Quote:
Originally Posted by Psalad
Fair enough point, and that is logical. However, when you combine the potential or ideal with the existing evidence.. nobody, to the best of my knowledge, has ever been able to reliably pick out higher sample rates (above 48k) in a scientific double blind test.
http://www.mixonline.com/news/profil...ng-rate/365968
While you could be right.. and I'd trust Lavry over me... I still go back to the evidence of testing, because I'm more interested in the real world.
What would be an interesting exercise is to evaluate some of the early CD players vs. the current converter technology. Even at 44.1, I'd imagine we would hear the poorly designed filters and lack of oversampling... but the technology has gotten so good, I doubt it makes a difference today.
But I'm not one who says things are necessarily "settled" scientifically because things never really are. For now the evidence sure seems to be pointing to there being no difference (or a difference that somehow doesn't show up in double blind testing but shows up in other evaluations).
Anyway, for me, it's 44.1, but today's world means for many with newer computers and tons of CPU, I guess it might not be a big deal to hedge your bets with 88.2. (BTW my ears are old at 52, so YMMV.. maybe the young guys can pick it out). |
I agree that it is not really settled. And I agree as I said, the other elements are very likely much more at play. And as you say and as Pokiehat's post indicates the computer's spec's will be a huge factor in determining what sample rate is even practical.
But for me simple logic would suggest that one of reasons it is not settled is that much of so called " evidence" suffer's from significant problems in its methodology . For example the test sighted in the above link your referring to.
Is to my mind indicative of much of "evidence" that often has basic problems of method from the get go, in terms of the difference between what people may perceive or expect the tests may be demonstrating, and what the tests are objectively showing.
For example:
While the author of article states.
It was designed to show whether real people, with good ears, can hear any differences between “high-resolution” audio and the 44.1kHz/16-bit CD standard.
I would suggest that considering the actual method used in the test he is sighting, that making that general of statement about this particular test, is arguably little more than expectation bias.
Unfortunately the test sighted as actually conducted, does not really show that, because of the method used, and logically that type of method for testing can not demonstrate that particular "stated design goal" per se .
The simple basic logical error in the method used in that test, is of course starting with a single file recorded in higher resolution , and then dithering that high res. file down to 44.1. for playback and testing.
So in reality the only thing that test actually demonstrates is that when you take a file already recorded in higher res. and play it back in both higher res and 44.1, that people practiced in critical listening were only able to detect which was which, 52% of the time.
That could in fact logically be stated to simply be demonstrating that recording in higher res. is actually beneficial when the file is later dithered down or compressed.
Unfortunately that method of testing (starting with a hi res file) does nothing to demonstrate the much more real world situation of both recording and playing back, in different sample rates, to see if people can hear the difference.
Which would mean using a single performance that is split and simultaneously recorded in both 44.1 and say 88 or 96. and then played back in 44.1 and 88 or 96. respectively and then perhaps another test played back again in only a single reduced resolution.
It seems logical to me that the only way would be taking the signal from the same front end pre amp outputs and splitting it:
Send it to two of the same brand and model of AD/DA converter (as I don't think one unit can handle two different sample rates at the same time)
One at 44 and one at say 88. and into two sessions ( one in 44 and one in 88 ) in the same brand DAW, on two of the exact same brand and config. of computers.
Then take the outputs from the computers back to the two AD/DA's and back through the same playback system.
Then see if people can determine which is which.
I have yet to actually read about a test with this type of methodology.