Agreed, but it isn’t 100% irrelevant either unless you always distribute your music in a lossless format.
I’m just saying that the reason the two different MP3 conversions don’t sound the same is not because one filtered more sharply above 20K hz, but because they do the MP3 encoding differently, as far as I understand.
To say that frequencies outside the range of hearing might change the way frequencies in our range of hearing sound, is like saying that colors in the dentist office change a little bit whenever they turn on the xray machine.
(How am I doing on the “stirring the bee’s nest” part?)
I think we’re now at a point where the skills of the people operating the equipment has more impact on the result than any differences in the equipment, where equipment is both hardware and software emulations of hardware. There is more to the equation e.g. the room you’re listening in, the speakers and some other factors that has more impact than the rather puny impact of WAV/AIFF → MP3, but the end result depends on your skills, whether you like it or not haha! What I try to say, I think, is there is less and less to hide behind as the equipment gets better, so we have to stop blaming our tools some day soon … 
And I’ve noticed the dentist’s room actually transform in a fragrant, pneumatic way as soon as I see the x-ray machine and everything feels much more orange-ish … which makes me happy, because the dentist has a vinyl player with gold plated contacts although the RIAA-circuit is switched off … and he hasn’t noticed … 
2 things:
MP3s are not lossless. Data is thrown out the window during conversion. When I did an analyzation years ago, the data “lost” was dynamic range and also the density/complexity of the sound.
Frequencies outside the range of hearing do create an audible effect of what is in our hearing range. You can easily use 2 frequencies with signal generators to see. How it relates to organic music, I dunno, but in a scientific test using sines, square and saws, you can hear and see the results.
I did a search here and in the old forum, but I can not find my posts about it. I had attached spectrograms that showed the lost data due to conversion. I don’t think I did sonograms of different sample rates.
Larry, I don’t know of any free programs that will do spectrograms, but I’m sure they exist. When I need that type of analysis, I use sound forge.
Who’s blaming tools? I think I may have missed it. 
I think there are an infinite amount of factors that create variables if a human being is involved. Gear is just gear. No matter how good or how poor, gear usually preforms in a predictable manner.
Didn’t the admin at GearSlutz once say that he does all of his mixing using a Soundblaster card?
Right, that was sort of my point; the difference in sound probably has more to do with the different ways that the MP3 conversion routines lose data. The one in Cubase is the licensed Fraunhofer conversion; the other one is probably the free crowd-sourced Lame converter. But maybe it’s because Lame cuts the frequencies off more steeply above hearing range, and it affects those in range.
Yes, I’m aware of this one, which is an electronic signal phenomenon, right? Aliasing? Which is why if you keep turning up the frequency on a sine wave generator way up you start to hear other weird frequencies coming in. I didn’t think this was a phenomenon with actual sound waves in the air, though. And I thought that this phenomenon was normally handled by plug-ins and da converters by oversampling, so we don’t have to worry about it? I admit I’m in over my head!
But I am trying to imagine bats in an old opera house, and their high-frequency emissions ruining the soprano’s aria. There’s probably a Mogami device that could prevent that!
There are high frequencies outside our threshold of hearing happening all the time. Don’t limit yourself to bats in opera houses because I don’t think I have ever seen any in there. The world is your oyster! 
I can’t remember the exact technique I used to test, but I had two tone generators going. One tone in the human hearing range and the other outside human hearing range. All I know it I could hear the add of the tone outside the range impact on the one within my hearing range. Like I said, how this relates to the real world, I don’t know, but I did have results that there was an effect created. There are 2 possibilities where the result was creates: If the result was created during internal summing and printed on mixdown, it would be heard on most playback systems. That is if the internal summing buss printed the effect of the interaction.
If happened within the speaker or in the air, the phenomena would most likely need a high quality speaker to reproduce the 2 tones in order for it to happen within the actual speaker or in the air. This would be tougher to reproduce as it would be the playback system that would create the phenomena.
It has been a number of years since I did this and where the results showed the interaction, I can’t remember and can’t fire up my rig ATM.
There are 2 possible reasons for this:
- Intermodulation distortion: If you have 2 (or more frequencies) in analog domain, all analog gear creates sums and subtractions of these frequencies.
- Aliasing: If you have a lousy A/D, D/A or sample rate converter, any signal having frequency above (target rate’s) Nyquist frequency will create sound at NyquistFreq - OrigFreq.
Both of these are something you don’t want to hear. And preserving ultrasonic content just to create these is not a good idea.
Question of the day: Can 2 notes create a harmony if one note falls outside the hearing range?
All I know it I could hear the add of the tone outside the range impact on the one within my hearing range. Like I said, how this relates to the real world,
I suspect you perceived and therefore think you heard a difference, but in actual fact I don’t believe there actually was one…
My experience is that most people hallucinate differences in sound, psycho-acoustics and all that, and I’ve also found that peoples perception of how one piece of gear sounds compared to another is directly proportional to how much money they just spent! Which of course has nothing to do with how it ACTUALLY sounds, something that can’t actually be determined by a human simply because human perception is so influenced by so many factors and therefore untrustworthy when it comes to true objective evaluation.
My guess is no one actually tried the procedure. I am very doubtful that I made up the results in my head, but I don’t know if I had anything to drink that night. 
A sequel (pre-quel?) to the link in the OP.
Factoid #1: CD mixtapes made at home in the Neolithic age: 6 bits at best
Factoid #2: THE BEST that tape can offer, irrespective of the studio (small project Studer vs. Abbey Road, e.g.): 13 bit recordings!
Factoid #3: For our 16-bit recordings, dither is barely necessary, if at all, in terms of what we can hear (17:00).
Lots of other good stuff too.