That's not correct either. 44,100 samples per second vs 96,000 samples per second is a measure of resolution, where frequency _is_ in the time domain. The Nyquist limit rises in addition to and in proportion with the sample rate, it is not the *only* difference.