At such high frequencies the wavelengths are so short that placement of a transducer is not very reliable, and the perceived frequency response can vary from trail to trail. If you're trying to use speaker in a no-anachoic environment, minute head movements can vary the results. In psychoacoutic testing we trusted only forced-choice tests in which the listener tries to identify which interval contained the test tone over many trials. Low level hearing is a stochastic process, so the concept of threshold is just that. And then there's the problem with calibration . . .
And what difference does it make anyway? Why not just listen to the audio system of interest?
db
And what difference does it make anyway? Why not just listen to the audio system of interest?
db