Measuring Microphone Characteristics and Performance

[Gerard]

I am looking for guidance on the best methods (and pitfalls) for measuring the characteristics of microphones. I haven't been able to find any detailed discussion here, so I thought it might be useful to start a new topic/thread for the subject.

I am very excited by the response to my thread asking about DIY RF Condenser Mics. When I have all the parts, I am looking forward to making some and being able to join in constructively in the discussion. But I will need to be able to measure the mics in a meaningful way. What do the experts here suggest?

Just to set the scene, I will be doing this at home; I do not have access to an audio workshop or an anechoic chamber. Also, though I live in the countryside, I am too close to a motorway and a waste recycling plant (both within a mile, in different directions) for it ever to be silent in respect of man-made noise, except very early on Sunday mornings. On Sunday mornings it is like I remember it as a very young child; just wind and animal noises.

I have a couple of inexpensive measurement mics: a Behringer ECM8000 and a home-made mic based on a Panasonic WM-61A capsule that I made long ago following the Elliott Sound Products (Project 93) design. I made this around 10 years ago (I thought it was longer), when it was still - just - possible still to get WM-61A capsules. I think I did not "Linkwitz" it, but I'd have to open it up to be sure. The capsule is bonded into an aluminium tube with bathroom sealant, so I'm a little reluctant to risk damaging it.

Anyway, there's the request and the starting message for a thread. Can anyone give practical, proven advice?

 

[Tubetec]

Room EQ Wizard,
That can perform all kinds of measurements , I think theres a way of calibrating the  Behringer ecm mic built in  . Once the mic is calibrated ,you probably then need to figure out how to calibrate whatever your going to use to drive an acoustic signal into the mic on test . Theres probably a learning curve to climb to get everything set up right , maybe some of the guys on the REW topic will know more.

 

[Gerard]

Ha! So much for my diligent searching. I've just noticed "testing and measuring a microphone? " (https://groupdiy.com/index.php?topic=69150.0). Off to see if it has the kind of advice I am looking for.

 

[RuudNL]

The advantage of RoomEQ Wizard is that you can import the frequency response of the measuring microphone.
After you have done that, you can create correction data for the loudspeaker.
When you import both microphone response and the correction for the speaker, you will measure a 'flat' response.
Now measure the microphone you want to test and you will see the difference.
This assumes the correction for the measurement microphone is correct...

 

[Khron]

That's how i ran some sweeps for some of my DIY mics, using a calibrated measuring mic as a reference

With one of my Tannoy monitors on a small stool, in the middle of my bed, and the mics maybe 0.5m in front of the speaker ;D

The advantage of RoomEQ Wizard is that you can import the frequency response of the measuring microphone.
After you have done that, you can create correction data for the loudspeaker.
When you import both microphone response and the correction for the speaker, you will measure a 'flat' response.
Now measure the microphone you want to test and you will see the difference.
This assumes the correction for the measurement microphone is correct...

 

[Gerard]

Given that I don't have a certified measurement mic, what assumptions can be made about the characteristics of the Behringer ECM8000? It has been well cared for, kept in its box and mostly in a dry, reasonably warm room.

Meanwhile, I must investigate if anyone can do a calibration of it here in Dublin, so that I have something to import into REW.


Also, what distance between sound source (loudspeaker) and mic is normally used for calibrations?

 

[Khron]

http://www.cross-spectrum.com/weblog/2009/07/
https://www.avforums.co.za/index.php?topic=3074.0
http://www.audioheritage.org/vbulletin/showthread.php?11167-JBL-Control-1-summertime-project-surround-array-and-more!&s=d987e489915ac9ee6563ba7b57dcb2b5&p=186536&viewfull=1#post186536
http://techtalk.parts-express.com/forum/tech-talk-forum/26599-mic-cal-files
http://realtraps.com/art_microphones.htm

 

[Gerard]

Thanks for those links. It's far worse than I imagined. Clearly the frequency response 'curve' on the ECM8000 box is pure 120 proof marketing moonshine.

 

[Khron]

Afaik, "120 proof" = 60%, so pick one ;D

But yeah, one does have to keep the price in mind, and the brand / manufacturer, so...

For what it's worth, i got my EMX-7150 from iSemCon in Germany a few years back, for something around 200e or so, complete with serialized calibration (plotted in a PDF, and in an .frd file), and is what i used as a reference when i ran some "haphazard" sweeps on my mics

That being said, they also have calibration services available (under 100e, if memory serves; you'll wanna double-check to be sure), where you can send in your mic and they'll run the measurements for you and provide the data.

 

[kingkorg]

I use the same technique mentioned here. However there is a bit of a problem with the method.

Well basically we all measure mostly cardioid, but the callibration mic is omni. Which catches more room reflections than cardioid and less low end.

Does anyone have idea what full range speaker is used by the big guys for measurements? Speaker crossover is an issue.

There is one thing wanted to try for a while now, but am too lazy. As anechoic chamber is something i'll never have access to, free field or nature is the next best thing.

Well, one has to be lucky and find a spot with no wind, animals, hunters, etc. but, still seems like intresting idea. And there is power issue.

 

[dmp]

How about just measuring the noise level?
I've read about the A-weighting and dB-A equivalent noise and SNL methods.
Curious what has been the most useful way to measure noise.
 

 

[Khron]

Well, in order to ignore the room effects, one can gate the window of the impulse response that's used to plot the frequency response.
That does, though, have the side-effect of limiting the low-frequency side of the plot.

Keeping in mind that the speed of sound is about 1ft/ms (or 0.3m/ms, or 3ms for 1m), and knowing roughly how far the nearest surfaces are, you can kinda-sorta half-ass an estimate Close enough for our purposes, anyway ;D

Re: low end - the "ChineseK-12"-equipped mics of mine that i've measured, with the speaker about 0.5m (2ft-ish) away (Tannoy coax, enclosure rear-ported), showed a very gradual "tilt", or tapering off of the response, from high to low. So that might not be all that much of a concern either(?).
https://www.dropbox.com/s/blkhgb8eq0z07y9/Screenshot%202019-01-11%2020.06.30.png?dl=0

On-location / mobile measurements shouldn't be much of an issue - a laptop and a 2-channel USB interface (most, if not all, are bus-powered anyway), and job done :

I use the same technique mentioned here. However there is a bit of a problem with the method.

Well basically we all measure mostly cardioid, but the callibration mic is omni. Which catches more room reflections than cardioid and less low end.

Does anyone have idea what full range speaker is used by the big guys for measurements? Speaker crossover is an issue.

There is one thing wanted to try for a while now, but am too lazy. As anechoic chamber is something i'll never have access to, free field or nature is the next best thing.

Well, one has to be lucky and find a spot with no wind, animals, hunters, etc. but, still seems like intresting idea. And there is power issue.

 

[Gerard]

Does anyone have idea what full range speaker is used by the big guys for measurements?

From other things I have read and from photos on his website, I believe Professor Angelo Farina at the University of Parma, Italy, uses very short sound sweeps in a large warehouse for calibrating Ambisonic microphones. Microphones (both DUT and reference) and the loudspeaker are suspended mid-way between floor and roof structure in the middle of the warehouse. He takes a window on the measurement that is shorter than the first echo return.

I have also seen suggested using pipes as impedance tubes to couple a transducer to the microphone. The suggestion was to use different pipes for different frequency bands. But you still need something like a 20 m long pipe to measure down to 30 Hz properly.

I would like to be able to measure a lot lower than that, to see how accurately my mics would record a low (32') organ stop! Besides, some Sennheiser RF mics could measure down to 0.1 Hz (MKH 110/1). While the DIY RF Condenser Mic being discussed elsewhere will not go that low, I'd like to see how low it does go. I'll be building examples with various different capsules.