Oversampling on the playback end meaning the DAC side is pointless because there is no new information between the samples. Just interpolation. Now over-sampling on the recording end meaning the ADC side makes total sense with at least 8x the Nyquist bandwidth given 2 samples per cycle at the Nyquist limit. At that oversampling bandwidth point you can show near perfect waveform reconstruction.

Anti-aliasing filters for 44.1kHz is at the Nyquist limit for a 20kHz analog waveform signal. So the filter is very restricted in design. My simulations show that you cannot make an IIR filter work however an FIR filter does work well for best faithful reconstruction of the input signal.

That leaves the last question which is the optimal order for the FIR filter as the order is determined by the number of series cascading of first-order filters. Now that I don’t really know which order is optimal because it is a balance of tolerating how much aliasing error you allow against impulse response and delay, damping factor, and the real bandwidth contained in musical signal.

It is strange that some designers claim no digital or analog filters in the playback system and also no aliasing error for a 44.1kHz source sampling.
Did they get the recording studio to record the CD with a 4kHz analog LPF at the microphone?

Maybe this is helpful to you?

https://www.facebook.com/share/p/1LqGkfb9QY/

https://www.facebook.com/share/v/1GTRGQznME/

It works on Windows, macOS and Linux. It is free and open source.

https://friture.org/index.html?fbclid=Iwb21leAO52BljbGNrA7nYFmV4dG4DYWVtAjExAHNydGMGYXBwX2lkDDM1MDY4NTUzMTcyOAABHsDYHbvngj6kTMgB281Sb4rn0um0eWhTE0ApU0mgNbXfMRJxeXY-6xqEYXUA_aem_YVAGa_9z6yolvLJAzu9vRw

Has anyone any thoughts on this or other recommendations?
https://www.facebook.com/share/p/1DUgYTheDZ/