So I’ve listened to all the arguments from people who don’t understand the Nyquist theorem for why audio higher than 44khz doesn’t actually matter and you can’t hear it bla bla bla. From literal decades of personal experience of hearing the difference from the production side and knowing that from a physics perspective that it’s just not true, I present objective evidence that you can hear frequencies above 20khz.
First: a sample of a track I’m currently mixing/mastering
https://drive.google.com/file/d/1WvSAkDAlV4g0joqmlgJcVSDKrLBq3bkO/view?usp=sharing
Second: the same exact sample at the same exact volume with a 22khz tone applied.
https://drive.google.com/file/d/1LK6n5zd6QsrzcfW9n967NPsAzp2BRaA8/view?usp=sharing
If you can hear the difference (spoiler alert: you can), then you objectively can hear frequencies above 20hkz and by extension you must necessarily concede that there is a point to having waveforms capable of representing higher frequencies.
Lookup the wiki on trigonometric identities and see what happens when 2 sine waves get multiplied together. What you get is called modulation products. For example, if you have 2 sine waves at different frequencies - say 95 kHz and 96 kHz - and you multiply them together, you get 2 totally different frequencies as the outcome. One will be the sum of the two input frequencies or 191 kHz - I’ll go out on a limb and posit that none of us mere mortals would be able to hear that. The other frequency would be the difference between the two input frequencies or 1 kHz - which at least most of us can hear.
So, OP, is it possible that whatever Jedi mind trick you are doing with 96 kHz is leading to a modulation product that lands within the audible range?