I downloaded the zip and loaded them into X3 after extraction. I could not tell which was which. Using selective solo and Nugen's Visualizer beta at 64 bits. The spectrum for both looks the same to me. But this is just a very casual and quick sort of thing. 

the 2nd one sounded 'thinner' to me...and couldn't tell if it was volume, listening on good headphones.   they both were very close and if the difference is that big I should start laying asphalt as a living and sell my guitars, it aint gonna happen....  although asphalt layers do make some good money!
 
I was gonna pop them in SoundForge, but I figured someone else would do that (like John above  ;-)  )... and be better capable of making such an analysis.

microapp
The sample Z3ta-2 files you posted sound to me like what happens when Z3ta is switched from low to high resolution.

Since I'm pretty sure low or high resolution relates to oversampling, then that would support the idea of using 96k with devices that don't offer oversampling to provide a similar improvement. Yes? 
 
Remember the whole point of the original TH2 thread was that I thought it sounded better at 96k, the same kind of difference I heard when switching Guitar Rig into HI mode. I suppose if every element of a software system within the computer offered oversampling, then 96k would be moot.

Beepster
This does seem to be a rather contentious issue which strikes me as a little weird.

 
Yes, it's like religion to some people or something. I honestly don't care one way or the other but if I hear a difference, I want to know why and if it's an improvement, then I want to implement whatever causes an improvement.

mixmkr
the 2nd one sounded 'thinner' to me...and couldn't tell if it was volume, listening on good headphones.   they both were very close and if the difference is that big I should start laying asphalt as a living and sell my guitars, it aint gonna happen....  although asphalt layers do make some good money!
 

[spoiler alert]
 
The second one is the one at 96kHz. The reason why it sounds thinner is because it doesn't have the "fatness" from the distortion. But also if you listen to the very highest frequencies, the second one reproduces the high frequencies better, and also, in the first one you can hear the "wooliness" from the foldover distortion at high frequencies.
 
To me it's definitely an audible difference, but then again, I do a lot of mastering and my ears are really calibrated to what's happening in the highs. I also have example files of the TH2 and AD at the different rates. Of course 96k fanbois will take me to task for bringing them back into 44.1...but that's kind of the point, 96 bakes improvements in the audible range that 44.1 is perfectly capable of reproducing. It's not like there's something wrong with 44.1, it's that under certain conditions there's something right about 96. Maybe it's about what bitflipper says, and the 96k is fixing something that maybe shouldn't be broken, but is.
 
However the other thing is little improvements are cumulative. I wrote an article once about how to produce quieter recordings and the thrust of it was you had to do it 1dB at a time. A little noise here, a lowered fader there, a re-oriented transformer to reduce hum pickup...after a while, it added up. If there's a slight improvement on individual tracks, it adds up to a significant difference with multiple tracks. I think this is why people often say "I dunno, 96k just sounds better to me." With some people I'm sure the placebo effect comes into play ("It goes up to 11, it must be better") but I think others hear that cumulative difference, even if they can't point out individual elements.
 
The reason I had no interest in 96k before is that I've done sessions at 96k that were released on CD and no one could ever tell the difference; I certainly couldn't. And no studies have proven that people can tell the difference between hi-res audio and CDs when listening to a playback medium. But I never thought to check whether recording at 96kHz could yield different results, and in this case, it came as quite a surprise but I can't deny what I'm hearing. The second file has better, and cleaner, high frequency response. 

John
OK but why would they be there anyway if (the higher frequencies) frequencies above half the sample rate are filtered out? And if this is happening than it must be considered a distortion. 

Distortion is exactly why it happens. The audio interface takes great pains to assure that no frequencies above Nyquist enter the system, and if you only mixed pure audio tracks with no intentional distortion you'd never have to worry about aliasing. But if distortion happens within a plugin, intentional or otherwise, harmonics are going to be generated. 
 
Plugins whose primary purpose is distortion, such as amp sims, are designed to deal with it because the designer knows up front that it's going to be a problem. That's why the better products offer 4x or higher oversampling, followed by filtering. Illegal frequencies are still generated internally within the plugin, but none of them make it back out to the DAW.
 

...wishing the problems weren't there in the first place doesn't make them go away; raising the sample rate does. Most of the foldover with 96kHz bounces back into a range that's above 20kHz so we don't hear it.

This implies that aliasing is an inherent and unavoidable byproduct of DSP. It isn't. Many types of processing never produce aliasing/distortion. For example, "boosting the treble" does not cause harmonic distortion. It will, however, exacerbate any existing distortion by boosting the harmonics.
 
Adding saturation does cause harmonic distortion, but again if your saturator causes aliasing it's the saturator's fault, not your choice of sample rates. Raising the sample rate would only be a band-aid to cover up perhaps half the fallout from uncontrolled harmonic distortion. That saturator will still sound bad at 96 KHz, just marginally less-bad.
 
Does 96 KHz push "most of the foldover...above 20 KHz"? Hmm, maybe. Depends on how you define "most". "Most" harmonics on most distorted sounds easily fall under the Nyquist frequency even at 44.1 KHz.
 
Let's look at a practical example, an electric guitar played through a high-gain amp sim. You play a very high note on your guitar, say with a fundamental frequency of 1.3 KHz (an octave above an open high-E string). The amp sim will generate harmonics at 3x, 5x, 7x, 9x, etc. The 15th harmonic is 19500 Hz, still legal at 44.1 KHz. You have to get up to the 17th harmonic before changing the sample rate would deliver any benefit. I didn't do the math, but the level of the 17th harmonic is going to be down more than 90 dB from the fundamental. IOW, inaudible. 
 
Synthesizers are a different can o' worms. A so-called "supersaw" contains both even and odd harmonics plus many sum and difference frequencies from the interplay between multiple pitch- and phase-modulated voices. Synthesizer designers have to anticipate such harmonic complexity. Improperly designed oscillators can indeed generate out-of-control harmonics. If Z3ta+ falls into that category, then I maintain that it's a design flaw. Zebra, for example, does not do this as far as I can tell. 
 

I agree upsampling and oversampling helps, but both rely on interpolation and in the case of oversampling, stuffing in zeroes and interpolating on playback because attempting to interpolate while recording creates its own issues. Running at 96kHz for sounds that are generated synthetically provides "real" data for each sample.

 
There is nothing inherently bad with how upsampling works, even the "stuffing in zeroes" part. In fact, this is what happens every time you record anything! Your audio interface isn't recording at 96 KHz, it's recording at some multiple of 96 KHz, at least 64x that rate. So you're using upsampling all the time. If it were messing up fidelity we'd still be recording directly at 96 KHz like they did in the early 60's when 96 KHz became the standard - precisely because of the lack of oversampling at the time.
 
Sample rate determines the highest frequency you can process. Period. It has nothing to do with "precision". Calculations aren't going to be more precise at 96 KHz. Precision is determined within the code itself, e.g. whether or not the programmer uses double-precision floating point values for all the internal math.
 

Saying I need better plug-ins is all well and good, but that's sort of like saying all my plug-ins should be 64-bit, which they should be. But some aren't, so I have to use bridging, which is a flawed technology but serves its purpose.

Careful, let's not bring bit-depth into the discussion. It's irrelevant to the topic. In fact I would not be the one to say that all your plugins should be 64-bit, because I know that would have no discernible impact on fidelity.
 
Now, I do understand and agree with your underlying argument, which is that you're stuck with the tools at hand, and a higher sample rate may improve the performance of some of them. I certainly couldn't afford to replace all of my plugins! So yeah, if I found out that most of my plugins were deficient I'd be seriously bummed and looking for any way to work around their limitations.
 
To put your premise another way, you could rephrase it thus: "if your plugins are substandard, a higher sample rate might or might not help". I'm good with that.
 

My goodness, a whole page of replies were posted while I was typing my last one! I'm glad to see there's such interest in what many might deem a boring topic.
 

One more question: Wouldn't a higher sample rate also spread out quantization noise over a wider bandwidth? I also wonder about jitter. Wouldn't a higher sample rate distribute any jitter over a larger number of samples, which when interpolated and filtered, would give better results?

 
Yes, quantization noise is going to have a broader distribution. Whether or not that explains why 96k sounds better to you depends on whether you could hear the quantization noise to begin with. Plus how accurate the quantization is to begin with. A given interface will be optimized for a specific sample rate, whatever rate the design engineer assumed most customers would be using. That's the main reason an interface might sound better at one rate over another: there are design compromises necessary to support a wide range of rates.
 
As for jitter, that's really a clock issue. I can't think of any reason why a clock would be more or less stable at one rate than another. The master oscillator runs at the same frequency regardless of sample rate.
 

Speaking of recording at higher sample rates somebody once told me 88.2 is the best to record at simply because it's 44.1 doubled. They said the math that is computed is handled better at this sample rate? I don't know a thing about it. I'm not that technical when it comes to this sort of thing, but it makes sense to me. I record at 24 bit/44.1 but I think I'm going to try recording at the higher sample rates. Why not?

Interesting discussion. @ Craig - the files do sound different with file 2 sounding clearer. I don't use too many synths other than a drum module and guitar sims. I'll have to give this a try later. I can see hard drives filling up quick vs the quality sound I have now at 44.1/24. It would have to be quite noticeable for me to switch, but I will give it an honest appraisal. Thanks for looking out. 

Geo524
Speaking of recording at higher sample rates somebody once told me 88.2 is the best to record at simply because it's 44.1 doubled. They said the math that is computed is handled better at this sample rate? I don't know a thing about it. I'm not that technical when it comes to this sort of thing, but it makes sense to me. I record at 24 bit/44.1 but I think I'm going to try recording at the higher sample rates. Why not?

Urban myth, unfortunately. One of those concepts that sounds reasonable but just ain't so.
 
As for the "why not", the only real downside is the extra disk space you'll use and the extra CPU load. If you have gobs of disk, and most do nowadays, that's not much of a barrier. What you're more likely to encounter is your CPU running out of steam sooner, meaning fewer soft synths and effects and more freezing. The upside is lower latency, if that's important to you - assuming your CPU can handle the same buffer sizes at the higher rate.