Topic: figure 8 mic for M/S stereo setup

[midside]

yep, that one.

[DSatz]

midside, the CMTS stereo microphones used specially adapted matched pairs of MK 6 three-pattern capsules from the CMT series. I owned two of these microphones at various times back in the 1970s; they were manufactured through most the 1980s, even after the rest of the CMT series was discontinued.

An omni can very well be the mid microphone for an M/S pair; given unity-gain matrixing, you'll have the equivalent of X/Y cardioids. I'm no great fan of X/Y cardioids myself, but a lot of people sure use them, so I hope that this helps explain the design choice. Also, stereo microphones in general are occasionally used either as a "failsafe" way of "backing up" the placement of a single microphone (imagine if you were miking a world leader for a live broadcast and one channel went out for whatever reason--a bad cable, perhaps, but there's no time to find out. You can't exactly interrupt and say, "please hold on a moment while I get another microphone"). Finally, there is a technique called "Straus-Paket" recording in which the signals from a concident cardioid and omni are recorded onto two tracks, and then the mix between them can be varied during playback to produce any result along the range from cardioid to wide cardioid to omni "retroactively."

John Willett, some figure-8 microphones are not well suited to be the "S" microphone in an M/S pair. Some Royer ribbons (not the ones they sell for classical music, but some of the models that are more "pop" oriented), for example, have different-sounding fronts and backs; the company admits this freely, says that it is an intentional aspect of their design, and recommends using one side versus the other for different applications. That's certainly one approach to marketing, but it would make me think twice before using that type of microphone for M/S.

However, please note that all the polar diagrams which you reproduced showed no difference whatsoever between front and back below 16 kHz--and 16 kHz is at the upper limit of the frequency response for this type of capsule. Furthermore, since the the published curves of different manufacturers are not drawn the same way, they cannot be compared directly; we must still compare actual microphones. But mainly I object to the fallacy of presenting a plausible explanation for a proposed phenomenon (a "back story," if you will) as if it were any evidence that the phenomenon itself is real. This fallacy is so rampant in the audiophile universe that I have become quite allergic to it. I'm sorry to be "getting on your case" but you made an overly categorical claim, and while you've said many interesting things and you clearly do "know what you're talking about," you still haven't backed up your claim with any particular evidence about actual microphones so far. That's what it would take, I think.

--best regards

[John Willett]

.....some figure-8 microphones are not well suited to be the "S" microphone in an M/S pair. Some Royer ribbons (not the ones they sell for classical music, but some of the models that are more "pop" oriented), for example, have different-sounding fronts and backs; the company admits this freely, says that it is an intentional aspect of their design, and recommends using one side versus the other for different applications. That's certainly one approach to marketing, but it would make me think twice before using that type of microphone for M/S.

Agreed, I have no problem with this.

However, please note that all the polar diagrams which you reproduced showed no difference whatsoever between front and back below 16 kHz--and 16 kHz is at the upper limit of the frequency response for this type of capsule. Furthermore, since the the published curves of different manufacturers are not drawn the same way, they cannot be compared directly; we must still compare actual microphones. But mainly I object to the fallacy of presenting a plausible explanation for a proposed phenomenon (a "back story," if you will) as if it were any evidence that the phenomenon itself is real. This fallacy is so rampant in the audiophile universe that I have become quite allergic to it. I'm sorry to be "getting on your case" but you made an overly categorical claim, and while you've said many interesting things and you clearly do "know what you're talking about," you still haven't backed up your claim with any particular evidence about actual microphones so far. That's what it would take, I think.

I agree partly, but Neumann do specifically have a passive but otherwise identical front plate in order to make the capsule truly symmetrical both sides.  It's to do with the acoustic impedance of the capsule and how the soundwaves impinge on the capsule.

I have discussed this with both Stephan Peus at Neumann and Manfred Hibbing at Sennheiser.

The Schoeps is slightly different at high frequencies - how important that is in the real world is debatable, Jörg Wüttke obviously thinks it is not important - personally I want identical characteristics both sides at all frequencies.

Personally I use the MKH 30/40 for my MS recordings, but I do also have some Neumann KM-D series and will, no doubt, get the KK 120-D head when it becomes available (I have the KK 183-D now with KK 131-D on order).

[Gutbucket]

...An omni can very well be the mid microphone for an M/S pair; given unity-gain matrixing, you'll have the equivalent of X/Y cardioids...

Just to clarify, this is equivalent to coincident cardioids facing directly to either side, 180 degree angle between them.  Not the typical 90 degree or so X/Y angle most would be familiar with.

[DSatz]

Gutbucket, right--with 1:1 matrixing, that's what you'd get. However, it's a perfectly usable arrangement for stereo music recording, under any circumstances in which an orchestra or other ensemble could reasonably be picked up in mono using a single omnidirectional microphone.

That's the perspective from which to understand this design choice, I think. When this type of microphone was developed, engineers in the European broadcasting organizations still took great care to maintain mono compatibility because stereo receivers weren't universal yet. The "Walkman revolution" hadn't happened yet, either; few people listened to radio through headphones. LPs were still being issued as dual inventory--you could buy either the stereo or the mono version of many records (e.g. the Beatles). So "intensity stereo" (i.e. coincident miking) was how European broadcasters and record companies mainly recorded classical music.

As you may know, Schoeps' multi-pattern capsules are single-diaphragm designs which switch their directional patterns mechanically (acoustically), rather than by varying the polarization voltage on a pair of membranes. The company already made a three-pattern (omni/cardioid/figure-8), side-addressed capsule called the MK 6. Its pattern-switching mechanism is very complicated, but there isn't a very large market for stereo microphones, so it would have been prohibitive to develop a new mechanism to include other patterns (such as supercardioid and/or wide cardioid) which we might like to have in a stereo mike. The initial prototype version of the CMTS simply had a pair of these 20-mm-diameter MK 6 capsules mounted on a 30-mm-diameter body, but that looked a little odd, so they adapted the capsules to 30 mm housings and produced the microphone that way. Thus it is what it is.

--Just as a final comment, though I've said it before on this board: X/Y cardioids with only a 90 degree included angle is a generally poor choice for music recording, the Røde NT4 notwithstanding. A cardioid pattern isn't as sharply directional as people often believe or wish it to be. If you only spread a pair of cardioid microphones to 90 degrees and don't space them apart, the resulting recording will be about 3/4 mono. All the direct sound sources will crowd in toward the center of the image in playback, as compared to where they were during the actual recording.

If mono compatibility isn't a requirement and you want to use cardioid microphones, an "ORTF" arrangement gives a far more even, realistic distribution of sound sources within the stereo image in most cases.

But if for some reason it is necessary to record X/Y with cardioids, I'd strongly suggest angling them farther apart. 120 degrees might be a good starting point, particularly with smaller microphones which don't lose much high-frequency response off-axis. The point at which two ideal cardioids would cross at their half-power points would be with slightly greater than 130 (!) degrees of spread between them, ideally.

--best regards

[Gutbucket]

Well put and informative. 
Agreed on the lackluster 90deg X/Y. I'm not a big X/Y fan, even with wider angles, generally preferring spaced techniques.

...The point at which two ideal cardioids would cross at their half-power points would be with slightly greater than 130 (!) degrees of spread between them, ideally...

Good to know, I've assumed it was closer to 120 deg. I suppose this varies somewhat between different cardioid mics.

[DSatz]

Gutbucket wrote:

> I suppose this varies somewhat between different cardioid mics.

Well, cardioid is cardioid is cardioid--the sensitivity is proportional to (1 + cos theta) / 2, where theta is the angle of sound incidence relative to the 0-degree axis of the microphone. But many "cardioid" microphones are cardioid only in the midrange, and their variability in pattern is itself ... variable.

Large-diaphragm microphones generally lose their high-frequency response off-axis, so at those frequencies they are more narrowly directional than cardioid--sometimes quite considerably so. As a result you can get rather dull sound from the center of the image if you angle an X/Y pair of them at what would otherwise be your optimum. Smaller microphones disturb the physical sound field proportionately less; one inch, for example, represents half a sound wavelength at 6 to 7 kHz.

Meanwhile, microphones using the common dual-diaphragm arrangement (regardless of size) tend to become less directional at low frequencies, broadening out to more of a "wide cardioid" pattern at 100 Hz and below. Unfortunately this means that any X/Y recording made with such microphones will be nearly mono in the bass. 20, 30 or 40 years ago that was helpful for cutting stereo LPs that would track easily on cheap turntables, but not very many of us need to worry about that any more.

That characteristic robs stereo recordings of the sense of spaciousness which they could otherwise have. When an X/Y stereo recording has been made with dual-diaphragm cardioids, it's sometimes worthwhile to run it through an M/S matrix "backwards" to derive sum and difference signals, boosting the low frequency content in the difference channel a few dB while reducing it in the sum channel by a similar amount, then rematrixing to left/right stereo.

Better yet, don't use large, dual-diaphragm microphones as cardioids for X/Y stereo in the first place; it's almost a recipe for a worst-case outcome. If a person must use cardioids for X/Y stereo, then small, single-diaphragm cardioids (that are still cardioid at low frequencies) are preferable for the reasons I just explained; there are quite a few good kinds available.

But X/Y generally works even better with good supercardioids or, in those rare and wonderful cases where the geometry and reverberation balance conspire to allow it, with figure-8s ("Blumlein"). X/Y just barely works at all as a stereo recording method with cardioids, and as I said, ORTF or other "slightly spaced" stereo recording arrangements give a much more even distribution of apparent sound sources in playback.

--best regards

[Gutbucket]

... it's sometimes worthwhile to run it through an M/S matrix "backwards" to derive sum and difference signals, boosting the low frequency content in the difference channel a few dB while reducing it in the sum channel by a similar amount, then rematrixing to left/right stereo.
...

I've tried this with good results. One related thing I haven't tried but I'm curious about is the 'Blumlein Shuffler' technique for converting low frequency phase differences into level differences when using spaced mic techniques.  Every try it?

Apologies for the threadjack.

[muj]

David,

how would you run a pair of SDC which are not flat, but may have a good hf boost. an X/Y would from my experience accentuate the hf amount in the recording.

[DSatz]

> [H]ow would you run a pair of SDC which are not flat, but may have a good hf boost?

I don't see where the high-frequency boost changes things very much--if it sounds good, leave it in; if it's too much, cut it back with a good equalizer. Or am I missing the point of your question?

[Massive Dynamic]

But if for some reason it is necessary to record X/Y with cardioids, I'd strongly suggest angling them farther apart. 120 degrees might be a good starting point, particularly with smaller microphones which don't lose much high-frequency response off-axis. The point at which two ideal cardioids would cross at their half-power points would be with slightly greater than 130 (!) degrees of spread between them, ideally.

Having an LSD2, I often run coincident cards instead of Blumlein or M/S. Should the included angle be varied based on the distance to the loudspeakers and the distance between them? Any rule of thumb?

[muj]

> [H]ow would you run a pair of SDC which are not flat, but may have a good hf boost?

I don't see where the high-frequency boost changes things very much--if it sounds good, leave it in; if it's too much, cut it back with a good equalizer. Or am I missing the point of your question?

nope---I should definately try it and see what's up. I usually go omnis, but have started to back to more directional mics.


 

[saneproductions]

I am waiting on my recorder to arrive and I still need to order a second mic. I also moved, so it has been kind of hectic. I still plan to pursue this type of recording. I really appreciate all of the advice!!

Mike