Topic: mic pattern variation with frequency technical question

[Gutbucket]

What techniques do microphone manufactures use to make the directivity of a directional mic constant over the full frequency spectrum?

I understand how varying the sum of coincident pressure and velocity transducers (omni + fig-8 microphones) creates a directional pattern. I also understand how the same thing can be achieved using only one capsule with a longer sound path length in one direction, leading to destructive interference reducing the response in a particular direction.

But that overly simplified technique would seem to work at only a single frequency. Vary the frequency/wavelength and the cancellation would require a different path length from one side of the diaphragm to the other. That implies that for a mic with rear vents a fixed distance from the back of the diaphragm, the direction of the 'null' or maximally reduced response would change direction over a relatively narrow frequency range and much of the audible range would fall outside of that range entirely. That's basically what happens with spherical ball attachments used to modify the directional sensitivity of omni mics and also what happens with the polar radiation of speaker drivers on open baffles.

Although no directional microphone has a perfectly consistent polar pattern across its full range, it is amazing to me that they can be made to be as consistent as they are. What am I missing?

[boojum]

GB, I have not checked, but I would think that Eargle's "Microphones" would be a good place to start.   

[John Willett]

The polar pattern is derived by delaying the sound getting to the rear of the diaphragm - I think it's more of a time thing than a frequency thing.

[Gutbucket]

Boojum- I may need to find a copy.

John- You're right about the delay, but frequency = time.  The lower the frequency, the longer the wavelength and the greater the path length required to the rear of the diaphragm for the signal to be 180 degrees out of phase with the signal reaching the front of the diaphragm.

Because each frequency has a different wavelength, a different path length is required.

[John Willett]

John- You're right about the delay, but frequency = time.  The lower the frequency, the longer the wavelength and the greater the path length required to the rear of the diaphragm for the signal to be 180 degrees out of phase with the signal reaching the front of the diaphragm.

Because each frequency has a different wavelength, a different path length is required.

I think you'll find the Library in the Microphone-Data website extremely useful.

[Gutbucket]

Thanks, I'll check that out.

[DSatz]

Gutbucket, I can pretty easily explain some of this, but then there's a part that's more difficult.

First of all, you're right that the cancellation for rear-incident sound has to work in a way that's basically independent of wavelength. Where I think you're getting confused is, you seem to assume that the cardioid pattern achieves cancellation for rear-incident sound by combining the sound energy with an equal dose of itself in opposite polarity at a single point--kind of like matter meeting anti-matter and being annihilated. But that's not what happens.

What causes the cancellation is that the acoustic chamber behind the backplate delays, attenuates and filters rear-incident sound just enough to match the effect of the longer path that the same sound takes getting around to the front of the diaphragm. And the backplate in a pressure-gradient microphone isn't entirely sealed. As a result, rear-incident sound energy reaches both sides of the diaphragm at the same time and in the same polarity--and for that reason there is no net motion of the diaphragm.

This explanation, while truthful, leaves out a couple of things that become much more important when you try to fill in the picture of how the same microphone responds to front-incident sound--especially if you consider that a delayed version of the same sound also reaches the rear sound inlets, passes through the acoustic chamber and eventually reaches the back of the diaphragm. That's a longer discussion. But I wanted to put this part of the explanation out for you, since your understanding is spot on: The delays caused by path length differences and the acoustic labyrinth are constant with respect to time, not phase.

--best regards

[Gutbucket]

Ahhh.  Makes sense now, thank you very much.

You anticipated my next question about the delayed rear arrival for front-incident sound.  Another question that immediately comes to mind concerns the techniques used to delay the sound in the rear chamber.