Mid-Side/Figure 8 Recording

  • Thread starter Thread starter buck78
  • Start date Start date
B

buck78

New member
After doing some reading on MS Stereo recording, I did some thinking (that can be dangerous).

I will be recording my father-in-law's bluegrass band in the near future. They are a four piece band (banjo, 2 guitars, bass). I am using a Boss BR-532 recorder which only allows 2-Track recording.

My inital concerns for recording were/are banjo and bass bleed. The bass can be fixed by over-dubbing that part after the fact. My fix for the banjo was/is to record that part on a separate track. To reduce the bleed into the other mic I figued (no pun) to set it to the figure 8 pattern (B3). By placing a make shift gobo between the mics and lining up the banjo in the figure 8 dead spot I should be able to reduce the bleed enough.

So both my guitarists are singing and playing into their respective sides of the figure 8 mic. Here's my question (finally!). If I re-record this track, while flipping the phase, will I be able to "separate" them into two tracks? I plan on doing this by running the signal from the recorder through my pre-amp (DMP-3) and flipping the phase reversal button.

If this doesn't work, I still feel that I will have a good recording. If it does work, I'll have a good recording and an even better mix. See, if I had a 16-track monster that could record 8 tracks at a time I would never learn this stuff.

Thanks for reading. Any input is welcome.

Buck
 
Sounds like it would work in theory. You may be able to get the separation with panning the 2 tracks though. Since it's your father-in-law's band, won't that allow you some time to experiment?

Is the bass electric or stand-up? If electric, why not record that DI and not use a mic? Then you have no bleed potential.

I don't really know. I'm just throwing out some ideas.:p
 
Banjo/Singing on Mic 1 (B-1 carodid)
Guitars 1 & 2/Singing (B-3 figure 8)
Bass DI

I'm attaching a picture to give everyone a better idea.
 

Attachments

  • mic-layout.webp
    mic-layout.webp
    11.8 KB · Views: 85
Sounds to me like you've got a pretty thoughtful and clever solution to your problem, there.

Whadya want from us, a medal or somethin' ? :D :D :D
 
I am assuming that this means you think this will actually work :D

That's what I was going for. I just wanted to make sure I was going down the right path. This will be my first "outside" (not myself screwing around) recording project. It's also a way to justify getting both the B-1 and the B-3!

The session date keeps getting pushed around. Hunting season and the holidays are currently winning the schedule battle. But I'll keep everyone posted on how things work out.

buck
 
This seems like a useful set up generally.

You will, however, not be able to separate the two guitarists. They're recorded on one track. There's no way to separate what came from the back lobe from what came from the front lobe.

The two guitar parts will be together on one track -- same as if they were sitting really, really close together.

I would probably not matrix the cardioid and figure-8 mics in the mix like you do when you use a M-S technique. Just for reference, if you did want to do a mid-side matrix at mixdown, you could do it by panning the cardioid (banjo) track to the center, panning the figure-8 track hard one way, and panning a polarity-inverted copy of the figure-8 track hard the other way. The splitting and panning of the figure-8 is really intended for a situation in which the mic is picking up the same sound, just coming off a different wall or whatever. The worst thing is that the guitars would vanish entirely if you summed to mono.
 
For MS to work correctly, I think the two mics would have to be coincident. You might also want to check-out the "3 to 1" rule - (see Harvey's Mega-Mic-Thread).

Hope this helps -

- Wil
 
I'd probably go for one of two approaches:

The "no bleed" approach

Record only two parts at a time and overdub. To get feel and timing right, you might have the others playing while you record the first two tracks, just keep them isolated. Bass would probably be a good one to get in your two "basic" tracks, as it's usually the rhythmic foundation (though that depends on the group). Maybe bass and whichever guitar is the more rhythmical "solid" (so to speak). Then overdub everything else. I would probably record at least seven tracks: bass, guitar 1, guitar 2, banjo, singer 1, singer 2, singer 3. If you really want to minimize bleed (and maximize mix-down flexibility), I'd overdub vocals separate from instruments. If they need to sing while playing, you can use two mics, get 'em as close as possible and try separating with a horizontal divider (like a music stand tipped back).

The "screw the bleed" approach

Record the whole group direct to two-track stereo. You can do this with a single pair of mics backed away from the entire group in a stereo set up (could be MS, or X-Y, or whatever works) ... or you could close mic everything with as many mics as your mixer has inputs and mix them the stereo on the fly.
 
Just to be clear...

I do not want to matrix the banjo mic at all. What I want to do is use the the portion of MS theory that separates what comes in on each lobe of the figure 8 mic.

My understanding of this is the following. The signal from the figure 8 is multed and the phase is flipped on one of the signals. These are panned hard left and hard right. The carodid mic is run straight up the middle. Adjust levels to suit. This is my understanding of MS Matrixing.

I only want to do the phase part on the figure 8. Mic placement is outlined on the diagram above. Will this work?

Again the reason for doing all of this is two fold. First, my recorder is limited to recording two tracks at a time and can play back four. Second, the banjo is MUCH louder than the other instruments and singers. The bass will be overdubbed after the inital recording of the song.

buck
 
The "matrix" part is really the whole flipping polarity and panning that you do with the S (figure 8) mic.

The key thing to take away from this is: it doesn't separate what's on the back lobe from what's on the front lobe!

The two guitars will be on one track ... same as if you set them next to each other and pointed a cardioid at both of them. Same (perhaps more easy to picture) as if they gathered around an omni mic. The only difference is that the guy who's playing and singing into the back of the figure 8 mic will get his signal inverted. But this doesn't really make a lot of difference, since the signals coming into the two sides of the mic are different anyway.*

As I noted before, the MS "matrixed" approach is really intended for stereo micing -- where you're recording what's effectively one sound source: one source, but with some lateral spread to it. It's mono-compatible (in a sense), because when you sum to mono, you lose both of the "S" sides, but you still have the "M." If there are whole instruments on the two mics, you would have quite another, and really pretty disastrous, result.


*Yes, there is such a thing as "absolute" polarity. Unlike a sine wave, your typical percussive sound is not symetrical, and flipping its polarity will have an effect. Most people don't notice inverted polarity at all ... some think it's obvious (these people become golden-eared mastering engineers or symphony conductors) ... depending on the sound, it may be perceivable to a normal person if you listen to it both ways in rapid succession. A significant number of home speakers are probably wired so as to invert polarity anyway -- so long as both speakers are wired the same way, nobody notices.
 
So where does the "awesome" MS stereo spread come from then? I guess I am not getting it yet.
 
That article set me straight.

As for the bass player going first. I don't think he's that good, to go it alone first :)

Thanks for bearing with me everyone.

buck
 
You must log in or register to reply here.
Back
Top