That's not what Shannon's theorem says. It doesn't state that sampling at twice the frequency is enough to be able to reproduce a perfect signal. What it states is that if you sample under twice the frequency, there is no way you will be able to reproduce a signal at that frequency... It's a minimum, not a maximum.
You don't even know what Nyquist-Shannon says at all. Here's a quote relevant to what you're talking about that 100% contradicts you.
"The Shannon sampling theory for non-uniform sampling states that a band-limited signal can be perfectly reconstructed from its samples if the average sampling rate satisfies the Nyquist condition. Therefore, although uniformly spaced samples may result in easier reconstruction algorithms, it is not a necessary condition for perfect reconstruction." --Nonuniform Sampling, Theory and Practice (ed. F. Marvasti), Kluwer Academic/Plenum Publishers, New York, 2000
In other words, it is a MAXIMUM requirement as well as a MINIMUM one. All frequencies can be PERFECTLY reconstructed if the sampling rate is twice the recorded frequency in a bandwidth-limited signal. In other words, what you said is 100% incorrect.
96kHz is pointless in terms of reproducing audio that a HUMAN BEING can hear and that is a scientific
FACT, not an opinion! The filtering issues are easily overcome in the present day, BTW. All this "HD Audio" is pure nonsense on that front.
However, recording audio at higher rates (bit-rates at least) does have real benefits in terms of head-room (i.e. avoiding clipping with unpredictable live sources). But this can easily be reduced to a lower number on the playback side with ZERO ill effect. It's as simple as using a normalizing filter to set the absolute levels and then dithering down to a lower bit-rate. You can even use noise-shaping and get near 18-bit audio out of a 16-bit CD if it really worries you.
Thus 16-bit audio may be a bit low in absolute terms for the best recordings out there. 18-bit would have been a better choice, but there's zero reason to EVER have >20-bit audio on the playback side, atl east that is meant to be heard by humans. You can only hear 120dB MAXIMUM dynamic range and 120dB is very VERY bad for your hearing (as in instant damage). Just because you CAN hear it, doesn't mean you SHOULD. Thus, in reality, almost no recordings have a usable realistic amount of dynamic range that would matter much over 16-bit and virtually NONE over 18-bit on the playback end.
For instance, a young person will hear a signal at 20kHz. To capture that signal, you need to sample at at least 40kHz. Then, that young person will be able to hear something, but you will have lost a lot of characteristics of the signal - for instance, you will not be able to know if the original signal was a sawtooth, a square or a sinusoid. So, significant information will have been lost.
WRONG WRONG WRONG. Go read about digital audio before you spread any more falsehoods. I get so SICK of reading this false information on the Internet. Go read some articles on digital audio. Try looking under OVERSAMPLING an FILTERING to start. What you are saying only holds true if you don't use a reconstruction filter (and there are some DACS out there now that do that since some crazy people actually think it sounds better to take perfect sine waves and turn them into crap, apparently). Playing back digital audio without a reconstruction filter (or not using something like oversampling on the recording side) is a bit like buying a car without a muffler or exhaust and then complaining that it's NOISY. Well DUH!
If you oversample when you record then filter the frequencies above 20kHz and use dithering (white noise), you will eliminate aliasing errors and all problems near the noise floor. If you then use a proper reconstruction filter on the playback side (often filtered in reverse at oversampling rates internally first or using other standards like Sony's 1-bit filters), you remove all the problems on the playback end as well. You can easily get a near perfect sine wave at 20kHz off a CD using this method. It's not difficult.
That's why CD recordings sounded metallic at first. The solution, which is
It never sounded metallic except on bad recordings or perhaps the very first CD players that used brick filtering and no oversampling. Anything else is a MASTERING problem. And there were and are today still a TON of mastering issues ranging from early on using LP masters that were optimized for a medium that couldn't handle stereo bass and could burn out cutting heads with too much treble to modern day "loudness wars" that assume "louder = better" when in reality, it just creates so much distortion (see the Red Hot Chili Peppers
Californication album as a good example of bad mastering to the point of CLIPPING all over the place). Why is it harsh sounding? It has almost zero dynamic range and is clipping square waves all over the place! Is that digital's fault? No, it's the mastering engineer's fault and possibly the band's fault as well (the pre-final master I have here is still pretty darn hot, but at least it's not clipping left and right).
applied on all CDs, was to cut the frequency around 16kHz to avoid the destruction of the characteristics of the signal around Shannon frequencies.
Where do you get this false information? LP masters are bandwidth limited purposely to avoid burning out the cutting heads. That is the only reason (other than crazy audiophile labels that believe this crap) it would be bandwidth limited. I've got news for you, though. There's hardly any usable (or audible) frequencies above 16kHz regardless. Your
average adult over 30 years old isn't likely to hear even a loud sine wave above 16kHz. I don't think they notice with music.
That's why 96kHz is interesting, because it keeps quality in the upper part of the spectrum.
Quality? For noise? MOST of the information above 20kHz is NOISE. It takes a lot more storage space to represent that inaudible (to humans) NOISE. How is that interesting?
Moreover, 24-96 is not only about 96kHz, it's also 24 bit. And there, you gain a lot. The problem with CD and digital capture in general is that the scale is linear while most of our senses use a logarithmic scale.
In case you don't understand physics (and I'm pretty certain you do not or you wouldn't keep making such statements), a SCALE is a measuring device used by humans. I can switch between linear and logarithmic scales on my scope. Switching mathematical measurement scales does not affect CONTENT ONE IOTA. In other words, digital audio reproduces bandwidth. That bandwidth can be measured in linear or logarithmic scales. There is no difference. The music content is the SAME. The fact you don't know this and think digital is somehow different than analog shows me you're spouting crap you don't understand in the slightest.
The result is that when you go at the bottom of your intensity, you have a very very low resolution in your sample, while the human ear (or eye) still have a good resolution. This is especially visible in photography: if you brighten the shadows, you will see a lot of banding, because the sample resolution is very low in the shadows. It's the same problem with audio: CD killed the dynamic range (hence the loudness war), because it's not that good when you have a lot of dynamic during the low volume ports.
Please don't compare video with audio. The so-called loudness wars are a mastering and recording issue. They have literally nothing to do with any limitation in the CD digital format, real or imaginary. "CD" did not "kill" anything. People killed dynamic range in the mastering booth, usually on hard drives these days. CDs are created later. Using 24/96kHz with such a mastered album would only get you a disc that contains more stored bits, but sounds every bit as bad. In fact, such an album could probably be accurately reproduced by a 6-bit 44kHz format because I doubt most of those albums contain even that much dynamic range.