Microphone calibration file | how it works
Dec 26, 2019 at 11:47 AM Thread Starter Post #1 of 5

Samk15

New Head-Fier
Joined
Sep 29, 2019
Posts
24
Likes
1
Location
Australia
Hi,

I looked at the calibration files of multiple microphones and couldn't figure out how it works when loaded into REW or similar software?

Lets say if I have a value of -0.9 for 29Hz and + 1.5 for 2Khz.
How the software translate and compensate these values?

Do the calibration file tell the software to invert those values in the ouput? Or these numbers are already inverted in the calibration file?
 
Last edited:
Dec 27, 2019 at 7:14 AM Post #3 of 5
In general a calibration file is simply how something measured. A microphone is measured by the manufacturer against a known reference of flat, and instead of being dead flat it gives -0.9dB at 29Hz. Then when you later use that microphone everything it measures is going to show a value 0.9dB too low at 29Hz(because that's how the mic measures). So of course the software should compensate for that error by adding 0.9dB at 29Hz to anything you measure with that mic. That's where the opposite effect of the value in the file comes from. the file gives the error, the software corrects it so we can get back to 0.

If you end up trying to create your own compensations, you might find this interesting https://www.head-fi.org/threads/general-iem-measurements-discussions.903455/page-11#post-14979787 or not ^_^
 
Dec 27, 2019 at 6:31 PM Post #4 of 5
Thank you.
I always find your posts interesting.
I started to measure one of my earphones and i noticed that there is a weird spikes in the frequency spectrum when nothing was attached to the mic input. So I assume that these spikes could be the result of circuit board. Therefore I wanted to make my own calibration file to eliminate any interference from the cable/circuit board before attaching the mic.
 
Dec 27, 2019 at 7:34 PM Post #5 of 5
If the circuit isn't closed(nothing is plugged into the input), then you're measuring noises instead of the signal. Those noises will hopefully be many magnitudes below whatever you're going to measure in practice with your mic, and should not matter for FR graphs. But in any case, noise should not be compensated because it's noise, not signal.

Now if you loop your system by connecting the output of your DAC or amp into the input of your ADC(soundcard), you can then measure the possible change in FR induced by the DAC+amp+ADC. REW even has everything ready for that under 'preferences', on the 'soundcard' tab there is an area called 'calibration'. The 'calibrate' button exists for the very purpose of measuring that loopback signal without a mic being involved. And the next button 'make cal...' generates a calibration file from that measurement to put everything back to zero before you start using the mic. That calibration does the same thing the mic calibration in the other tab does in term of FR compensation, and you can use both at the same time if needed.
 

Users who are viewing this thread

Back
Top