I won't pretend to be an expert in some of the technical parts you wrote about, but I can offer what I think is sort of an answer to your question.
 
Does the bit depth and sample rate matter?  Yes, but mostly if you're playing content recorded at a higher bit-depth/sample rate than you currently have the NU Audio card set to.  The other times it objectively matters is sometimes when you're working on a DAW (digital audio workstation), and you're working on a track recorded at a certain sample rate, but you want to work on it in a sample rate that is not a direct multiple (e.g. a source file at 44.1KHz, but you want to work at 192KHz).  You can sometimes run into audio issues due to the conversion.
 
Whether there's a positive difference by upsampling audio or even just setting your bit depth/sample rate higher than your source content will be a debate that will probably never end.  The technical side of it says that there is no difference than if you played the same source file at its native bit-depth/sample rate, whereas some audiophiles believe that they can subjectively hear more detail or more content when upsampling or setting the bit-depth/sample rate to the max that their audio device can support.  I've read enough, listened to enough, and seen enough that I know there's no way I can prove to the latter people that they're only hearing what they want to hear.  At the same time, there's no way that I can be sure that they aren't hearing what they say they're hearing. 
 
From my understanding of how bit-depth/sample rate works when the source is a lower depth/rate is that it's a lot more like the audio card has the potential to process content at the set depth/rate, but the excess of that potential goes unused.  A higher sample rate is more useful, since the cycles are much faster and you will be less likely to lose any content, but unless there's some deficiency when the depth/sample rate is set to match the source's values, there should be no actual improvement or change.
 
In other words, bit-depth/sample rate isn't a matter of upsampling anything.  It can't read what isn't there.  Some people claim that setting the bit-depth/sample rate adds some effect to the audio, but if it does, it's not some data magically hidden in the track.  If the bit-depth and/or sample rate is too low, on the other hand, it just means that some detail will be lost.  You can experiment to see if you notice a difference, but most people won't.
 
Going down the path of audio as vibrations and recording performance is probably the wrong angle to take here.  Music (or audio, in general) is recorded and stored digitally via algorithms to keep it as close to the original source as possible, while maintaining the ability to convert back to an analog signal.  This is why audio has the popular phrase of "garbage in, garbage out."  There's nothing that an audio card should do to "improve" the source.  Essentially, the question is more along the lines of what the audio card can do to prevent the source from degrading or acquiring artifacts as it processes the file from analog to digital, or from digital to analog.  Along the way, the various components will probably have some effect due to minimizing electronic interference, preventing signal distortion, and protecting the data path from the digital source through conversion and as it leaves as an analog signal.
 
Moreover, although you are correct that audio in real-life does not have those limitations, we aren't really talking about real-life audio - we're talking about recorded/created and stored audio.  That audio has already happened, it has occurred, and it has been preserved.  As a part of its storage process, it's subject to other digital limitations, such as data retrieval and playback.  Similarly, even vinyl requires certain playback speeds, depending on the disc.
 
This is why you get a number of different crowds when you talk about things like upsampling.  Upsampling doesn't objectively improve the quality, since it typically requires an algorithm to fill in the blanks of the things that are missing.  It's why AA support on graphics cards can only do so much to fix tile limitations in games.  On the flip side, NVIDIA DSR takes a high resolution and downsamples it to your current resolution, resulting in a better picture quality, but the overall quality depends on the source video files.  DSR is the opposite of what you're suggesting, however, since DSR deals with more data being scaled down to a smaller resolution - compression, essentially. 
 
Upsampling (in theory) takes the same amount of data and tries to fit it onto a larger canvas.  In the case of video, it basically takes the edges and stretches them to the now wider corners, and tries to fill in the gaps via an algorithm.  I'm sure you've seen the generally poor quality when you stretch an SD signal to HD to avoid the black bars; hint: it's not great.  Another example is when you try to use a desktop background with a wallpaper designed for smaller resolutions; it stretches and becomes distorted. 
 
With audio, is that even necessary?  DACs are not designed to do this.  DACs are not designed to add data to an existing source, and it would have no real way to do this in practice.  It's the same if you take a CD recording and try to upsample it to DSD or a higher bit-rate/sample rate.  The audio data stays the same, but there's a lot of extra padding in the empty space.  Some people claim this is still an improvement, while the technical testing suggests there's no difference, other than a higher file size.  Maybe the more salient question is - do you want a hardware component or software solution to alter the audio to add something that wasn't originally there?
 

usa423
Many years ago I noticed a phenomena with digital audio that if I began a project as 24 bit 96 khz to start, then ended in 16 bit 44khz. It sounded better to me then just starting in 16 bit 44 khz and ending as such. There might be a technical way to explain this by some means. I don't know.

 
Right, because you're talking about downsampling/compression.  Assuming your sources were at 24/96, you're starting at a higher quality, then later compressing it to 16/44.  Generally speaking, that's how it should generally work.  There will be more data in the 24/96 source to compress down to 16/44.  The problem is that most of what you're talking about is upsampling - starting from a lower "quality" bit-depth/sample rate and trying to "upgrade" the audio source by changing the bit-depth/sample rate.
 

usa423
But my ultimate question here today is something in the same ball park as the above statement. If the Nu-Audio actually has 32 bit 384 khz playback capability, but applications like I dunno...Dolby Atmos for headphones,.....force me to use 16 bit 44khz, is there still a benefit to using the Nu-Audio over just any other PC audio solution that is limited to just 16 bit 44-48 khz??????????????????????

I think this is the crux of what you're trying to ask, and I think there are a couple fallacies here.
 
First, you're equating bit-depth/sample rate directly to a certain level of quality in the overall source.  If you have a terrible quality audio source recorded at 32/384, it won't sound better than a good 16/44 recording.  If the source is capable of being recorded at both 32/384 and 16/44 to compare results, that's a different matter.  Ultimately, the quality of the audio source is the overall factor when it comes to the quality of the audio.
 
Second, by asking if there is a benefit to using the NU Audio card over another (but really, you could ask that about any card), it seems to presume that 16/44 is such a low quality that you cannot hear a difference between audio cards.  16/44, as I'm sure you know, is CD quality audio, and a lossless rip should easily allow you to tell the difference between audio cards.  Same thing with 16/48, which is DVD quality.  Likewise, games typically do not exceed recordings of 16/48, so you are not losing anything by using a spatial option that limits it to 16/48. 
 
That brings us ultimately to the last point.  Yes, there's a difference with the NU Audio Card, even at low bit-depth and sample rates.  Through my own testing of competitor cards, there is a noticeable difference at 16/44 and 16/48, but it's also the same difference at 24/192 and higher.  The overall audio signature is consistent for each audio solution, so the differences should remain the same.  The bit-depth and sample rate of a source do not have the overall impact that you seem to believe it does when it comes to the overall audio quality.  As noted, I won't pretend to know what audio sounds like to you under those circumstances, but I can advise you based on my technical understanding of these specs.

EVGATech_LeeMI think this is the crux of what you're trying to ask, and I think there are a couple fallacies here.
 
First, you're equating bit-depth/sample rate directly to a certain level of quality in the overall source.  If you have a terrible quality audio source recorded at 32/384, it won't sound better than a good 16/44 recording.  If the source is capable of being recorded at both 32/384 and 16/44 to compare results, that's a different matter.  Ultimately, the quality of the audio source is the overall factor when it comes to the quality of the audio.
 
Second, by asking if there is a benefit to using the NU Audio card over another (but really, you could ask that about any card), it seems to presume that 16/44 is such a low quality that you cannot hear a difference between audio cards.  16/44, as I'm sure you know, is CD quality audio, and a lossless rip should easily allow you to tell the difference between audio cards.  Same thing with 16/48, which is DVD quality.  Likewise, games typically do not exceed recordings of 16/48, so you are not losing anything by using a spatial option that limits it to 16/48. 
 
That brings us ultimately to the last point.  Yes, there's a difference with the NU Audio Card, even at low bit-depth and sample rates.  Through my own testing of competitor cards, there is a noticeable difference at 16/44 and 16/48, but it's also the same difference at 24/192 and higher.  The overall audio signature is consistent for each audio solution, so the differences should remain the same.  The bit-depth and sample rate of a source do not have the overall impact that you seem to believe it does when it comes to the overall audio quality.  As noted, I won't pretend to know what audio sounds like to you under those circumstances, but I can advise you based on my technical understanding of these specs.

Thanks Lee for taking the time to reply to this and offering a very well thought out reply. I'm not necessarily equating the bit-depth / sample rate to the quality of the source material. But for example, lets say we had a special room, a really nice piano, and a really good piano player. To capture this signal, we have 1 mic, a Neuman U87, fed into a Mackie mixing board with the 48V phantom power turned to power the condensor mic. Out of the mixer we go straight into the line in of a Nu-Audio sound card.
 
Now we have a choice of what bit depth and sample rate to use. Anything from 16 bit 44.1 KHZ up to 32 bit 384 khz. The higher the bit depth and sample rate the larger the recorded files will be. But regardless of any opinions of perceived quality changes by anyone for the various bit depths and sample rates, it has to logically be capturing more audio data. Data that is actually real to the original analog electrical signal. No parlor tricks. So would that not be kind of like watching a movie really shot in 3D like Avatar VS a movie that is made 3D in post like Jurassic Park? Or a real 4K blu ray VS an upconverted 1080p blu ray?  Sorry I keep comparing digital audio to video, its just that the behavior of waves in a pond have often been used in teaching about sound, electricity, and radio. As its something we can see visually and then consider the behavior of things we cannot see visually.
 
I've also had it debated to me before that there is a quality difference in the ADCs and DACs themselves that can yield different perceived qualites even when all are the same bit depth and sample rate like 16 bit 44.1KHZ cd standard.

ty_ger07
Here you go, this should help a lot:



My opinion is that a sound card is a waste of money and it is mostly placebo.

The math guarantees that at "only" a 44.1 kHz sample rate, every human audible frequency is perfectly reproduced using even the most basic audio gear. For decades, it has been easy to buy a simple, cheap stereo system with a DAC which perfectly reproduces the sound exactly as the artist intended. Why do we have $100 sound cards, $200 sound cards, and $1000 stereo systems? In my opinion, it's because some people just want to believe that they can spend more money to get something exceptional. In reality, in my opinion, it's all smoke and mirrors and all you are paying for is the ability to apply hardware and software filters to alter the sound as you prefer against the artist's original wishes. I buy this card, I buy this op amp, I buy this headset, so on and so forth, I subtract the left and right audio by 25%, and so on, and now the sound is brighter. Great.... but is that because the $25 version is inferior, or is it just because you wanted to pay more money to feel like a unique individual?

Thanks for posting the video.
 
The math guarantees it? But what makes you think the issue here (my personal issue) is about frequency range or response? What about individual user perceivability?
 
 
"For decades, it has been easy to buy a simple, cheap stereo system with a DAC which perfectly reproduces the sound exactly as the artist intended"
 
That is your opinion, its not a fact, as I don't think that comment is truly 100 percent accurate, and again goes back to individual user perceivability. There is non alcholic beer that taste the same going in, but it doesn't get you drunk though.
 
 
"Why do we have $100 sound cards, $200 sound cards, and $1000 stereo systems? In my opinion, it's because some people just want to believe that they can spend more money to get something exceptional."
 
I know what you mean. There is lots of marketing and voodoo speculation in the world of audio. Like a pair of speakers that are 20,000 + dollars. But I do not think that is the case here. This sound card offers abilities that no other desktop mainstream PC audio solution has for recording. Not that I'm aware of. Feel free to link me to any product that has 32 bit 384khz recording ability available to the public and @ or lower than EVGA's price.
 
"but is that because the $25 version is inferior, or is it just because you wanted to pay more money to feel like a unique individual?"
 
First off, I am a unique individual, I'm not trying anything.
 
And well, I wanted to buy the best available recording an playback sound card or interface that at least has the ABILITY to use said rates on both recording and playback. Doesn't mean you have to actually use that. If a car has a 4 speed manual transmission then you are not going have a 5th gear to use. But if you had the ability, you might get by with just 4 gears, but that 5th gear is there in case in the future it becomes more mainstream. As was pointed out in this thread, actually having things that are natively recorded at such high rates usually yields much better results than things first recorded at lower settings.
 
Many say they don't care, nor hear a difference. But while I think its possible for anyone to fail a taste test (acute), I personally think their opinions would change if they got to spend a decent amount of time with said products over weeks and months (chronic). But I didn't honestly care who made it or made it first. It could have been Creative Labs, or someone else. I was actually quite surprised that a popular video card maker would be the first to do it and offer it to the market.
 
 

usa423

ty_ger07

ty_ger07

Again, keep in mind that can't hear over 20 kHz. As explained in the video above, all sounds that you can hear can be perfectly captured and 100% reproduced at 44.1 kHz. Higher frequency recordings are litterally needlessly larger file sizes and offer decreased compatibility for your audience. If you want to be an elitist about it and purposely limit your audience size for no good reason, go right ahead.

Not being elite about anything. I appreciate saving money whenever I can, I'm not rich, not even wealthy honestly. There is quite a big difference between this argument and 20,000 dollar house speakers IMO.

"Again, keep in mind that can't hear over 20 kHz"

Can you explain who came up with that number for hearing ability of a human being?  I wasn't aware Nyquist was credited for discovery of the frequency range of the human ear. But that is really besides the point here. As far as I know, sound waves in real life are not a series of sequentially played samples.

Larger file sizes for nothing? Is that your opinion or do you think that is a fact?

Can you hear above 20 kHz? No. Just as your eyes cannot see above and below certain wavelengths, your ears cannot hear above and below certain wavelengths. It's science.  It's physics.  It's fact. There is an overwhelming amount of evidence on the subject. I can't believe that you want to argue this fact. It's not debatable.

Ok, that being factual, the technology and math guarantees that at 2 samples per 20 kHz wavelength, all human perceivable sound can be recorded and perfectly reproduced at "just" a 44.1 kHz sample rate. Did you even watch the video?

Yes, larger file sizes for nothing. It is my opinion. But it is also fact -- would have to be true -- if you first accept the notion that 44.1 kHz can perfectly record and reproduce all human audible sound.
 
You talked about "ripples in a pond of sound" which are being lost at 44.1 kHz sample rate.  Yes, sure.  But those "ripples in a pond of sound" are imperceivable to a human in the first place, so "losing" them is losing nothing of the human sound experience.
 
Watch the video above.
 
Then watch this video (again explaining why 44.1 kHz is all that is necessary AND explaining/demonstrating bit rate/depth):

 
Really.  Watch them!  Together, it is 40 minutes of your time to dispel all of the bad information that you have learned/accepted/assumed during your previous decade of dedication to sound.  If you really care so much about sound, you cannot afford to not watch these two videos.  Your questions are answered.
 
 
If you want to hear some of the same information coming from the mouth of an "audiophile", you can watch this: https://www.youtube.com/watch?v=IiZqYnd5g8M .  I think that the demonstrations in the first two embedded videos above were already adequate, but the audiophile video just linked does talk a bit more about 16 bit versus 24 bit (at 29:17 ) and why 24 bit might be theoretically better, but that 16 bit in reality is almost always sufficient (and the difference often imperceptible). Skip to 44:13 to skip all the sample rate and bit rate/depth information already discussed, and listen to what really matters to audiophiles: format/compression.

If you dedicate 2 hours to the subject and watch all three videos completely, sequentially, you will be an audio expert and no longer an "audiofool" for marketing. Every question you had is covered in great detail.

How can you say to me that YOU cannot do this and do that yet you don't know me.

If you could, you would be a super human and I think that would have been one of the first things that you mentioned. As was said in one of those videos, medical researches would be banging on your door to learn how you can hear higher frequencies than anyone else.

A difference in how the audio is reconstituted and how it sounds coming out of speakers.

Again, as designed decades ago, at the designed 44.1 kHz sample rate, the input matches the output. It is guaranteed. There is no other output possible which will fit the bit stream. From a theoretical standpoint at least. If the amplifier is garbage or the circuit design causes a bunch of electrical interference, all bets are off. But, that wouldn't be the fault of the ADC, DAC, or the design of the technology. Again, I am repeating myself. So, "difference in how the audio is reconstituted"? No, not even possible. Again, the input matches the output. How could it be reconstituted differently?

Things that were not in the audio to start with. Thus higher sample rates and bit depths the ADC would be or should be capturing more REAL data of the audio signal.

There is no human-audible frequencies between the points. There is no junk data kept and no data is tossed out which should have been stored. At 44.1 kHz sample rate, all you need are a minimum of 2 points to characterize any wave form from 0 Hz all the way up to 22 kHz. Sound is a wave. At 44.1 kHz sample rate, two points will tell you where a 22 kHz sine wave's 2 peaks are; that is the bare minimum needed to perfectly replicate the sound. All frequencies below 22 kHz have more than 2 points characterizing their sine wave. All human-audible sound is quantified in such a bit stream. None is junk. Almost none is extra either. At higher sample frequencies, you are just filling the data stream will junk frequency data which you can't hear anyway. At higher sample rates, those junk frequencies just lead to distortion in your speakers which needlessly distort the lower frequencies which you can hear. Thus, those needlessly high sample rates needlessly reduce the audio quality.

As he said in the 3rd video, record some samples for yourself, sum the inverse of one with the other, and hear how little difference there is (usually no audible difference).

Hoggle
Honestly I have found that you can tell a difference with good quality audio gear but will admit that you start getting diminishing return with headphones above about $169. Going up to $300 does sound better but not twice as good and going up to $500 isn’t three times better. Yet you can tell a difference. It’s like 1440p and 4K some will notice a difference and some will not.

I'm not saying that there can't be a difference. Sure there is a difference if one is lacking compared to another. But, in an ideal world, where both perfectly reproduce audio as intended decades ago, there is no difference. That's a difference between theory and implementation. I can't usually hear much of a difference. Maybe some can. Any difference should occur before the ADC when recording or after the DAC when playing. 44.1 kHz 16 bit or 24 bit should be perfect for everyone in the world and sound identical on any device. The only difference in sound should be in the amp and noise induced into the input and output. Of course, purposely using audiophile wankery to purposely alter the sound, well, all bets are off then, huh? I would argue that wasn't the artist's original intent.

The nonsense about higher sample rates is just that, nonsense. That is mostly what I have been discussing above. No one should ever need to record stereo at higher than 44.1 kHz 24-bit (and usually 16-bit is all that is necessary). Sure, it's cool that EVGA's sound card supports higher. That is good. Increases compatibility. It's a selling point. But anyone who owns the sound card should make it their mission to not record at ridiculous sample rates and perpetuate the lunacy. All they would be doing is restricting the size of their audience and perpetuating bad habits and bad beliefs.

You also have people who will tell you that vinyl sounds better then CD because it’s analog even if it’s a digital remaster album....

That's also nonsense. There is a difference in sound, sure. Someone may prefer one sound to another, sure. But in implementation, the digital one is the one which matches the original; everytime. The digital one is more pure/perfect. If someone prefers the flawed/imperfect sound of vinyl -- which does not match the original recording -- that's a matter of preference. From a measurable sense, the digital version is better.