Topic: Fundamental mic capsule ponderings

[SmokinJoe]

So I've been reading and thinking lately... perhaps falling into that category "the guy reads a few articles off the web and thinks he's an expert, but he don't know shit."

Fundamentally, there is (a) the omni pressure transducer, (b) the figure 8 pressure gradient transducer, and (c) the baffle approach to turn a figure 8 into a directional mic by delaying the sound wave a tiny bit.  The omni is simple enough that even I can understand it, no questions there.

First question... if I understand ribbon mics correctly, there is a ribbon hanging in space fixed at one end, with a magnetic field trying to hold the other end in place.  When sound pressure hits the ribbon, there is a small lateral excursion, which tugs the magnet on the end, and that generates a tiny voltage.  It seems to me if pressure comes from the front of the mic, or the back of the mic, it will generate the same magnitude excursion, and tug on the magnetic field the same amount, which generates the same little voltage.  The magnetic fields doesn't care if the ribbon is being pushed left or right.  If that's true, the same sound pressure impulse coming from the front or back should generate the signal voltage, no phase inversion when coming from the back side.  Yet "everyone knows" the back side of the mic is out of phase with the front.  Is the voltage generated from the ribbon swinging like a pendulum?  I assumed it was a linear motion, like a solenoid.  What am I missing?

I'm probably explaining it terribly, but I think I understand the fundamental idea behind directional (such as cardioid) mics.  There is a baffle system or passage ways, and the sound pressure wave coming from the back of the mic has to run through a maze, which delays it a tiny bit, just enough so the same sound pressure hits the front of the diaphragm at the same time, canceling each other out.   If I think about that, it's absolutely amazing that it works as well as it does, right up there with internal combustion engines and making a computer out of a bunch of transistors.  Multi-pattern LDC's (like AKG414) have 2 of these modified pressure gradient transducers facing in opposite directions.  Sum the voltages from the 2 cap in different ways, and you have yourself multi-patterns.  OK, I think I understand that.  My question is... in cardioid mode the back capsule is just not used at all, right?  But it's this big set of flat plates, it's got to be "in the way".  I would think a fixed card cap (my ADK-TC) would sound a hell of a lot better than double cap mics in cardioid mode (my TL's) but that doesn't seem to make a big difference.  What am I missing?  It's just all in the design of baffling I suppose.

Second question... a Schoeps MK8 is a single diaphragm condenser, intended to be very close to the "primary" figure 8 pattern, no doubling up on bastardized card caps here.  So how do they do that?  I'm guessing the backing plate is quite open and airy, which allows free air flow as much as possible, while still having enough meat left for capacitance.  I'm sure those Schoeps engineers are pretty sharp guys, and it took them a while to get it right, so why doesn't every other company make one like it?  Is it just because there isn't a big market? I suppose most people don't understand philosophically that it is very cool to have this pure and simple primary figure 8... most people would rather have a multi-pattern.  Still I would think there would be more examples out there, including a bunch of cheap Chinese copy cat models.  What am I missing?

[Gutbucket]

At the risk of being just another armchair electro/acoustic yahoo, I'll dive in a bit.  As for what I can offer, I have a better understanding of the general principles than the practical engineering aspects of microphone transducer design.

First off, in a ribbon microphone, the ribbon is attached at both ends. Its wavy corrugation and tension allows it to mechanically flex in the middle within the magnetic field.  Those two elements form a basic electric motor structure, or more specifically in this case a generator.  The ribbon is a simplified 'coil' moving back and forth in the static magnetic field.  Instead of a coil with many loops of conductor there is only one conductor path, the short ribbon itself, which generates miniscule voltages requiring a step up transformer to be useful. Push the ribbon one way and it generates a small voltage by moving the electrical 'coil' through the magnetic field, push it the other way and get the same voltage with opposite polarity when it moves through the same field in the other direction.  It's a simple direct current generator, but because the ribbon changes direction and moves back and forth through the field, it generates a small AC signal. The alternating current fequency corresponds to the changes of direction of the coil in the field. Contrast that with a similar, basic rotary DC generator in which the coil keeps passing through the field in the same direction as the stator rotates, generating direct current.

A dynamic diaphragm mic works the same way, but instead of the air pressure variations moving the conductor directly, it instead moves the diaphragm which has a much longer conductor attached to it in the form of a coil.  That makes it a more efficient generator since a longer conductor path is moving within a similar magnetic field.  In microphone terms that increased electrical generation efficiency equates to the microphone being more sensitive.

For more on how the basic fundamentals relate, refer to the Wikipedia entry on Electric Machines- A category which includes all motors, generators, microphones and transformers.

More on the other stuff later, unless someone else chimes in.

Oh, one other correction-

Fundamentally, there is (a) the omni pressure transducer, (b) the figure 8 pressure gradient transducer, and (c) the baffle approach to turn a figure 8 into a directional mic by delaying the sound wave a tiny bit.  The omni is simple enough that even I can understand it, no questions there.

Not fundamental in the same way.  Also, such a baffle works with any pattern by simply blocking the sound from the other side of the baffle.  Which frequencies it can block is determined by the size of the baffle and the distance of the microphone from it, and to a lesser extent by it's shape.

[Gutbucket]

One thing that may help in thinking about some of this which may be more familiar is considering how different types of loudspeakers and headphones operate. As electric machines they use the identical principles. There are important differences between speakers and microphones, but the basic generator/motor relationship is the same.   For every microphone type there is a corresponding speaker type.  The two main categories are those that work via electrostatic means verses those that work via elecromagnetics.

Electrostatic-
Condensor (capacitor) mics = electrostatic speakers and headphones

Electromagnetic-
'Standard' dynamic mics (diaphram & coil) = cone speakers (moving coils in magnetic field), most dynamic headphones
Ribbon mics = ribbon tweeters (moving ribbon conductor in magnetic field)
Fostex made some mics that work on the same principle as Magnapan speakers and the old Infinity EMIT tweeter drivers, and 'orthodynamic' headphones operate on the same idea.

There are also other categories such as piezo mics and tweeters.

[SmokinJoe]

Thanks for the ribbon explanation.  I should have done a better google search before I posted.  The key is a ribbon "motor" really is like a motor.
The whole ribbon is between the magnets http://www.rimagraphics.com/microphone.html
I thought 1 end moved up and down like a solenoid.

[MIQ]

Hi Smokin,

The free mini-book "Microphones Methods of Operation and Type Examples" by Neumann is really great for learning about microphone fundamentals.  You can find a link to it in this thread: http://taperssection.com/index.php?topic=139307.0  It has circuit schematics and math but its really not much more intense than some of the conversations you are already having with Gut.

Also mentioned in that thread is another good intro book "The Microphone Book" by John Eargle.  It's got more topics and is not free but is another source of info on mic fundamentals. 

-MIQ

[SmokinJoe]

Thanks for the link to the Neumann document.  I had browsed that document a couple of years ago, but had lost track of it.

Math doesn't scare me! I got an A in Differential Equations.... ummm... 27 years ago.... OK, I used to know math. 

[Gutbucket]

Forgot about that great Neumann document even though I had a copy here on the drive.  Should answer your questions way better than I can.

[DSatz]

SmokinJoe, to be honest, while the Schoeps MK 8 is a pure, single-diaphragm pressure gradient transducer, they didn't invent that approach. The earliest commercial condenser microphone, the Neumann CMV 3 with its various interchangeable capsule heads (1928) included a figure-8 capsule (model CM 7) with single-diaphragm construction and a symmetrical, push-pull backplate arrangement that was half a century or so ahead of Sennheiser's (see attached page from an old Telefunken brochure for the CMV 3). And even today there are single-diaphragm figure-8 capsules or microphones from other manufacturers, including Neumann (the AK 120 active capsule for the KM 100 series) and Sennheiser (the MKH 30 microphone).

Nonetheless I must admit I'm very fond of the Schoeps MK 8, particularly since Jerry Bruck and I are who convinced Schoeps to make this type of capsule for the Colette series.

--best regards

[John Willett]

SmokinJoe, to be honest, while the Schoeps MK 8 is a pure, single-diaphragm pressure gradient transducer, they didn't invent that approach. The earliest commercial condenser microphone, the Neumann CMV 3 with its various interchangeable capsule heads (1928) included a figure-8 capsule (model CM 7) with single-diaphragm construction and a symmetrical, push-pull backplate arrangement that was half a century or so ahead of Sennheiser's (see attached page from an old Telefunken brochure for the CMV 3). And even today there are single-diaphragm figure-8 capsules or microphones from other manufacturers, including Neumann (the AK 120 active capsule for the KM 100 series) and Sennheiser (the MKH 30 microphone).

My understanding is that the Neumann fig-8 is not a push-pull capsule like the MKH 30.

I think the problem is that having an active front-plate on a capsule of very high impedance like normal AF condensers can be quite problamatical.

My understanding is that the front plate on the Neumann figure-8 is for acoustic purposes only.  It is not an active front plate, but balances the back plate acoustically so that the figure-8 pattern is truly symmetrical front and back.

The Schoeps fig-8 does not have a truly symmetrical pattern as it is about 4 or 5dB down at 16kHz on the rear lobe.  This does not matter, of course, used normally; but if it was used as a side mic. in an MS set-up then the right channel would be a few dB down at 16kHz compared to the left channel.  Schoeps lovers say this does not matter of course 

[fguidry]

SmokinJoe, to be honest, while the Schoeps MK 8 is a pure, single-diaphragm pressure gradient transducer, they didn't invent that approach. The earliest commercial condenser microphone, the Neumann CMV 3 with its various interchangeable capsule heads (1928) included a figure-8 capsule (model CM 7) with single-diaphragm construction and a symmetrical, push-pull backplate arrangement that was half a century or so ahead of Sennheiser's (see attached page from an old Telefunken brochure for the CMV 3). And even today there are single-diaphragm figure-8 capsules or microphones from other manufacturers, including Neumann (the AK 120 active capsule for the KM 100 series) and Sennheiser (the MKH 30 microphone).

By understanding is that the Neumann fig-8 is not a push-pull capsule like the MKH 30.

I think the problem is that having an active front-plate on a capsule of very high impedance like normal AF condensers can be quite problamatical.

My understanding is that the front plate on the Neumann figure-8 is for acoustic purposes only.  It is not an active front plate, but balances the back plate acoustically so that the figure-8 pattern is truly symmetrical front and back.

The Schoeps fig-8 does not have a truly symmetrical pattern as it is about 4 or 5dB down at 16kHz on the rear lobe.  This does not matter, of course, used normally; but if it was used as a side mic. in an MS set-up then the right channel would be a few dB down at 16kHz compared to the left channel.  Schoeps lovers say this does not matter of course 

John, any chance you could whip up a sample clip demonstrating this issue and the impact on an actual recording?

Fran

[DSatz]

John, I didn't say that the Neumann AK 20 was a push-pull capsule; please go back and re-read my message. Also, you are both misstating and misinterpreting the polar response of the MK 8 at the top of its frequency range.

16 kHz or thereabouts is the physical upper limit of the honest response of any figure-8 capsule of this size. The physical driving force of the pressure gradient itself goes into an extremely rapid decline at high frequencies--there's a nice graph in Dr. Boré's little book that illustrates this fact, and I've attached a copy of it here.

As a result, most cardioids (for example) are actually functioning as pressure transducers at such frequencies; if they didn't, their response would roll off just like that of the Schoeps and Neumann figure-8 capsules. But because those figure-8 capsules are acoustically symmetrical they show no pressure response at the top of the audio range--and neither Schoeps nor Neumann uses electronic equalization to flatten and extend the capsule's response (although we both know another German manufacturer that does so).

16 kHz is a standard frequency for polar response graphs, so that is what Schoeps uses there. If they were to use, say, 14.5 kHz then there wouldn't even be the ~2 dB (not "about 4 or 5" as you wrongly claimed!) front/back difference in their polar response graph. I might wish that it weren't there at 16 kHz, but not because it's audible--rather, so that you and I wouldn't have to go over this same territory for what I believe is now the third time on this board.

--best regards

[Gutbucket]

Very interesting, thanks for the insight.  Am I correct in interpreting the speration of points (25mm in that graph) as the total path lengh from the front to the back of the diaphram through the backplates and around any immediate capsule housing?  If so I'd assume that Ft and more significantly the 1st comb-like dip above it to shift upwards in frequency with a shorter path length and downwards in frequency with a longer one.  Is that the case?

[John Willett]

John, I didn't say that the Neumann AK 20 was a push-pull capsule; please go back and re-read my message. Also, you are both misstating and misinterpreting the polar response of the MK 8 at the top of its frequency range.

16 kHz or thereabouts is the physical upper limit of the honest response of any figure-8 capsule of this size. The physical driving force of the pressure gradient itself goes into an extremely rapid decline at high frequencies--there's a nice graph in Dr. Boré's little book that illustrates this fact, and I've attached a copy of it here.

As a result, most cardioids (for example) are actually functioning as pressure transducers at such frequencies; if they didn't, their response would roll off just like that of the Schoeps and Neumann figure-8 capsules. But because those figure-8 capsules are acoustically symmetrical they show no pressure response at the top of the audio range--and neither Schoeps nor Neumann uses electronic equalization to flatten and extend the capsule's response (although we both know another German manufacturer that does so).

16 kHz is a standard frequency for polar response graphs, so that is what Schoeps uses there. If they were to use, say, 14.5 kHz then there wouldn't even be the ~2 dB (not "about 4 or 5" as you wrongly claimed!) front/back difference in their polar response graph. I might wish that it weren't there at 16 kHz, but not because it's audible--rather, so that you and I wouldn't have to go over this same territory for what I believe is now the third time on this board.

--best regards

According to Schoeps, the MK8 is 4dB down at the rear at 16kHz.



The Neumann KK120 is 0dB down at the rear and both lobes are fully symmetrical.

But the Schoeps has a smoother top-end response and the front lobe of the MK8 is smoother than the front lobe of the KK120 at 16kHz.

I'm not waving a flag for any particular capsule, just pointing out what is said in the published specs.

It's just that I would prefer both lobes of a fig-8 to be identical when using it as a side mic. in an MS set-up.

This, of course, does not matter at all when it's used for a Blumlein crossed pair or for a Faulkner Phased Array or used as a spot mic and the slight drop on the rear lobe would actually be more of an advantage used like this.

It's quite obvious to me why the Schoeps and Neumann capsules have been made this way and the compromises that the designers decided to make (and, yes, I know the designers of both mics and have talked to them both over the years).

Neumann went down the route of having a truly symmetrical polar-pattern at the expense of a slightly less smooth very top end.

Schoeps went down the route of having a better front lobe and smoother top end response at the expense of a slightly drooping top end on the rear lobe.

Both methods are very valid and they just decided to make slightly different compromises - as all capsules are a compromise of some sort.

Personally I treat Schoeps, Neumann, Sennheiser, DPA and Gefell as all equal and all the very top quality.  They are all slightly different as each designer decides where to make the compromises.  All I am doing is discussing what the manufacturers say about their own products.

I know you are very close to Schoeps, but I am not anti Schoeps at all, probably the opposite in many ways; just that I, personally, would prefer a truly symmetrical polar-pattern for a fig-8 used as a side mic. in MS, even though the effects of the droop are audibly very slight.

[fguidry]

... I, personally, would prefer a truly symmetrical polar-pattern for a fig-8 used as a side mic. in MS, even though the effects of the droop are audibly very slight.

If you're unable to create sample clips demonstrating this slight audible effect, perhaps you could point us to a commercial recording that demonstrates the issue.

Fran

[DSatz]

John, in Schoeps' polar graph the concentric circles are 5 dB apart. The -5 dB point is the circle within the 0 dB circle, NOT the point beneath it where the label says "-5 dB"; I think you have been mistaking that. The 16 kHz plot crosses the vertical axis about halfway between the -5 dB circle and the 0 dB circle, so the discrepancy shown in the graph is about 2 dB--not 4 or 5 dB. If the graph indicated what you thought it did, I might agree with you more, but it doesn't, and I don't.

Manufacturers, including the best ones, routinely smooth out bumps and dips that fall within their tolerance limits. What you are criticising in Schoeps' graph is something that most manufacturers would simply choose not to reveal. Yet in your message above you seem to take it for granted that graphs from different manufacturers, who use different measurement methods and different approaches toward deciding which response features they will publish or not publish, can be directly compared even in their smallest details. I think you know very well that this is not so.

About M/S: It might surprise you how little "S"-channel signal one picks up in live recording at 16 kHz relative to "M". Such high frequencies are usually only present in direct sound to any meaningful extent. The spectrum of arriving direct sound has a distinct high-frequency rolloff, but the spectrum of arriving reflected sound at that same point in space will have a much steeper rolloff, unless one is recording in an "echoic chamber".

Thus the higher the frequency, the greater the proportion of sound energy that arrives alongside a side-facing microphone--where the rejection of its figure-8 pattern is greatest. So the highest frequencies tend to be reproduced more toward the center of the stereo image. And yet this is OK because human hearing doesn't localize such high frequencies very well; such signal components are at best part of the "sheen" of the sound, for those who hear that high any more.

--best regards