Hmmmm. This discussion about the 3:1 rule differs from my understanding. I thought the guideline applies to the difference between two mics being used on one source, not the distance of each mic from the sound source. For example, when micing a choir, if two mics are placed ~ 3 feet in front of the vocalists, they should be ~ 9 feet apart. Is this outdated? If so, there is a lot of misinformation on the internet because a quick google search pulls up numerous references consistent with what I described. And they state that the main objective is to minimize comb filtering. So it's not simply a matter of placing the second mic farther from the source such that the recording level is diminished and thereby reduces phase problems. This explanation doesn't actually make sense to me, so I'm not defending it.
There's a muddle of different circumstances here.
Single source, two mics, combined sound.
Single source, two mics, stereo image.
Two sources, two mics, isolation required.
Traditionally the 3-1 rule applies where there are 2 sources, 2 mics and isolation is required.
Putting a 57 and a 421 on a cab is not the same. You've moved those mics until you like what you heard.
The 'problem' created by <3-1 with 2 mics 2 sources is the goal with a cab and 2 mics!
With a choir you're dealing with stereo image.
If you take the very first image on the page, those two loudspeakers could represent altos and sopranos.
Depending on the size of the choir, you could get 3-1 on the extreme l+r of the choir, 1-1 on the people bang in the middle, and the rest is the best of a bad job.
If you're trying to create a stereo image, that's not a bad setup, right?
Idk, the choir example you gave is probably a tried a tested happy medium or whatever, but if someone said "tell me about 3-1 rule", I wouldn't start talking about choirs.
Thinking about it, I observe 3-1 rule when stereo micing an acoustic.
Not because I read it somewhere and think it's a rule that I should follow in that context, but because that's what I've found to sound the best through trial and error.
Basically it happens to fit, kinda, but it's not what the rule is for.