SouthSIDE Glen said:
An example of what I was talking about would be stereo overhead mics over a drum set. By the 3:1 rule, the overheads should be a minimum of three times the distance from each other as they are from the top of the drum kit.
"Coinicdent" mics are two mics that are setup to record in stereo where the mics are pointed angled away from each other with their pickups either right on top of each other (what's called in the States "X/Y miking".) There are other variations on such miking techniques, a Google search on coincident microphones should give you plenty of detail.
But to get back to 3:1, the idea is that the mics should be farther away from each other than they are the source, regardless of whether its one source or more.
G.
So if my drum OH mics are three feet over my kit, they should be nine feet apart? You may want to rethink that, no offense intended, you always post great stuff. That would be hard, at the very least, if your OHs are more than a foot or so up. At that point, some kind of spaced or coincident pair technique is probably a bit easier, and what seems to happen a lot for true stereo overheads. 3:1 doesn't really give full true stereo, as I see it. It really seems to be meant to keep phase issues down by controlling bleed, and therefore makes a poor stereo mic'ing choice, compared to others. It seems some center material is always going to be lost.
Jakob,
Some good info here, which I feel like expounding on. It is late, and I am bored. Your instinct about number one in your post is correct, and here is why.
The 3:1 rule covers a lot of situations, and is a guideline to prevent phase anomalies. It is very gain, source, and placement dependent. It applies to some semi-stereo techniques, close-mic'ing of multiple sources to prevent phase problems from bleed, reflections, and probably others I have forgotten.
The basic idea is this:
Phase problems happen when two out-of-phase signals get heard together, at similar levels. Some freqs reinforce, some cancel. A propellorhead a long time ago figured out that if two out-of-phase signals are heard together, you can't hear the phase cancellation if one of the signals is at least 9db quieter than the other. You can't hear the quieter signal very much at all, really, but that's another thing.
Two similar sources in a room (say two strummed acoustic guitars) are each miced at the same distance (say one foot) by identical mics with identical gain.
If the mics (and guitars) are at least three feet apart (3:1), in each mic the close source (the close guitar) will be at least 9db louder than the far guitar. You can play both tracks back together, no phase problems. The bleed between mics is low enough in level it causes no problems.
So this is part of how the general 3:1 rule came about. All things being equal, sources, mics, gain, etc., if the far source is at least three times the distance of the near source, the far source will be at least 9db quieter in the mics than the near source. You can play both tracks, no problems. I don't know if it was originally designed and researched as a stereo technique, but as a stereo technique it has been replaced by many vastly superior ones. 3:1 has lots of uses, though.
Hopefully you can see that a strict 3:1 rule won't do shit if you are mic'ing a hammered dulcimer from one foot, a raging Marshall stack from one foot, and they are three feet apart. If you play both those tracks back at the same level, the Marshall will be very phase-affected, and the dulcimer track will be useless.
It's about relative levels of direct and far sounds in the same mic during recording and relative levels of out of phase signals in different channels on playback, not an absolute 3:1 rule. You observe 3:1 to give you usable tracks, able to be mixed at the right levels with no phase problems.
If two mics are on the same source, at different distances, if you listen to both at the same volume, there will be phase anomalies. This is where the idea of moving the far amp mic around to find out where it sounds the best comes in, as well as Snow Lizard's pink noise technique. You are trying to minimize the effects, you can't kill them completely. That is why 3:1 doesn't kill phase problems when you use two mics on one amp from different distances. 3:1 isn't meant to address the phase issues from that. So yes, number two is wrong.
Never mind the physical impossibility you pointed out, of having two mics be farther apart than their combined distance to the source. I laughed when I read that, I got my ass chewed by my mentor when I pointed that same thing out. It led me to study the subject in a lot of detail.
In these digital days, you have a bit more flexibility. Say you find a great room mic location, but it causes massive phase problems. You can nudge the track until it doesn't cause as many problems, instead of moving the mic and losing your sound.
The concept behind the 3:1 rule also helps explain phase problems caused by reflections. Picture a speaker talking into a podium mic across a hard desk. His voice reflects off the desk, and the reflection arrives at the mic shortly after his voice. This can cause phase cancellation if his voice is less than 9db louder than the reflection at the mic, as it violates the concept behind the 3:1 rule. The cures?
Move the mic close enough to make the speaker's voice more than 9db louder than the reflection, or damp the reflection so it is more than 9db quieter than his voice.
3:1 in Stereo Mic'ing:
First of all, a little basic background on how stereo mic techniques and human hearing work.
Human directional location works on two principles. If a signal is louder in one ear than the other, our brain thinks it comes from the louder side. Our heads are kind of sound baffles, and our ears are fairly directional. Also, our head acts as a delay between our ears. A sound of equal intensity in both ears but delayed slightly on one side will seem to come from the non-delayed side. Our brains use amplitude and delay to provide very accurate spatial location of sound sources.
Spaced pairs use delay to generate the stereo image. Spacing the mics introduces the delay for the stereo recording. A signal on the right will hit the right mic first, then the left.
Coincident pairs work by the fact that directional (cardioid, mostly) mics are used, but facing different directions. A sound coming from the far right will enter one of the mics where it is very sensitive. The other mic picks up the sound at the same time, but reduced in level, as the mic is not facing the sound. When you listen to both mics, the postion in the stereo field is determined by the relative volume of the source in both channels. Something appearing to be in the center is actually in both mics at the same volume. This is how the pan pot on a mixer works, by varying the levels of a channel in both ears at once.
3:1 is sometimes used in close stereo mic'ing. In the case of a piano, say, with one mic covering the low strings and one covering the highs. 3:1 is a good place to start to keep the bleed and therefore phase problems down.
The key thing to notice is that left and right are kept as isolated as possible, to reduce phase problems. Essentially each half of the piano is treated as a separate source. So, two sources, two mics, 3:1, no problems. Not real stereo, either, though, as I see it. Because as you might guess, this leaves a hole in the middle. It's the same with drum overheads. A pair of mics one foot up and three apart is going to miss a lot if you really want to avoid phase from bleeding.
So you gotta toss 3:1 and go with your ears.
Any spaced pair is either going to follow 3:1, and not capture the whole image, or needs to be placed so phase problems are minimized.
Spaced and coincident pairs have their own problems, and solutions, but that is another thread.
Anyway, that's my 3.14159265 cents
First of all, you read and quoted my reply before I deleted it because I explained it badly or wrongly. "Drugs" was my lame explanation for deleting it. Sorry 
