The
ASR review shows up to 84kHz with a sample rate of 192kHz.
In case you don't know this, you are limited (by "math") to half of the sample rate (Nyquist theory). This is logical since you need at-least one sample for the positive-half of the waveform and another sample for the negative-half, So, CDs (at 44.1kHz) can't contain anything over 22,050Hz. In practice there are (imperfect) filters so you can't go all the way to 22,050 with a CD. If you "try" to go higher (without properly filtering) you get aliasing... Additional lower-frequencies that weren't in the original.
At higher sampling rates you don't
always get higher audio capability because audio interfaces are made for audio, and ultrasonics may be filtered out just as "good practice", or so they can use the same filter for all sample rates.
After listening to quite a few recordings I convinced myself that the same recording with frequencies exceeding 20kHz filtered out sounds worse.
Either there is "something wrong" with your filter, or maybe you are fooling yourself. It's easy to fool yourself unless you do a proper, level-matched, scientific,
ABX Test.
It turns out that even if you can hear up to 20KHz in a hearing test with "loud" test tones, when you're listening to music, the highest frequencies (which are harmonics, not "musical notes") are weak and masked (drowned-out) by stronger and slightly lower frequencies.
And as long as there isn't some kind of unusual intermodulation distortion, the ultrasonic frequencies have no effect on what you're actually hearing down in the audio range.
P.S.
Filter cut-off is defined as the -3dB point, no matter how sharp the filter. So a 20kHz low-pass filter is 3dB
down at 20kHz and a 15kHz filter is 3dB down at 15kHz. If you want flat audio at 20kHz a low-pass filter needs a higher cutoff frequency,