MS stereo confusion

[simpleybass]

I've been thinking about the MS stereo set up suggested in another forum https://homerecording.com/bbs/showthread.php?t=179999&page=2, and I'm confused about the principle, so I decided start a thread focused on this topic. I've read several web pages about XY and MS stereo mic setups, but I still don't get it.

When I mic XY, I have a left and right signal. That's obvious. So if the mandolin player is standing on the left, the resulting mix makes the listener hear him louder out of the left speaker (to a greater or lesser exetent depending how far its panned during mix-down). In my mind, that's the nature of "stereo".

With the MS signal, we'd have a center M mic (cardioid B1), and a side S mic (figure eight B3, with null facing center). This S signal is split with a Y cable, with one connector phase-reversed and panned left, and the other panned right. I think this means that if the S signals were both panned center, they would cancel each other out.

The part I can't understand is how you get any difference between the resulting left and right channels with this set up.

Maybe I don't understand how a figure eight pattern mic works. I THINK it mixes sound coming in from both sides of the diaphragm, because it is a mono mic, afterall. So if I stand facing one side of the diaphragm, and you stand facing the other side of the diaphragm, and we both sing, doesn't the mic mix our voices into a mono output? Logic follows that the Y cord would duplicate the signal, but both of our voices would be equal on both signals.

My brain malfunctions when it tries to understand how you get left vs right out of this setup. It seems to me that you would get a center vs. side, but they are mixed into a left and right speaker in a basic stereo set up, so isn't the resulting sound coming out of the left speaker the same as the right? That's mono, correct?

Please help me undertand this. Sorry if this is tedious. Thanks.

 

[Harvey Gerst]

Ok, this M-S stuff works by additive/subtractive matrixing. True figure 8 mics have the best off axis response of any pressure gradient design and a perfect null in the middle (at exactly right angles to the front of the mic. When you turn a figure 8 mic sideways to the sound, none of the sound from dead center is heard. Only from the left side, or the right side (but one side is out of phase, cuz it's coming off the back of the mic).

You add a cardioid pointed at the center of the music. and not only does it fill in the hole, but it is combined with the figure 8 to create sum and difference combinations (usually thru a matrixing box) to let you control the absolute width of the stereo image. You can dial in anything from a perfect mono signal to wide stereo, all with perfect phase coherency.

A more complete source for how it does this is available from my friend Wes Dooley at http://www.wesdooley.com/aea/library.html. He makes M-S matrix boxes and he has a complete article on how M-S stereo works on his web site. Wes is one of the leading authorities on M-S Stereo miking. (I can't afford any of his damn matrix boxes, but it's good reading to understand the principles.

 

[jkokura]

Have you tried it yet? I don't care about the science - I think it sounds cool! I don't wonder why corn flakes work for me, I just like eating them...in the same way, I don't wonder about how M-S works, I just use it to get that cool 'stereo' sound that I want...I've only used it on Acoustic, but I want to try using it on OH...

Jacob

 

[simpleybass]

hmmmm

Thanks Harvey. I don't totally get it, but I think it's stirring my brain up.

And jokura, I think you are right, I just need to try it.

I was able to find XLR Y cords, but none that said one side was phase-reversed, so I assume I'd have to make the modification. Maybe I'll try to make one. I guess I'll need a soldering gun, too. Somebody said switch wires 2 and 3 one side of the Y cable.

Hopefully I won't do it wrong and short out something...

 

[Harvey Gerst]

Lemme try this again:

Imagine a guitar far left, a mandolin far right, and a banjo in the middle. OK, the "M" mic is saying, "Wow, I can really hear the banjo great, but the other two instruments are a little bit lower in volume." (All the instruments sound positive to him.)

The "S" mic replies, "What banjo? All I hear is a positive guitar and a reversed polarity mandolin."

Why does the "S" mic know the mandolin is reverse polarity? Cuz the mandolin sucks (came in from the back side of mic), while the guitar blows (came in from the front side of mic).

So we take the "S" signal and split it to two channels, but we reverse the back side polarity, so that both the guitar and the mandolin both blow equally. All it really is is a negative image of the other side, but the polarity will make a difference, as you will soon see.

If we put just the two "S" channels up in volume, they should cancel out. Why?

When you reversed polarity, you wound up with "guitar +" and "mandolin -" on the first channel, and you wound up with "guitar -" and "mandolin +" on the second channel. Now, we add in the "M" mic (which is a big drunk that hears everything as a "+").

Still with me? Good.

So now, these two mics are listening to the band and the drunk "M" mic is going, "God, I love that banjo." The "S" mic channels are saying, "What banjo?" But somebody at least is hearing the damn banjo, even if it's just the drunk "M" mic.

The "M" mic also says, "Hey, I can hear a guitar, too". The "S" channels ask, "A guitar? Are you positive?" The stock "S" channel replies, Hey, "I'm positive there's a guitar I hear, too." The reversed "S" channel add in his now negative guitar vibes and the guitar appears on the left.

The "M" mic also says, "Hey, I can hear a mandolin, too". The "S" channels ask, "A mandolin? Are you positive?" The stock "S" channel replies, Hey, "I'm gonna be negative about this; there's no mandolin here." But the reverse polarized channel says, "Hey, I've got a positive mandolin over here", and the mandolin appears on the right.

And that's how M/S works.

 

[Gilliland]

All right, I'll take a stab at it. The simplest way for me to think of it is algebraically. Let's start with some basics. If we simply had an XY stereo pair, we'd get two signals - left and right. The two signals would be similar in some ways and different in others. We'll designate these two signals as L and R.

In an M/S setup, our center "Mid" mic is capturing a directional center signal, pure mono. It's the common part of the set of signals from the XY that we discussed above. Let's think of that as the sum of left plus right. However, since we're using only one mic to capture it, the resulting magnitude will be half that of the sum of the XY pair:

M = (L+R)/2

OK, the figure-8 "Side" mic has a null pointing toward that same center, so it is capturing NONE of the center signal. Instead, it is capturing the signals that are reaching it purely from the two sides. Let's think of that as the difference between left and right:

S = (L-R)/2

So now we've got two signals M and S. What happens if we take the M signal and add the S signal to it:

M + S = (L+R)/2+(L-R)/2 = 2L/2 = L

As you can see, the Rs cancel, leaving just the L signal.

Now what happens if we take the same M signal and add the inverse of the S signal. That's the same as subtracting:

M - S = (L+R)/2-(L-R)/2 = (L+R)/2+(R-L)/2 = 2R/2 = R

As you can see, this time the Ls cancel, leaving just the R signal.

I apologize to the math-impaired, but this is pretty basic arithmetic. In any event, it's the simplest way that I know of to explain how M/S works.

 

[simpleybass]

ok

I think I like the explaination that included a banjo and a drunk.

But seriously, you both helped me a lot. And Harvey's link led me to some fairly technical scientific articles that helped too.

Now the real question... would you prefer x/y or m/s stereo for recording a bluegrass group, and why? Anyone feel free to give their preference.

The reason I ask is... I'm wondering if the m/s setup will actually sound any better. It seems like a trick to simulate x/y. Those articles seemed to suggest that the main benefit of the m/s setup was for television where the ms stereo can be readily converted to mono.

Just a continuing converstion here. You guys are really helpful. Thanks.

 

[Harvey Gerst]

The biggest benefit is the ability to dial in any stereo spread you desire - after the fact. You only need to actually record two channels - the inverted second "S" channel can be done later. Mic placement is less critical, and you don't need absolutely matched mics; just a good figure 8 for the S, and a cardiod or omni for the M side.

 

[Gilliland]

Now the real question... would you prefer x/y or m/s stereo for recording a bluegrass group, and why? Anyone feel free to give their preference.

The reason I ask is... I'm wondering if the m/s setup will actually sound any better. It seems like a trick to simulate x/y. Those articles seemed to suggest that the main benefit of the m/s setup was for television where the ms stereo can be readily converted to mono.

If the microphones are "perfect", then M/S is exactly interchangeable with X/Y. The real advantage of M/S is that you can vary the width of the stereo separation by simply raising or lowering the "Side" fader. There's no particular benefit to it other than that and, of course, its inherent mono compatibility.

With respect to a bluegrass group, there are some groups that really like to do that retro "let's all play around one mic" thing that's become so popular lately. An M/S config gives them their center mono mic, yet lets you record in real stereo.

 

[kid klash]

And if all these explanations still leave you wondering, pick up the new (March 2006) issue of Electronic Musician... they cover (literally - it's on the front cover of the magazine) the topic quite well in their article named "A Practical Guide To Mid-Side Miking; Take Complete Control of Stereo Imaging".

Enjoy!

 

[soundchaser59]

I understand most of the blue collar science behind it, not like Harvey does, but enough to know that it should work.

And when I try it it works.

And when I compare it to takes using other methods, the M-S method ALWAYS sounds the best.

That's my story and I'm stickin to it......

 

[Harvey Gerst]

The trick is to realize that you're just pushing air around and then figure out how these damn mic thingies pick that up. The rest of it is all smoke and mirrors.

It can be explained with math, or you can use drunks staggering around a room, or "dueling banjo" references. Both can work well in visualizing what's going on. It's all about air molecules bumping into each other, and positioning the mic to pick up the desired bumping of those molecules.

 

[Robert D]

Harvey,
There's a group of angry women who want to talk to you. They call themselves MADMM (Mothers Against Drunk Mid Mics).

 

[kid klash]

And when I compare it to takes using other methods, the M-S method ALWAYS sounds the best.

I agree! I haven't looked back since I tried M-S on a drum kit ("overheads", which are actually positioned at ear-height, in front of the kit) many moons ago.