. Also, the microphones would be placed about 20 feet in front of the orchestra. Of course the distance from the orchestra would change depending on the room acoustics and how lively I wanted the recording to be. Also, I always used moderately wide cardioid microphones…rarely did I use omnis (too much audience noise).

This always produced a wonderful stereo image and a balanced soundfield. Also, many of my recordings were played on the radio and I never once had a problem with audible phase cancellation.

As years went on, I experimented with X/Y microphone placement and found that I really did not like the stereo image I got. But as I did more and more recordings that were being played on the radio, I felt like I needed to “always be sure” that there would not be a problem with phase cancellation, so I stuck with the X/Y setups. Occasionally I would experiment with near-coincident micing, but really did not like the results.

Anyway, when I am recording for me, I always use a spaced A/B setup. When it is for the radio, then I always use coincident X/Y mic placement.

I would not be concerned too much with phase cancellation. Your best bet is to place the microphones, pan them to center and listen to what they sound like to be sure there is no frequency cancellation. Once you are sure, then you can pan them as you like.