Aloha l and welcome to the board
But can human beings ‘feel’ it?
{‘-’}
No.
If you don’t have an organ of perception, there is nothing you can sense (= feel).
Being able to “feel” (physically, in terms of kinesthetic sensations, if this is what you mean) frequencies higher than 20 kHz would imply such an insane amount of energy that it would kill you instantly.
Some insects and animals don’t have ears, yet they can sense the vibrations by other means.
I’m not saying that humans DO have such abilities to ‘feel’ higher frequencies (though lower frequencies are felt), but I wouldn’t be as arrogant to claim there are none.
By the way, what happened to your sentence with ‘science and logic’? I had whole lot of things lined up to deal with your lack of that!
Patanjali, as long as there is no hard evidence to backup any claims of higher sample rates being useful, everybody should refuse to waste 50% or more of his CPU power just for some unsubstantiated hearsay and esoteric waffle, seriously.
I need to see significant differences in perception, under strictly controlled circumstances (double blinding, open source and heavily examined LPF design, etc…) between 44,1 kHz and 192 kHz recordings to even CONSIDER those claims being anything more than esoteric mumbo-jumbo.
If it would not cost CPU cycles (and memory, but this is a non-issue nowadays), I wouldn’t be such a nitpicker, but people pay through their noses for moar CPU power (and have to utilize awkward metatechniques such as “freezing”, which is a horrible concept in itself) for exactly NOTHING.
Please understand that I really appreciate your strife for better, more beautiful and clearer sound. I’m all for that and I try to do the same (mostly by learning, but also by getting good equipment, such as the SPL Gainstation 1 I just ordered) - but some efforts are not only futile but even counterproductive.
Believe me, I really like high end equipment (SPL stuff) and I use stuff like Kramer Master Tape and the Slate VBC and all that, which add nuances at most - but I would never even consider going > 48 kHz (and I do 48 kHz only because KORG forces me to).
Science simply doesn’t back any higher sample rates (except in select mathematical processes, this is why some plugins perform upsampling, which is good and necessary and great, but this is about mathematics only, algebraic necessities and all that).
And, here…
I would like to see those as well, but there needs to be some discussion about how that may be achieved. Such things don’t have to be the province of proprietary research or musings by learned experts in obscure forums. Such collaboration can be had by any group interested in knowing the truth.
However, having open discussions about such things is very difficult when you just use dismissive and erroneous arguments to ruin getting to anywhere near such a forum. Those tactics are the enemies of science and engineering.
Suppression, ignorance and choosing the least confronting options are not what the peaks of human achievement are made of. I choose NOT to be mediocre!
Right, Patanjali, and I’m with you there, but this discussion about higher sample rates strongly remind me of “EUR 5,000.- audio cables” and “Argentum nitricum D12”.
It’s not that I would not be open to meaningful evidence, quite the opposite, but it’s basic logic which dictates for me that it’s pointless to go higher than something like 44.1 kHz of sampling rate. This opinion is based on where science stands. We would actually need different mathematics and physics (or biology, if you want to debate human ability to perceive beyond 20 kHz!), not just updates or errata.
There is a difference between “reason allows for something to be valid, even if the probability is low” (if this was the case here, I’d all be “go 96 kHz, just to be on the safe side, should science find out that there IS something”!) and “this is utter nonsense and has nothing to do with reality at all” (which is what is the case here - given high quality LPFs… I totally agree that low quality LPFs may cause horrible differences in sample rates, but this has nothing to do with “Nyquist being invalid” or something similar).
I’d rather investigate IF and WHY there are filter differences between sample rates in some cases… I’m not one of the tinfoil hat guys, but there may be economic interest behind that. 
I’m following this with interest, though I confess to not understanding the mathmatical stuff in here.
That said it’s interesting to note that it’s been shown that humans are affected by “sound” outside the limits of hearing, of course, we don’t perceive it as sound.
[sounds] above the human audible range (max. 20 kHz) activate the midbrain and diencephalon and evoke various physiological, psychological and behavioral responses.
Humans can be affected by “sounds” outside our hearing limits, like in ultrasound treatment (in medical equipment). BUT…
…But to do that we have to bandlimit this sound waves to the upper frequencies only, focusing the energy needed in that frequency band only.
…But within a musical context (I guess we are doing music here), we cannot feel anything up in these frequencies above 20kHz. Because if we provided that amout of energy to play the higher freq to be “felt through” within a musical context, the amount of power provided in the hearing range would have killed us.
…Besides that we have no amplifier to provide such power in a full frequency playback (remember we talk about within a musical context).
Of corse there are frequencies above the human hearing range, as in ultrasound and Roentgen to take two different frequency bands that excist (above our hearing limits). Of corse they excist (nobody has denied that), but not at all present at a “feelable” level within any form of musical context.
So if you want to discuss medical equipment, I suggest another forum
Again, take a look at the Fletcher-Munson curves (phon curves). How much dBSPL (dB Sound Preassure Level) would you have to provide to make a 48kHz signal loud enough, to even get close to “feelable”?
And then switch of the High-pass Filter…yes sir…what did you say?
Please admit that what we hear, HAS TO BE within the human hearing range. And please discuss this Sample rate debate within a musical (full frequency) context. Please don’t make this harder then it is.
Have nice weekend 
Indeed, BUT guess what happens when we take a waform of only limited time … we’ll get DISTORTION. This distortion is greatest at the very beginning and very end of the waveform.
Now let me give you a homework: calculate how many samples it takes in worst case (from beginning or end of the file) to have this distortion created by time-chopping to be lower than quatisation distortion (you can choose the sampling parameters yourself). Or you may also choose to calculate how many samples it takes to make this distortion less than … let’s say … 1%. That would be more like a real-world problem.
BTW, we DAW users all should be very familiar on problems using Shannon-Nyqvist where it’s not applicable: those clicks what we get when slicing and combining waveforms without using appropriate methods.
OK, OK folks … I have a quick solution: let’s all make an appointment with our local audiologist, and after we’ve had our hearing scientifically measured, those who can hear above 20kHz can set their interfaces to the higher sample rates, buy more expensive gear and bigger disks – oh, and buy shares in PonoMusic, because it’s clearly so much better, people will be queueing up to buy!
Now, where did I leave my gold-plated speaker cables … oh, there they are, plugged into my tube amplifier, beside the record deck …
I think that would be more like a first-world problem.
if we can’t hear a dog whistle, but when we blow on it our dog barks and we hear it… the same goes for frequencies that fall outside of our hearing range. They affect what is in our hearing range. I did a test a few years back and I posted somewhere about it. You can clearly see a node created in our hearing range buy something outside our hearing range.
From a result-oriented POV 44/24 is absolutely fine. I don’t negate effects of higher samplerates but would consider them as pretty much meaningless for myself. We are flooded with pure luxury today, aren’t we?
Doesn’t the magic of music happen in complete different spheres anyway? The idea, the talent of performers as well as engineers have much more impact in an audio world that has removed so many limitations we had to face just 10 or 15 years ago than hypothetical considerations on sample rates - which seem to remain emotional/subjective in this as well as in previous discussions about the same topic.
Fetishizing about things like sample rates can provide a good excuse for lack of ability or creativity.
Are you referring to the guard bands, ADDED to the front and end of a block of samples being processed, that ramp up and down from zero, and designed to maintain bandwidth limiting and pretend that the block is doubly infinite?
If so, technically there is no distortion at the front and end of the actual band, but processing it as such would treat it like a square wave (where not zero), which would generate frequency components well beyond the bandwidth to which the signal is required (according to the Nyquist et al theorem) to be constrained. Careful design of the guard band slope patterns would keep bandwidth, but would also add to the coefficient values for each of the conversion frequencies, which presumably would have to be factored out before using the coefficients in any calculations. The people that come up with this stuff are SMART!
I am not familiar with all the maths, so maybe you can stop being a smart-a** and let us all know THE ANSWER.
HAHA, you have a point. In fact most of this geeky stuff is pretty much always the least of the problems associated with recorded music.
DG
Yes, just listen to some of those 192kHz advocates music present in this …
And I suspect Dan Lavry, who makes some of the better converters out there, know a thing or two about digital audio, Sample Rates and Bits.
I think I can relate to, and learn more from Dan Lavry, Bob Katz (sorry you ended on this list ), Ethan Winer, Nika Aldrich, Thomas Lund - TC.Electronic, and a few others, than some mediocre alternative “medicine” man.
PS. What those 96/192kHz advocates should be more concerned about (than what they think they can hear at 40kHz, at 120 dB below zero), is intermodulation distortion. Which certainly can force its way into our audible range.
Since analog circuits are almost never linear at super-high frequencies, they will introduce a special type of distortion called intermodulation distortion.
This means that two super-sonic frequencies that cannot be heard, say 22 kHz and 32 kHz, can create an intermodulation distortion down in the audible range, in this case at the “difference frequency” of 10kHz. This is a real danger whenever super-sonic frequencies are not filtered out.
But what do those above mentioned know about digital audio, sample rates and bits etc… ?
You obviously cannot read very well! In this thread (here), I have cited Bob Katz’s recommendation in Paul Gilreath’s The Guide to MIDI Orchestration he recommends recording at 96k, or upsampling to that as soon as possible, and keeping all processing at that until ready to downsample at the very end to the target deliverable sample rate. That was in 2006, a few years after Mastering Audio (2002).
It appears that you have no real idea of where Bob Katz stands on sample rates, so I wonder what else you have misunderstood about the others you cite.
It is appealing to false authority to cite people based on the respect others show them, when you cite them in error. That is disrespectful to those whom you cite, those to whom you cite them, and, of course, to yourself, to whom you do the greatest disservice!
Of course, Bob may have changed his mind over the years, but since I had already cited (in the course of THIS discussion) a stance he took at one time, it was your obligation to explicitly cite his changed stance, if there is any.
Anyway! 
While it’s ‘nice’ for tools to have ‘capability’ to operate at higher sample rates, I think it’s a waste, and won’t ever bother to record higher than 44.1 KHz. Besides, the end listener isn’t going to be listening at 96 KHz. The talent, performance, vibe and especially the song is far more important than recording at ridiculously high sample rates.
I remember the re-birth of Aerosmith in the late 80’s/early 90’s…their newer higher rate recordings were so much clearer! But you know, I liked their older, duller, and warmer sound MUCH better!
Well, geneally speaking these above are more trustable sources than both you and me. He kind of represent some of the professionals out there on digital audio in general, and I have read more of their papers than just that book.
Sorry if I didn’t call in for all those spokesmen last opinions, and that I rely on their official documentation. Sorry to Bob Katz if his name falsely got on my list in the hurry
That said, his statement in your thread may has something to do with midi instruments and electronically generated sounds (never been performed before a microphone, through pre-amps etc), where you have a new set of challenges.
You don’t have to defend your self to me, so keep on recording those intermodulation distortions. It may just be the distortion you prefer. What do I know?
But please don’t tell me that you can hear over 20kHz. Please don’t tell me that you have a set of ears unknown to man.
Where did you take your sound engineer theory education/excam?
I finished mine in 1997 at NISS, the biggest audio engineering school in Oslo Norway (after recording for about 18 years). Prior to that I have had practical multitrack courses +++ at the same school.
Why are you better than Dan Lavry on Digital Sampling Theory (and therby converter design)?
You gotta have some sort of education?
Do you have an education on the Human Auditory System?
I joined a Tinnitus research group (6 months) at Norways biggest hospital due to my mothers involvement, where a good amount of Auditory doctors also attended (I then found it appropriate with some questions - for learning).
In what profession do you have your degree?
So if you are after me, I have my ground covered. I don’t worry about that.
We hear what we wanna hear. Nothing I can do about that either. But…
…But please discuss this subject in a musical context, as in the practical use of pro audio equipment in recording.
You can question Nyquists Theorem as much as you like, but we are never the less talkin about bandlimited signals in analog to digital conversion. Nobody has disaproved this since 1947, neither have you.
Keep on doing whatever you’re doing. So will I. Have a nice weekend