"MS" technique?

[Markd102]

What is "MS" technique?
My apologies if this is a stupid question. I keep hearing about it, but I haven't a clue what it is. Could somebody please take the time to explain it to me?
Thanks
Mark

 

[skippy]

MS stands for "mid-side". It's a stereo miking technique developed by the Germans (Lauridsen) in the '40s that offers some advantages over other common stereo techniques. Two examples of other common techniques are X-Y, using a pair of cardioid mics mounted at 90deg on a single stand, with the capsules as close as possible ("coincident"). Another is ORTF, which uses a pair of cardioids spaced about 10" apart and set up with a 110deg angle between them ("semi-coincident", thanks to the French). These give you right and left signals directly, and your soundstage is pretty much dependent on how well the mics are matched.

M-S uses a coincident pair of mics, but they are completely different mics. The "mid" mic is a cardiod, pointed straight up the middle. The "side" mic is a figure-8, pointed perpendicular to the mid mic. The outputs of these guys are then fed into a matrixing circuit that produces left (mid+side) and right (mid-side).

The thing that makes it work is that the two lobes of a figure-8 are out of phase with each other. One big advantage it offers is the control of the soundstage, by varying the relative gains of the mid and side preamps: you can change the stereo spread from barely-stereo to holyshit-stereo with a gain knob. Like any coincident-stereo technique, it is also mono-compatible...

It's just another technique that is sometimes very useful- like recording barbershop quartets, for example. Mathematically, it would seem that there's not much difference between the result of MS and the result of X-Y. But the sonic possibilities are limited only by the characteristics of the mics you choose, and *that* can be a powerful tool...

 

[Gidge]

either skippy:

a) made that up

b) is suffering from vertigo due to the atmospheric pressure caused by Colorado's altitude

c) gave an excellent explaination of an interesting technique

 

[alanhyatt]

Actually, Skippy is quite correct. While the explanation may sound technical, it is a simple thing to use, as long as you have the proper equipment. Most Americans are not much into the use of M/S Recording. The technique is more used in Europe.

 

[skippy]

Hey, I'm basically a nerd. And I love M+S for a lot of the stuff I do: pick the right mics and the right room and the right talent and the right day, and you can do some serious damage.

But ultimately: ask a nerd question, and I'll give you a nerd answer. Cain't help it...

However, there's much to be said for the interaction of a good Pinot Noir and any over-6000-ft-above-mean-sea-level elevation of the studio. So maybe all of Gidge's alternatives are equally true!

 

[Track Rat]

Even though I'm in St. Louis, many times this is the highest place in the known cosmos.
Pfffftttttt......I think my karma just backed over my dogma....

 

[Tekker]

How exactly does the matrixing circuit work? What would be the difference between the matrixing unit and copying the figure 8 mic and reversing the phase on one of them?

-tkr

 

[skippy]

There really isn't any difference at all- that's essentially all the matrix circuitry does, anyway.

Of course, if your board has balanced direct outs and/or channel phase switches, M+S is trivial: you can do a poor man's M+S matrix right in the board. Pan the mid mic to the center of a pair, pan the side mic hard left on the pair, and (if you don't have phase switches) make a phase-swap patch cable: just swap tip and ring in the patch cable you run from the side mic's direct out into the (balanced) line in of another channel. Pan this inverted side channel hard right on the pair, and turn the summing buses for the pair into your matrix. The only hard part is making sure that the levels of the two side mic channels are equal, so that your imaging doesn't become lopsided.

In fact, there are a number of inexpensive mixers that have their channel insert outs inverted in phase with respect to the mic input and summing bus (older Studiomaster boards come to mind). If you have one of those boards, you can use an insert steal to get a single-ended, but inverted, version of the side mic for your poor man's matrix. Lather, rinse, repeat.

Viola', instant M+S, absolutely for free- with no outlay other than for the mics. Takes a bit of tweeking in some cases, but works very well... It pays to find out the relative phase of the various outputs of your board with respect to the mic ins: both so that you can do tricks like this, and so that you can avoid phase-cancellation headaches when you _don't _ want an inversion in the path (a much more common annoyance with entry-level gear).

And so on... There's a million ways to skin that cat, and they don't all have to cost money.

 

[Gidge]

that elevation had my head alittle cloudy the last time I was in Colorado so Im going to answer a and c to my previous question.....

 

[alanhyatt]

Skippy,

I agree with you on the "poor mans matrix", but there are too many variables and issues with older consoles with regard to which pin is hot. This will tear the interior of many a brain cell, as I have run into it quite a bit, especially on British and Japanese consoles, but yes it can be done, but you have to really know what to look for.

An easy way out is to find a mic pre with decoders on them, so the decoded signal just plugs into line level inputs. We are building an 8 channel mic pre right now with the first two channels offering decoders for M&S matrix, and left/right summing busses...and of course lots of other features needed to make it a good unit. Target price is $799.99, available June 2001.

 

[skippy]

Sure: you have to have a bit of technical wherewithal to find out what the phase relationships are. However, if you do develop those skills, you can save yourself some grief, and also be able to do "stupid mixer tricks" like this to amaze and amuse your friends... (;-)

It's not always easy: as you point out, with complex pro gear sometimes it can be mind-numbingly hard, unless you have access to, or can borrow, a 2-channel oscilloscope for a weekend. But just the same, if by some mischance a few readers here *are* technically inclined, it can certainly pay dividends for them to get that familiar with the behavior of their own (usually simpler)equipment. I always like to encourage folks to improve their techie chops: it makes me feel less lonely over here in the nerd camp, especially since I'm really just a drummer (;-).

The easiest way to do M+S is to buy hardware with the M+S matrix built in, to be sure. But there are people who frequent this site who are either so poor (or so _tight_!) that they try to reuse *solder*... For them, the alternative of hacking around to do the poor-man's tricks is a way of life. And some just like the challenge as part of their home recording/hobby experience. That's just a part of life hereabouts...

So many folks have talked about using insert steals as cheater direct outs, in threads here over the last couple of years, that I thought it'd be good to bring up that cautionary tale. Easy way to tell if your insert sends are out of phase with respect to the channel inputs? Patch a signal to the line in on a channel, and route it to a bus. Patch its insert steal to another channel's line in, and put it on the *same bus*. Bring up the first channel fader, leaving the second channel down. Then, bring up the second. If the signal from the bus gets _quieter_ as you bring up the second, the send is out of phase (the in-phase and out-of-phase will cancel each other as the second fader comes up). For the purposes of playing this particular poor man's game, this can tell you a lot- and no test equipment is needed, other than the Mark I Earbone.

June, 2001 for that preamp, eh? Is it out, or is that a typo? And if it's not out, do you need beta-testers? I've got a live recording date coming up Dec. 1 that just cries out for some M+S work, and the Truth 8-ch pre with M+S was too rich for my blood...

I've got a little bit larger gear budget than many folks here, but the accountant in me rears his ugly head from time to time just the same!

 

[Tekker]

Hmmmm, do you think I can do all that with my Behringer 1602a?

I was thinking about doing it with a DAW, just for recording. It's nice to know that I don't have to go out and get anymore equipment.......oh wait I do, I still need a figure 8 mic, dang.

-tkr

 

[PinkStrat]

MS Stereo mic'ing technique question

Hi Skippy,

I really appreciate the info on the M+S mic'ing technique. I have a pretty nice variety of microphones in the old mic closet and I have a simple question? Can I pull off the M+S technique by simply using two separate mic pres [that have phase reversal switches on both]? Is this just a matter of setting the side figure eight mic out of phase with the mid microphone? I have yet to try this on acoustic guitar! I have used simple stereo mic'ing in the past but I would love to try this MS technique now. Thanks for any helpful information in advance!

 

[tubedude]

I like M-S.
The cheapest way I know of to accomplish it is by way of Waves S1 plug-in. Not only does it work great, but its phase compensated and it sounds really fat.
The Amek system 9098 has M-S decoding along with two Neve preamps... theres a nice deal...
Enjoy!

 

[skippy]

Pinkstrat: the problem is that you need *both* in-phase and out-of-phase versions of the side mic's signal. If you have separate pres with balanced outs, this is easy: mult the output of the side mic pre into the line ins on two channels, but use a cable that swaps tip and ring for *one* of them- giving you a channel with +side (in phase) and a channel with -side (out of phase), in addition to your channel with mid.

Then the phase switch on the side mic pre will reverse the stereo image. More ways to amaze your friends...

Bottom line is that the left signal comes from (mid mic + side mic), and the right signal comes from (mid mic - side mic). You have to have both sexes of the side mic signal, and they have to be matched in level and exactly 180deg out of phase with each other. If the levels of the +side and -side channels aren't matched, or if you have eq on one but not the other, expect the stereo image to get severely whacked. Most eqs insert some phase anomalies, which kills the cancellation of the lobes of the figure 8 that make all this magic work. You should be able to sum to mono, pull down the mid mic fader, and have exactly _0_ signal left from the sum of +side and -side (that will be all that's left). If there's anything left over, *that's* what's whacking your stereo image. Trim the side channel gains until it goes away.

Of course, if the sound of the severely whacked stereo image pleases you, feel free to use it anyway! I've been known to do that occasionally...

Tube, I haven't played with the Waves plugin- but that's because I'm still an analog, hardware-based guy at heart. I've been seriously considering expanding my plugin armory with some of the Waves stuff, now that my DAW is running well. It's good to know that feature is out there!

 

[PinkStrat]

Thanks for all of your input on the M/S technique, Skippy! I am having an audio maven friend of mine make up a couple of high quality balanced female XLR to 1/4 inch tip/ring adaptors as we speak. Then, I will be able to do two M/S techniques; the first one [per your suggestion] using the single figure eight mic as the side mic with a cardioid as the mid mic. And secondly, [and even better yet perhaps?] using TWO figure-eight mics together...I cannot wait to hear the results. I am all-too curious to hear how "holy-shit stereo" I can get by cranking that gain control on my mic preamp. Thanks again for all of the audio enlightenment! I'll let you know how it all turns out.

 

[Wil Davis]

...And secondly, [and even better yet perhaps?] using TWO figure-eight mics together...

Turn them so that their axes are 45 degrees to the source, and you have a standard Blumlein coincident pair. (see also Harvey Gerst's "Great Microphonium Thread" :> )

- Wil

 

[skippy]

Two figure 8s? Are you trying to do the old EMI style stereo (Blumlein), which is basically XY with figure 8s?

That's fun to play with, but only in a _really_ good room: the stuff coming off the back wall is there in all its glory, in the (out of phase) rear lobes... But in the right room, it's a great technique as well. It's a great way to get all the audience noise for classical recordings, since there's no backside rejection at all. But let 'er rip, and see how it sounds!

The Blumlein stereo technique doesn't *need* a matrix as such. It was really popular in the early days of stereo, because you can matrix the signal to get as much vertical or lateral component as you want at the cutterhead of a lathe: this goes back to the days when "direct-to-disk" meant you were cutting lacquer with the outputs of the board....

Here's some more fun reading, if anybody still cares. http://www.dpamicrophones.com/index.htm

Hey, Wil! You beat me to the punch- I type too slow...

 

[Wil Davis]

Oh, by the way - just a thought...

The way I do MS is with a pair of GT MD 6TMs, one set to fig-8, and the other set to cardioid. I have a splitter-cable which goes from XLR-F to two XLR-Ms, one of which is reversed in phase. The three signals are then recorded on three tracks, and then when I come to mix, I set one side of the fig-8 panned hard left, the other (reversed) side panned hard right, and the track from the cardioid panned dead centre. The two mics I use each has its own power-supply, so I've never needed to use phantom power with this set up. I'm wondering if I used my splitter-cable with another mic which did require phantom power, if there might be a problem... just a thought!

- Wil

 

[Wil Davis]

Sorry Skippy, I didn't mean to steal your thunder!

I was totally blown away when I first came across MS - it is a really simple technique, and so rewarding (for such little effort), and has many advantages! If I'm recording the room (i.e. if the acoustics are fantastic) then MS is my first choice!

- Wil