Advice on characterizing microphones

[emw]

I was wanting to do some rough frequency response characterization of some microphone / preamp / soundcard combinations. I am not worried about phase at all just relative frequency responses. Ideally I want to characterize it between 5Khz and 24Khz, but would be happy with 10K to 20K.

I plan on getting a speaker with a relatively flat response from 10K to 20K playing white noise, measure the frequency response using REW RTA with a UMM-6 microphone. I would then repeat this process with a microphone on a sound card, take the difference (division?) of the results using REW and have a frequency response curve.

I expect it will have some wild results at the very top end, but will this work? Is there a better way to do this? Should I be using stepped sine instead?

 

[Wayne A. Pflughaupt]

Don’t see any reason why the idea wouldn’t work, as long as your UMM-6 has a calibration file. You’ll have your baseline measurement with the calibrated mic and flat speaker, so it’ll be able to identify any differences subsequent equipment displays

For the record, the calibration file that one that comes with the UMM-6 mic is trash. You’ll want a calibration from a lab like Cross Spectrum if you need a truly flat baseline reference.

Regards,
Wayne