Topic: Fundamental mic capsule ponderings

[John Willett]

I do take what you say - but with so many people saying the importance of having matched pairs for stereo why would you deliberately choose a mic. that would not give the equivalent of a matched pair, however small the difference?  That's all.

[DSatz]

John, I actually am a proponent of close matching for directional microphones that are used either as coincident or as closely-spaced pairs. Despite that I really do think all the (rather many) things that I said in my last message.

"On the third hand," so to speak, I can also imagine that if I felt a strong loyalty to a manufacturer that competes with Schoeps, I might well take the front/back discrepancy at 16 kHz as an opportunity to take a good jab at the competition.

And yet "on the fourth hand" (as soon as I finish writing this message, I think I'll go enjoy a nice game of bridge with myself), even with the loyalty that I do have, I take your point as well; I would prefer it if the 180° response of the MK 8 at 16 kHz could be spot-on identical to its 0° response. What my experience and knowledge and rational thinking tell me about interpreting that specification may be one thing, but it is so much nicer to have a simpler picture that is free of controversy and that doesn't require any explanation at all (let alone the rivers I pour forth sometimes).

However, I would really appreciate it if you would kindly acknowledge that the graph doesn't indicate a "4 or 5 dB" difference between the MK 8's front and back response at any frequency; it is simply not what the graph says, nor is it the reality of the capsule's performance.

--best regards

[John Willett]

John, I actually am a proponent of close matching for directional microphones that are used either as coincident or as closely-spaced pairs. Despite that I really do think all the (rather many) things that I said in my last message.

"On the third hand," so to speak, I can also imagine that if I felt a strong loyalty to a manufacturer that competes with Schoeps, I might well take the front/back discrepancy at 16 kHz as an opportunity to take a good jab at the competition.

Actually "loyalty" doesn't really come into it as I have a great respect for Schoeps and I went into detail as to why I thought the designers made the decisions they did.

I did point out the pros and cons of both the Neumann and Schoeps designs and why I thought they are each made the way they are.

And Schoeps do make some magic mics, especially IMHO some of their omnis, open omni and hypo-cardioid.

It's just that as a side mic. for MS I would like the fig-8 to be as symmetrical as possible.

However, I would really appreciate it if you would kindly acknowledge that the graph doesn't indicate a "4 or 5 dB" difference between the MK 8's front and back response at any frequency; it is simply not what the graph says, nor is it the reality of the capsule's performance.

Sorry, I thought my previous reply accepted this.  I was taking the minus sign as the mark and it therefore looked like 4dB down.

Now you have explained it, I can see that it's just 2.5dB down here.

[DSatz]

John, when I wrote earlier that "I can also imagine that if I felt a strong loyalty to a manufacturer that competes with Schoeps, I might well take the front/back discrepancy at 16 kHz as an opportunity to take a good jab at the competition," I truly was speaking about myself--I know that "I can resist everything but temptation," as Oscar Wilde once wrote.

The really nasty bit about bias is that it works on a person even when he or she is trying to be above it all. I rewrite many of the messages that I post on this board, because often I find later on that I'd completely missed the balance that I'd thought I'd achieved. Somehow that never changes--as long as I express opinions and views, I seem bound to recreating this same problem again and again, and for all I know, the messages with which I'm the most satisfied today may well be exactly the ones that, some years later, I will particularly feel embarrassed by.

From that standpoint it would probably be better to be more "objective" than I generally am about most things. (Insert your favorite ironic remark here about the likelihood of that ...)

--best regards

[ghellquist]

Aah. Now the guys are at it again. A "not-very-theoretical", "almost-beginner-level" question moved over to arguments about a few different figure-8 mics.
Now, please, start your own thread instead, don´t steal this one as well.

Back to fundamentals.
We have covered the figure 8 mic. The classical example is a ribbon mic. It contains a ribbon, fixed at both ends, and hanging in a rather hefty magnetic field. Move the ribbon backward and it gives a positive signal, move in forwards and it gives a negative signal.
The important point is to understand is that the ribbon mic measures the "movement" of air around the ribbon. Give it a low enough frequency and the air will simply move around the mic, not giving any signal at all.
 
The omni mic instead consists of a closed volume of air and a diaphragm in front of it. This mic measures "air pressure". It can measure even very low frequencys, say the change in air pressure between a high pressure and a low pressure in the weather. Of course, this is impractical, so there is a small vent allowing air pressure to equalize.

The interesting thing is that the pressure is not sensitive to which direction sound comes from. But movement is very sensitive to the actual direction.

There are different ways to make a cardiod mics. The simplest in theory is to combine one figure 8 mic and one omni mic. There are actual examples of this mic design from the early days. The theory goes as follows: place both mics ( omni and figure 8 ) very close to each other and add the two signals to each others. If a sound comes from the front of this setup, the omni mic will give a signal, say plus 1 V. The figure 8 will as well give a positive signal, say 1V as well. If we add these two signals, a sound from the front will give +2V.

On the other hand, a signal from the back will work differently. The omni mic will still give a positive signal, say 1V. But the figure 8 mic will give a negative signal, say -1V. Remember that the omni mic is not sensitive to direction to the sound source, but the figure 8 is. Add these two signals and we get zero volts for a signal coming from the back.

Follow the situation "around the clock" and we get the normal cardioid pattern. It takes a bit of math but it can be, I will not do it here though.

// Gunnar

[ghellquist]

Now for the situation with one diapraghm and a labyrinth.
Somehow this will combine a figure 8 and an omni in the same capsule. How does this happen?

If you look closely on a mic capsule you will see that towards the front there is as diaphragm. This obviously caters for the signal coming from the front giving a positive signal.

On the back of the mic there is a kind of labyrinth. It works by delaying the sound waves in the air slightly. So if a signal comes from the back of the mic, it is delayed slightly. Enough so that the signal going in the back of the mic and the signal going around the mic impinging on the membrane from the front will reach the membrane at the same time. The effect is that no signal from the back will be seen by the mic.

A sound wave from the front first hits the membrane in the front and creates a positive signal. It also goes goes around the edge of the mic, through the labyrinth and is attenuated and delayed. The net effect is a slightly lower signal but still enough to  be useable.

One thing to note about this design of cardioid is that it is frequency dependant. Looking at signals from the front. With high frequencys the acoustic delay is long enough for the "back signal" to arrive much later than the front signal allowing high frequencys to not be attenuated. On low frequencys, the delay is short compared to the frequency of the sound wave and hence the signal is very effectively dampened. This is part of the reason why cardioid mics fall off in the bass frequencys.

One of the realworld problems with cardioc mic designs is that they by necessity are artistic compromises. If you look at a typical LDC mic you will find that the backplate has a very complex pattern of holes. Not all holes goes through the backplate fully, there may be different diameter holes. This is part of the design giving the mic its signature sound. Getting this exactly right, every time, is quite complicated and involves very small tolerances for manufacturing errors.

Now, if we created the cardioid mic, why not point one forward and one backward and combine the output from them. The nice thing is that if we:
 -  add the two signals we get an omni mic (more or less)
 -  subtract the two signals we get a figure 8 mic (more or less)

And the two cardiods can actually share the same backplate. One membrane each way and a common backplate ( with the acoustic delay labyrinth ) is how it is done in many switchable studio mics.

There are some other fundamental designs of mics, one of them is the interference tube used in shotgun mics, let us save that for another time.

Gunnar

[Gutbucket]

The omni mic instead consists of a closed volume of air and a diaphragm in front of it. This mic measures "air pressure". It can measure even very low frequencys, say the change in air pressure between a high pressure and a low pressure in the weather. Of course, this is impractical, so there is a small vent allowing air pressure to equalize.

Not impractical, just unwanted for this application.  There is nothing impractical about a barometer which is simply a pressure transducer tuned for long term response around 0Hz.  An absolute pressure omni microphone would have DC offset that correlates to barometric pressure, which isn't something recordist's usually want. 

Though it would be cool for a super recording system designed to record absolute instead of relative values of signal level and pressure.  Cue the thought experiments and all that..

[John Willett]

There are different ways to make a cardiod mics. The simplest in theory is to combine one figure 8 mic and one omni mic. There are actual examples of this mic design from the early days. The theory goes as follows: place both mics ( omni and figure 8 ) very close to each other and add the two signals to each others. If a sound comes from the front of this setup, the omni mic will give a signal, say plus 1 V. The figure 8 will as well give a positive signal, say 1V as well. If we add these two signals, a sound from the front will give +2V.

On the other hand, a signal from the back will work differently. The omni mic will still give a positive signal, say 1V. But the figure 8 mic will give a negative signal, say -1V. Remember that the omni mic is not sensitive to direction to the sound source, but the figure 8 is. Add these two signals and we get zero volts for a signal coming from the back.

Follow the situation "around the clock" and we get the normal cardioid pattern. It takes a bit of math but it can be, I will not do it here though.

// Gunnar

Maybe this will help explain - two screenshots from a Powerpoint presentation I do:

The first shows omni and fig-8 polar-patterns superimposed with +ve and -ve symbols.
The second shows how these make a cardioid pattern.



Click on the thumbnail for the full size picture.

I hope this helps.

[DSatz]

Gutbucket, it really is impractical for a capsule to have pressure response down to DC, because ordinary day-to-day changes in barometric pressure would affect the resting tension of the diaphragm (altering the capsule's frequency response) and the diaphragm could even burst if the capsule was shipped or transported by air.

--best regards

[Gutbucket]

Understood.  I meant only that such a thing is not impractical to build from an engineering standpoint, though I conceed it's certainly impractical for a microphone as typically designed (in addition to being undesirable).

What I should have pointed out more clearly is that impracticality stems from engineering tradeoffs- microphones designed for linear response to small pressure changes happening up to tens of thousands of times per second are the result of different engineering choices than what is appropriate for barometric pressure sensors.  That required speed, sensitivity and accuracy of microphones tend to make them far more delicate devices.  It is certainly possible to build a microphone that has response to 0hz and is not destroyed by it, but the resulting device would be somewhat different than both current pressure omni microphones and barometric pressure sensors.  The combination of the two could be one approach.  Since there is no real demand for such a microphone/sensor there has been no effort made to develop one, but there is nothing which inherently prevents anyone from doing so.

I realize I'm belaboring the 'real-world' limitations of actual microphones.

[edit- ponderous]

[aracu]

I do take what you say - but with so many people saying the importance of having matched pairs for stereo why would you deliberately choose a mic. that would not give the equivalent of a matched pair, however small the difference?  That's all.

In choosing any specific mic, you would weigh the comprimises in the design which you were aware of, with the benefits of the design for practical use.