Topic: Schoeps Mk41 vs. Schoeps M-S "hypercardioid"

[thunderbolt]

I'm wondering if anyone has ever simultaneously recorded with a Mk41 pair and with a mid-side pair (i.e., Mk 4 and Mk eight and compared the results.  As I understand it, mid-side should approximate crossed hypercardioids when a cardioid is the mid and the mix is 50/50.  Assuming the angles between mics were adjusted to encompass an equivalent SRA, would they sound similar?

[acidjack]

Paging hi and lo.  I ran M-S MK5+MK8 and he ran MK41 side by side the other night at the same venue.  You can also compare my M-S Schoeps to MBHO Hypers right under them:
Schoeps M-S: http://archive.org/details/gpn2013-02-23.capitol_acidjack_schoeps
MBHO: http://archive.org/details/gpn2013-02-23.capitol_acidjack_MBHO

Keep in mind that the whole point of M-S is that it allows you to vary the width of your recording, so it wouldn't necessarily sound hyper-like unless you wanted it to. The more S you mix in, the more omnidirectional the recording becomes. 

Beyond that, while I know in *theory* they would sound similar, I think that especially if you are indoors and at some distance, having what are essentially two hypers pointed outwards at 180 degrees and a center cardiod pointed forward will not sound the same as two hypers at a given angle. 

[hi and lo]

Yep. I'll try to get that MK41 recording posted soon. They weren't on the same stand, but same height and maybe 2ft apart.

I agree with AJ about the flexibility of M/S and that it won't necessarily sound like a pair of hypers. It's only going to sound like hypers if the ratio of S to M is low and, making some theoretical assumptions, you could make it sound like a pair of cardioids, sub-cardiods, or omnis by varying the ratio.

[myke2241]

personally i like the MS recording more. the hypers phantom center is just mud... although the sides sound very good. Vox in your MS just pops out.

[acidjack]

personally i like the MS recording more. the hypers phantom center is just mud... although the sides sound very good. Vox in your MS just pops out.

I agree. I thought M-S worked well here.  It is especially useful in this venue because you have to be behind the board.  So, the people right in front of you are not talking, but (IME) the people off to the sides outside of our little section are all talking extremely loudly (it's a very chatty venue).  For both that and the more upfront vocals, I thought M-S was key.

[hi and lo]

personally i like the MS recording more. the hypers phantom center is just mud... although the sides sound very good. Vox in your MS just pops out.

I agree. I thought M-S worked well here.  It is especially useful in this venue because you have to be behind the board.  So, the people right in front of you are not talking, but (IME) the people off to the sides outside of our little section are all talking extremely loudly (it's a very chatty venue).  For both that and the more upfront vocals, I thought M-S was key.

That was my preference as well. The hypers just don't have the same 'pop' that the M/S recording has, however I think a three channel setup (hypers + center cardioid?) could be even better.

[myke2241]

MS can be decoded into three channels. So if you want something more directional i would just swap out the " M " channel with a hyper or super. Although this would create more dramatic movement in the image and give you some proximity effect.

[Todd R]

I'm wondering if anyone has ever simultaneously recorded with a Mk41 pair and with a mid-side pair (i.e., Mk 4 and Mk eight and compared the results.  As I understand it, mid-side should approximate crossed hypercardioids when a cardioid is the mid and the mix is 50/50.  Assuming the angles between mics were adjusted to encompass an equivalent SRA, would they sound similar?

A 50:50 mix of a mid cardioid with the side fig8 would (with ideal card/fig8 mics) yields a crossed hypercard pattern, but with an included angle of 120 degrees.  The relative mix creates both the pattern and the angle, and of course MS is a coincident mic pattern.  To get an angle of 90 degrees, yields something like a subcard pattern.

As to whether they would theoretically sound similar, the biggest difference is probably that a 50:50 mix gives you a hypercard pattern, but the mk41 is actually a supercard pattern, not a hypercard pattern.  Then of course most tapers are probably using a near-coincident pattern with the mk41's (like DIN), so the MS decoded mic would not have capsule separation, and typically tapers tend to use a 90 degree angle or less, not the effective 120 degree of the decoded MS.

Nothing against MS at all, and I wish I've had experience with it.  But since the mic pattern is tied to the effective angle as you decode MS, it isn't likely to yield the same effective results as an actual pair of mics (since the taper isn't as likely to choose that pattern and angle combo).  Whether the MS pair or the actual mk41 pair sounds better depends on tastes and the actual application of course.

[F.O.Bean]

Well said Todd!!!

[Brian Skalinder]

In addition to the issues Todd R noted, the M-S would also pick up its mid component from the on-axis Mid mic, whereas an XY hypercardioid pair at 120º would pick up its mid component off-axis to one degree or another (based on the width of the sound stage relative to the XY array).  Depending on the quality of the hypercardioid mics and their off axis response at various frequencies (and a variety of playback variables), this may or may not produce an audible difference.

[MIQ]

Mid=card
M/S Mix=50/50
SRA=80 deg

Miq

[MIQ]

Coincident pair of Hypers
Angle=120 deg
SRA= 76 deg

Similar SRA but as Brian noted you are relying on the off axis response to capture the center info when using the coincident hypers.

Miq

[MIQ]

Coincident Supercards
Mic Angle = 120 deg
SRA = 100 deg

Considerably different SRA.

Miq

[Tom McCreadie]

A 50:50 mix of a mid cardioid with the side fig8 would (with ideal card/fig8 mics) yields a crossed hypercard pattern, but with an included angle of 120 degrees.  The relative mix creates both the pattern and the angle, and of course MS is a coincident mic pattern.  To get an angle of 90 degrees, yields something like a subcard pattern.

If you do the maths, a 1:1 M/S mix of an ideal cardioid and Fig-8 actually translates to a virtual XY pair of included angle 126.9 deg.
And this virtual mic has a "V=a + b.cos(theta)"  polar pattern of V = 0.309 + 0.691.cos(theta).

So It's quite reasonable to label such an outcome as "supercardioid". Various literature sources - Woram, Witteck (Schoeps), Jecklin, Sengpiel, Eargle, Dooley & Streicher etc - list a supecardioid as having "a" ranging from 0.33 to 0.37, with "b" from 0.67 to 0.63. A hypercardioid is generally put in the ballpark of  "ä" = 0.25 and "b" = 0.75. The MK41 is supercardioid.

All academic of course, for we never have perfect mics. [and my nitpicking about names reminds me of those German technical hobby mags - Hi-Fi, photo etc.-  where a stressed reader would send in a letter with: "Please help. I can't sleep!  Please clarify: is Model ABC in the Ubermittelklasse or the Unterspitzenklase?"  :-) ]

But one can _never_ generate a subcardioid pattern in the virtual X(Y) mic, starting out with a cardioid for M. The pattern of the virtual mic is always an intermediate of the two 'parent' patterns and can never be fatter than that of the starting M mic, no matter how much extra M gain applied. One needs to start out with something fatter for M than a subcardioid  - e.g omni. In such a scenario, to achieve a subcardioid pattern of, say, V = 0.70 + 0.30.cos(theta), an M:S ratio of 2.33 is needed (i.e. omni M level is  7.36dB higher than the Fig-8 S). The included angle of this XY virtual pair is - no surprises here - 180 deg. i.e back to back subcardioids.

Some folks describe MS as being "more flexible" than XY. Not quite true, for the width can always be readjusted by the additional matrix step:
XY -> MS -> M'S' -> X'Y'. I guess you could say that XY is not less-flexible, merely longer-winded...like this post.   :-)

[MIQ]

Various literature sources - Woram, Witteck (Schoeps), Jecklin, Sengpiel, Eargle, Dooley & Streicher etc - list a supecardioid as having "a" ranging from 0.33 to 0.37, with "b" from 0.67 to 0.63. A hypercardioid is generally put in the ballpark of  "ä" = 0.25 and "b" = 0.75. The MK41 is supercardioid.

This. It is a little confusing subtlety and as you noted mostly academic. It's something you rarely see mentioned with respect to determining (calculating) the SRA.  The visualizers above both include two Supercard settings.  One for a= Cm=0.37 and one for Cm= 0.34.  These values seem to correspond reasonably well to the popular Supercard polar responses available from mic manufacturers.

Miq

[myke2241]

XY relies on phantom center. MS actually has a center. Although the math may say they are the same, they are indeed different in how they physically pickup sound. A this makes a difference.

[Tom McCreadie]

XY relies on phantom center. MS actually has a center. Although the math may say they are the same, they are indeed different in how they physically pickup sound. A this makes a difference.

Not really. The concept of "phantom center" is purely a property associated with playback through two L,R loudspeakers, usually in your sitting room, Every method of recording whatsoever, when played back through two speakers, has to rely on the principle of a phantom center to give the illusion of centered sounds coming from a third, centered speaker that isn't there. If all mics involved had theoretical patterns, with perfect off-axis responses at all frequencies, then the M/S and XY techniques would be absolutely indistinguishable in the quality of the phantom center - or the entire stereo image, for that matter. Of course, the mics are never perfect, so, in a classical or acoustical music context, at least, M/S will generally give a more satisfying, solid center because the M mic gets to point its "best, on-axis side" at the important center stuff.

In the context of a rock venue, though, where almost all the sound might happen to come from L, R  PA stacks, you might be grappling with _two_ phantom-imaging instances: that from the stacks for the attendees during the show..and then later again from speaker playback, at home on the sofa. In that case it might make sense to employ XY and have their "best, on-axis side" pointing at the stacks? Think of it as avoiding a double whammy and postponing all the phantom witchcraft until you get home.   :-) 

[DSatz]

To sum up a part of this discussion, for every possible M/S arrangement there is a theoretically equivalent X/Y arrangement, although it might require a pair of "theoretical microphones" with some in-between pattern that you can't buy off the shelf.

But when people say, OK, given that principle, M/S is still better than X/Y because in M/S, the M microphone is facing directly forward while in X/Y, both mikes are facing somewhat to the side, and are therefore picking up a lot of the direct sound from off-axis angles (where their frequency response varies from the spec-sheet curve)--I have two things to say in response.

Actually, the first thing I have to say is a rhetorical question: "But what about the "S" channel microphone? You're aiming its "null" front and center. You can't get more off-axis than that."

And the second thing is, the observation is completely valid that X/Y stereo recording places huge demands on the off-axis response characteristics of microphones. That's really fundamental. Actually it's not just X/Y stereo, but also closely-spaced pairs or practically any other combination of two or more main microphones that you'd ever be likely to mount on the same mike stand. Ideally we shouldn't classify microphones as "cardioids" or "omnis," etc.--that's too vague and general for our purposes. Ideally we should describe them the way they function both on-axis and off-axis. But since our terminology isn't that precise, we need to keep that aspect in mind when talking about microphones in such oversimplified terms as "omni" or "cardioid," etc.

--best regards

Quick P.S.: The actual hypercardioid pattern is very rare. Most microphones labeled as either hyper- or supercardioid are actually a mixture that usually comes closer to being supercardioid (but isn't exactly supercardioid, either). This applies to Schoeps, Neumann, AKG, Sennheiser, Beyer and many others.

[thunderbolt]

This is extremely interesting, and some of you guys have the maths calculated beyond what I can grasp.  What I find fascinating is, as DSatz says, we do place huge demands on the off-axis response of our mics.  And it is very true, that you can't get more extreme than facing the nulls of a figure eight toward your source.

On a subjective level, I do think that the M-S is appreciably better, for the reasons listed.  There is a central focus, yes at the expense of crowd noise on the side, but the central image and vox do sound more "direct" and present, and it definitely has more "pop."

AJ, how much of the fig eight did you mix relative to the card mid in your post-processing?  Sounds like you nailed it.

BR,
Corbin

[noahbickart]

I am finding that running MS allows me to use a "wider" capsule for the M than I ever would running ORTF, DIN, etc.  my schoeps mk22 is getting a great deal of use.

Also, look at the pattern created by running an mk41 m/s some time- it's quai blumlien! Cool.

[Tom McCreadie]

Actually, the first thing I have to say is a rhetorical question: "But what about the "S" channel microphone? You're aiming its "null" front and center. You can't get more off-axis than that."

But then the MS zealot would probably retort: "So what! The sound signal captured from the null direction of the S is, by the very nature of 'null', of such a tiny level anyway that - even though it may be brimming with 'off-axis distortion' - it still can't make a huge impact on the overall audio picture."

(Not that I'm an MS zealot or have a horse in this race...just throwing in some arguments from both sides of the fence.)

[noahbickart]

Actually, the first thing I have to say is a rhetorical question: "But what about the "S" channel microphone? You're aiming its "null" front and center. You can't get more off-axis than that."

But then the MS zealot would probably retort: "So what! The sound signal captured from the null direction of the S is, by the very nature of 'null', of such a tiny level anyway that - even though it may be brimming with 'off-axis distortion' - it still can't make a huge impact on the overall audio picture."

(Not that I'm an MS zealot or have a horse in this race...just throwing in some arguments from both sides of the fence.)

Right, but his point is that precisely that when you go to a concert and try to record it- intentionally arranging one of your microphones such that its null is pointed at the musicians sounds counter intuitive, which it is.

But, we all know however how important capturing the space in which a performance occurs is. Having pinpoint control over the ratio of direct to reflected sound is the reason to use m/s.

[Scooter123]

I own both and sometimes run redundant rigs, with the only difference being the mikes. 

There is no discernable difference between the mk41 and the mk4.  Well maybe there is, but I couldn't notice anything that jumped out at me. 

[DSatz]

Tom, you've hit on the reason I put my little rhetorical question in quote marks: In the particular way that I put it, it isn't actually any better-founded than the statement I was using it to criticize. And its logical weakness is just as you said: The null of a well-made figure-8 is typically 20+ dB below the sensitivity of the "M" channel microphone across the entire audio range. So within a certain wedge of sound-arrival angles, the "S" channel's contribution just isn't audibly significant. This corresponds to the very center of the stereo image in playback--it pretty much all comes from the "M" mike.

But when you get, say, 30 degrees or farther away from the null, it's quite another matter. I think I've already told the story of my sad attempt, over a decade ago by now, to use a pair of Neumann KM 86 microphones for Blumlein stereo with a very energetic-sounding soprano soloist. The KM 86 was an excellent microphone for a lot of things, but look at its polar diagram in the figure-8 setting as shown below: It has a sizeable high-frequency response peak around ±45° (for those not versed in reading polar diagrams, the main point to note is the difference between the 8 kHz and the 1 kHz response around those angles). And those, of course, are the angles that face forward in a Blumlein setup; consequently the soprano voice came out harsh, almost spitty sounding. For the same reason, that figure-8 would be a poor choice as the "S" microphone in an M/S pair.

My point ("and I do have one," as Ellen DeGeneres says) is that the wrong virtue is being claimed for the front-facing microphone in an M/S pair. Historically, M/S was introduced as a mono-compatible stereo recording approach--as contrasted with spaced-omni stereo, which causes huge comb filter effects when you sum the channels. That distinction harks back to the 1950s when stereo records and FM radio were still new, and most people, even hi-fi enthusiasts, still had mono systems (I remember the single, big Electro-Voice speaker in our living room). In western Europe the classical music broadcasters were still publicly controlled, and they adopted a policy of using compatible stereo recording methods so as not to stick the majority of their listeners with bad sound.

(Incidentally, the whole way stereo FM works is directly analogous to M/S recording--the main carrier conveys the mono, L+R sum, a subcarrier carries an L-R difference signal, and these two signals are converted to L/R stereo by sum-and-difference matrixing inside the tuner or radio. A similar analogy could even be made with the vertical and horizontal modulation of phonograph grooves, with horizontal = M and vertical = S, except that stereo phonograph cartridges didn't generally pick up vertical and horizontal modulation separately.)

The real advantage of a front-facing "M" microphone isn't any beneficial effect on the resulting L/R stereo signal. Rather, it's the benefit that it has for the sound quality of the compatible mono signal--because that signal isn't derived by summing the off-axis response of two microphones as it would be given an X/Y recording (imagine how a mono version of my Blumlein recording would have sounded given the ±45° shown in the polar diagrams!), but rather, it is delivered directly by a single, high-quality, forward-facing "M" microphone--the "S" microphone being totally out of the picture where mono playback is concerned.

I could say more about this, and I probably will eventually, but I've talked long enough for one evening already.

--best regards

[Tom McCreadie]

Having pinpoint control over the ratio of direct to reflected sound is the reason to use m/s.

I'm unsure that MS has this purported big advantage of pinpoint control over the direct to reverb sound ratio. After all, the direct/reverb outcome depends on 3 decisions you make: (1) mic stand placement; (2) capsule-pattern choice for the X/Y mics or the M mic; (3) XY capsule-angling (or M/S ratio). And both our competing methods, MS or XY, allow you freedom to make those 3 choices. Indeed, MS might even be at a slight disadvantage, as it really needs an on-the-fly conversion to XY for a proper live listening evaluation (to nail the correct stand placement - a crucial parameter that, if judged wrongly, can never be "fixed in post.")

But granted, MS does have one significant additional element of control flexibility: you are freed from the shackles of having to employ matched, same-brand, same-pattern mic pairs. Not only could you use, say, a Neumann brand as M paired with a Schoeps as S, but you also then have access to a practical continuum of patterns for M: omni, omni with high freq beaming, subcardioid, wide cardioid, bit less wide cardioid, narrow cardioid, supercardioid....Fig-8 etc  This puts a whole arsenal of possibilities at your disposal.

[F.O.Bean]

I own both and sometimes run redundant rigs, with the only difference being the mikes. 

There is no discernable difference between the mk41 and the mk4.  Well maybe there is, but I couldn't notice anything that jumped out at me. 

I agree completely. Tonedeaf and I run side by side all the time. Me running mk41>LB him mk4>psp2. the preamps are the only difference I can hear, and its MINIMAL at that

[acidjack]

I think it's a testament to the MK41 that it has such similar tonal balance and response as it does to the MK4; however I wouldn't say it sounds "the same" at all.  The supercardiod pattern clearly picks up more direct and less reflected sound, which is why there are some rooms where I absolutely always run them.  In the ones where I feel better about running MK5s, the 41s still do OK, but the cardiod pattern clearly gives a more open, full sound.  What's incredible about the MK41 is that it doesn't have the harsh tin-can sound of most of the other hypers out there. 

[F.O.Bean]

I think it's a testament to the MK41 that it has such similar tonal balance and response as it does to the MK4; however I wouldn't say it sounds "the same" at all.  The supercardiod pattern clearly picks up more direct and less reflected sound, which is why there are some rooms where I absolutely always run them.  In the ones where I feel better about running MK5s, the 41s still do OK, but the cardiod pattern clearly gives a more open, full sound.  What's incredible about the MK41 is that it doesn't have the harsh tin-can sound of most of the other hypers out there. 

100% agreed. The mk41 is super full and balanced

[Scooter123]

"What's incredible about the MK41 is that it doesn't have the harsh tin-can sound of most of the other hypers out there"

By that you mean a warm, muddy sound. 

I like warm muddy sounds...........

[F.O.Bean]

"What's incredible about the MK41 is that it doesn't have the harsh tin-can sound of most of the other hypers out there"

By that you mean a warm, muddy sound. 

I like warm muddy sounds...........

Tonedeaf and I do too