I see, Shannon-Nyquist is guaranteeing frequency sampling process to be reversible.Well that's the trick. If you are not able to trust the demonstration (not just the words) then you'r only recourse is to get your maths to the poiint where you can follow the maths of shannon-nyquist.
Is what I guess in my last answer, the subspace of signals on the band limited space is the only considered to be possible to reconstruct. So nice, Shanon-Nyquist, beautiful theorem.
This leads more to physics than to maths, in the sense we suppose physicals signals to be band limited simple frequency composed.
I suppose this is true, or conversely engineers wouldn’t use it. In the context I studied at the faculty, the set of functions being considered was not constrained to this requirement.
So quantization variable is the only non reversible transfer function. I saw also the harmonic distortion added of quantization and need for sacrificing SNR adding dither noise to avoid distortion on some frequencies.
Now I understand the maths, but what in the hell is the reason I hear better on the high digital volume and analogue attenuation?