Microphone technical information query

Linguist

New member
Hi all. I've been lurking in these forums regarding topics about microphone's frequency response, so while my question isn't about home recording, you guys seem to know a lot more about me regarding microphones' technical aspects.

Basically, I'm doing a research project looking at the effect different microphone types have on assessing voice disorders (I'm a speech therapist/pathologist).

I've got a research paper here that talks about different tests the researchers did on microphones. These tests were frequency response, sensitivity, total harmonic distortion (THD), and coherence. I've heard about all of these except the last one, coherence.

I've searched online and can't seem to find anything about microphone coherence. I'm wondering if you guys might know what it is, or if it's a different name for something more common? Or maybe it's something not even specific to microphones, and just electrical in/out devices in general.

I'll paste a quote from the paper that described what coherence is. They don't say much about how they tested it, just that they used white noise stimulus.

Coherence is a frequency-specific measure relating the output of a system to the input at each frequency. The coherence plot displays the degree of linear relationship between the output and the input while the transfer function displays the output as it relates to the input. A coherence value of 1 indicates maximum coherence (minimal distortion), while 0 is the lowest value possible. If the coherence function has a value between 0.9 and 1.0, distortion is assumed to be negligible.

Any help about what this could be would be greatly appreciated.
 
I've only ever heard of coherence relating to microphones when multiple microphones are in use. Could they be referring to multiple mics (or at least a single stereo microphone)? If not, it's a new on on me.
 
Agreed, only heard it referred to for the time domain in a pair, and maybe it can apply to a single mic's quality there as well?
Coherence is a frequency-specific measure relating the output of a system to the input at each frequency. The coherence plot displays the degree of linear relationship between the output and the input while the transfer function displays the output as it relates to the input. A coherence value of 1 indicates maximum coherence (minimal distortion), while 0 is the lowest value possible. If the coherence function has a value between 0.9 and 1.0, distortion is assumed to be negligible.
But to me this actually seems rather non specific. What kind of variation(s?) from the input?
 
This is a layman's view, but coherent is defined as 'logical and consistent', or 'Physics (of waves) having a constant phase relationship.'

I'd guess that in your case, coherence means comparing input to output in order to determine how accurate, or true to the source, the microphone is.
 
Thanks for you replies.

There are four microphones being tested, and that's what I found while searching - microphone coherence in an array. But, from that quote it sounds like they are comparing the input signal to the resulting microphone output signal, like what Steenamaroo said.

How they calculate how similar the two signals are to get a value of 0 - 1 with '1' being maximum coherence is what is a mystery to me.

They don't give any reason why they tested this 'coherence' in the paper, and there is no other mention of coherence after that quote. There is a sentence after that says 'In this case, all four microphones had a coherence of 0.9'.

In my study I will be recording one person speaking on several microphones (probably around 6) all at the same time. Is the coherence of several microphones recording at once something I should look into? If so, what is it/how do you calculate how similar the microphones are?

I don't necessarily want the microphones to be similar, since this is a study about how difference in microphone type (i.e. consumer/high end, cadioid/omni, condenser/dynamic) affects signals.
 
... In my study I will be recording one person speaking on several microphones (probably around 6) all at the same time. Is the coherence of several microphones recording at once something I should look into? If so, what is it/how do you calculate how similar the microphones are?

I don't necessarily want the microphones to be similar, since this is a study about how difference in microphone type (i.e. consumer/high end, cadioid/omni, condenser/dynamic) affects signals.
Definitely. The rule in general is if there are arrival time differences (distance differences) there will be phase cancelations at some frequencies -the greater the time difference the lower the frequencies involved. See ‘XY stereo mic placement vs. almost any other placement styles.
Also even if the physical distances are the same, there can be radiation direction and reflection differences from side-to-side variations.
 
Test like these will test 3x thing IMO. The speakers (or other sources) generating the test sounds. The room / environment of the test. And the mics. So it's really a grain of salt thing, even with optimal circumstances on all three legs.
 
Back
Top