How are you creating your matrix? Are you doing it within and audio console? Are you doing it within software? Or, are you using an actual MS decoder?

If you are using an audio console, the Mid microphone be it an Omni or cardioid microphone, must go into a single input on your console, pan to center. Your figure 8 microphone must be split and fed into 2 separate inputs. 1 panned left in phase in the second input, panned right, phase inverted. The levels on the figure 8 microphone must be balanced so that if you pan both to center, it will cancel completely. Since it is the same signal with one channel out of phase to the other.

Next, take the figure 8 inputs and Pan them both back to left and right. Now with the figure 8 levels matched together, slowly bring up your Middle microphone panned to center and voilà! You can adjust the stereo width by slightly increasing or decreasing your center Middle microphone. If you remove the Middle microphone completely, you will hear your figure 8 microphone completely out of phase through both of your speakers. This is what makes this matrix work. Your Middle microphone should not disappear! But in mono that figure 8 microphone does disappear if the levels are balanced properly.

I know that MS should work regardless of Omni or cardioid microphone although the true definition of MS generally indicates a middle cardioid microphone.

If you are using software and losing the Middle microphone, reverse the channel's as what you're describing indicates that your Middle and figure 8 microphone are reversed.

I love MS miking!
MS Remy Ann David
Why do you think they call me "MS Remy Ann David"?
It's not because I'm beautiful.

Thanks Remy, but I may not have expressed myself clearly.

I use a Grace Design V3 which you can use as an MS decoder. A delicious bit of kit which always excites envy (carefully disguised as professional interest).

When I say that the omni element "disappears" when paired with a figure of 8 I mean this figuratively - according to the diagram I saw in Streicher's book, the end result of matrixing with an omni as the centre element (with 1:1 ratio of mid and figure of 8) appears to be two channels L and R in a figure of 8.

Rob

Regards,

Using an omni as M mic is perfectly good in many situations. The resulting pattern is like two cardoid mics back to back. This might look like a pattern 8, but the difference is that both lobes are positve, not one + and one - as in a figure 8 mic. It gives a true stereo effect as one "cardoid" one listens to the left and the other only to the right. In contrast a figure 8 listens both right and left at the same time.

A bit of curiosa: As I understand the first cardoid mics were made out of an omni capsule plus a figure 8 ribbon. Adding the two signals created a cardoid (half of the stereo pair we are making in M/S with omni). Now, theory and practice is not quite the same thing which means the lobes never are perfectly accoring to the theory.

I like the omni and 8 on a medium distance in good acoustics and when close micing. As always omni-s has no proximity effect and goes an octave or so lower in the bass. My ears generally preferr omni-s to cardoids in most situations, but then my ear does not always like the rooms they are in.

Gunnar

Firstly, Happy New Year. I hope everyone had a good and peaceful Christmas and that the coming year is healthy and productive for all :)

Gunnar makes a good point. The matrixed pattern is akin to cardioids rather than fig-8s but they don't sound like normal crossed/back to back cardioids.

As a VERY rough rule of thumb, the matrixed "pair" of mics you get from an MS pair has polar patterns which are a step or two (depending upon how you define your polar pattern steps) tighter than the selected M mic. i.e (and again this is VERY rough):

Omni M = Cardioids
Cardioid M = Hypercardioids
Hypercardioid M = Supercardioids (assuming you use "Supercardioid" to describe a pattern similar to Hypercardioid but tighter - some people use the terms the other way round)
Supercardioid M = er...very tight super cardioids with somewhat enlarged rear lobes
etc..

The odd one out is when you use a Fig-8 as the M mic. Then matrixing gives you a pair of Fig-8s.

The polar patterns still tighten with frequency but, because one of the mics faces directly forward, the behaviour of the pair simulated by matrixing MS is different from using the complimentary pair of mics. They would have, to a varying extent, their off axis response characteristic facing the centre stage whereas an MS rig captures the centre stage on axis. An MS rig with fig-8 M mic matrixed to give a crossed pair of fig-8s sounds quite different from a conventional Blumlein pair. Also, the absolute accuracy of the MS pair is heavily dependent upon the type of Fig-8 S mic used.

Conventional switchable pattern mics using multiple diphragms to achieve Fig-8 don't give a true response, neither do single diaphragm switchable/fixed Fig-8 condensors which use acoustic interference to switch/define patterns (e.g. Schoeps MK6/MK8 or Shure KSM). Ignoring the Soundfield system, the only way that I know of to get a truly symmetrical fig-8 response is to use a pure pressure gradient capsule such as a ribbon mic or, in condensor mics, a Sennheiser MKH 30 (with it's ingenious true single diaphragm capsule). (EDIT - I changed the order of the last sentence as it was even messier than my usual gibberish!)

For many years (though I've moved on from it in the last ten years or so), my preferred rig for just about anything I recorded was an MS pair combined with a sort of variable size, loose Decca Tree - probably more of a pair of outriggers and a centre mono omni, often placed so as to be de-correlated from the other arrays. My own preference for the MS was (and still is) normally for a Schoeps MK6 (or MK8) as I've become used to, and now quite like, the slightly uneven frequency response between the sides and frequently use it to fiddle the balance a little. In practice, I find that ignoring the theory and quest for mathematical perfection is fine as long as the ends justify the means. I have no problem with using an AKG C426 or a pair of Brauner VM1s in MS regardless of the theory as I like how it sounds and anyway, it's not much worse in theory than the all but impossible to fix alignment delays beweeen the M and S capsules :wink: I usually find myself preferring a cardioid mid mic over an omni as I find that it gives a stereo image and more obvious width control that I prefer. The downside is that you lose LF extension (unless you're cheating :wink: and using a Sennheiser MKH40 cardioid, with it's pre-EQ'ed LF response). Nowadays, if I'm using an MS main pair, I normally use omni (or subcardioid) outriggers of some sort as well so I get some LF back from those.

I don't know who else here has tried it, but if you have access to the mics and can make up some mounting hardware, you can have a lot of fun doing surround recordings with an M-S-M setup using a rearward facing M mic in addition to the front one. It's not the best/most accurate/most spectacular surround setup I've used but it's very compact for live work, only needs 3 tracks on the recorder and with the right source it can produce satisfying results.

Excellent points so far.

Something else you'll want to take into consideration is the actual, physical polar pattern of the various types of mics.

Dual-diaphragm fig-8 mics' patterns look more like:
OO
(work with me here...)
Single element pressure gradient systems' (ribbons) patterns will look more like:
(X)
(Again, work with me here...)
With the former, you have much smaller nulls. With the latter, you have a far more defined null pattern. The main reason (though there are more) has to do with the fact that, a dual diaphragm mic system is relying on the cancellation of similar signals at the sides which only happens directly perpendicular to the diaphragms and at the respective rear-lobes of each diaphragm. However, the pressure gradient mic doesn't generate its pattern by cancellation, rather by simple off-axis rejection. Since the width of the ribbon is significantly greater than its thickness, only vibrations coming in direct (varying degrees of directness) contact with the larger surface will cause a vibration of the element.

There are a couple other minor laws of physics at play here, but these are the essentials.

Also, bear in mind the frequency response of omni microphones. Most omnis are quite directional at higher frequencies. (And don't forget, it's the higher frequencies which help us determine the directivity within a space - barring the point of diminishing returns at near ultrasonic frequencies.) Therefore, the pattern will essentially morph with the frequency.

When in doubt, draw circles on a page representing the patterns of each mic. Then indicate the phase relationship of each of the patterns. Anytime you have a + and a - overlapping, shade that a darker color to indicate its cancellation. You will very quickly see what patterns remain and thus the proper orientation of the pattern.

Enjoy!

J.

Forgot to mention earlier, there's a couple of entertaining little bits of software downloadable free from the Soundfield website which create very simple visualisations of the effects of various adjustments in MS (and some other techniques including the main Soundfield processes) http://www.soundfield.com/downloads/software.php

Many thanks. I need to read the posts thus far with some care. There is a lot of experience behind them. As I said, M/S is certainly not the most intuitive mode of acquiring stereo imaging. I will get my omnis soon and am really looking forward to using them.

Rob

Conventional switchable pattern mics using multiple diphragms to achieve Fig-8 don't give a true response, neither do single diaphragm switchable/fixed Fig-8 condensors which use acoustic interference to switch/define patterns (e.g. Schoeps MK6/MK8 or Shure KSM). Ignoring the Soundfield system, the only way that I know of to get a truly symmetrical fig-8 response is to use a pure pressure gradient capsule such as a ribbon mic or, in condensor mics, a Sennheiser MKH 30

The Neumann KM120 is a pure single diaphram symmetrical fig 8 and works wonders in an MS pair or as part of a blumlein pair. Don't let the Neumann bashers tell you this mic is bad, its a great fig 8, honey smooth and flat. The Schoeps MK8 does not use interferance like the MK6. The MK8 is also single diaphram, but its curves show it to be slightly asymmetric at 16kHz, otherwise its also a real beauty.

The published low freq rolloff in these mics is not evident when recording with them, a little EQ and you have a rich low freq response without the omni WOOF.

It is perfectly valid to use an omni for the center mic in a M/S pair. Doing so is a fairly obvious way of crowding more singers around a coincident pair when recording background vocals.

You can also use it for other things: rec.audio.pro's RAP CD in Blue set http://www.hoohahrecords.com/rap/ included a track on which I used this technique to record acoustic guitar. I used a DPA 4003 for M and an AKG 414 for S, and cranked the heck out of the side gain to get more ambience. I remember getting a lot of favorable feedback on it the time, but your mileage may vary.

The RAP CD in Blue was a very interesting compilation of different recording techniques, but unfortunately it has been sold out for years. If you really want to hear the track in question ("The Things I See"), I believe you can still buy the original artist's CD, "Songs from the Pit", here:

http://net.indra.com/~sqr/links.htm

Any M/S stereo array has an equivalent X/Y coincident array, though the latter may not actually be realizable with standard patterns. Sennheiser's Manfred Hibbing published a complete set of graphs which show the equivalent X/Y pairs that result as the relative M/S gain is changed. There are graphs for four different M patterns, if I recall.

AES members can download Hibbing's paper for $5 by going here. (It costs $20 for nonmembers.)

HERE

Here is a complete citation, in case you have a friendly librarian:

Hibbing, Manfred. Journal of the Audio Engineering Society, Volume 37 Number 10 pp. 823-831; October 1989

David L. Rick
Seventh String Recording

DavidSpearritt wrote: The Neumann KM120 is a pure single diaphram symmetrical fig 8 and works wonders in an MS pair or as part of a blumlein pair. Don't let the Neumann bashers tell you this mic is bad, its a great fig 8, honey smooth and flat.

David, thanks for pointing out the AK20. I use all the other KM100 series capsules but I'd disregarded the 20 as I felt that I didn't need another fig-8 and didn't expect it to offer anything over the Schoeps. Since reading your comment, above, I've spoken to the tech manager at Sennheiser UK (Neumann distributors in the UK) and, having checked with the mic's designer at Neumann, they've confirmed that it is indeed a single diaphragm completely symmetrical condensor capsule :) They even sent engineering drawings of the capsule to explain how it's done, and very neat it is! The drawings also help to explain the unusual polar pattern shown for 16KHz in the published documentation.

The Neumann engineer also explained the differences in LF response between the AK20 and Schoeps MK6/MK8. It seems that Schoeps use a slightly different method in presenting their measurement which gives a slightly enhanced LF response compared to Neumann's method of presentation. He states categorically that the AK20 is no less extended in it's LF response and rolls off no higher then either of the Schoeps capsules.

I'll be trying out a pair next week :)

0VU wrote: I'll be trying out a pair next week :)

0VU, what did you think of the KM120's?

8-)

Excellent.

I bought them :)

Great stuff, this is a real sleeper of a mic in my opinion, they seem to suffer from being collectively considered with the KM18* series, but the capsule and sound is something else.