Some marketers such as 2L took down their Hi-res download test bench page, I believe, once they saw individuals like me exposing the public to how to easily use the files* to conduct
valid, proper double blind, level matched, synchronized, statistically analyzed tests such as with the free Foobar component "ABX"
They took it down years ago but there are still
snapshots on the wayback machine (internet archive) which seem to have at least some functionality, I think, where one can find cuts to download. Not sure if these work, but here. The pages look different, not sure why, but look for the play arrow at the bottom of each of these links:
CD 44.1/16-bit res
192kHz/24-bit res
The reason track 2L-053-04,
Joseph Haydn: String Quartet In D, Op. 76, No. 5 - Finale - Presto, Engegård Quartet,
is the most important cut in my mind is because it was also used in a
dubious JAES paper by one of the Hi-Res promulgators. Could he locate
any individual who could hear the distinction on his chosen cut?
NO, not even one.* However he pushes that after testing dozens of people he added up all their scores and instead of 50% correct answers and 50% wrong (so no audible difference could be demonstrated) he found that their combined scores were 53% correct and 47% wrong. When dealing with this large a number of responses that does indeed technically meet traditional statistical significance. But think about it for a minute in terms of "
meaningful significance" to an audiophile. . . .
Joe: "Wow, I just bought a new Hi-res player and test track.
It's amazing!"
Friend: "Cool. So I guess you can hear the difference against CD quality. Right?"
Joe: "Um, me personally? No. About half the time I guess right and about half the time I get it wrong."
Friend: "Oh, well I guess you mean other people in your family can hear it. Right?"
Joe: "Well, no, not really."
Friend: "So how many people did you have to test to find a person who could hear it?"
Joe: "Um, I, um. . . I haven't actually found anyone who can hear it on their own."
Friend: "So why are you stating it is 'amazing' then?!"
Joe: "Because after testing 100 audiophiles I know, using $80,000 speakers, giving them 10 questions each and then combining all their responses together, they successfully got 530 answers correct and 470 wrong."
Yikes.
That's a valid reason to buy something?!
Not in my book!
*Can you see why for the JAES Hi-res author/promoter it would be in his and his company's best interest to not claim he
found anyone? Because if he had then other scientists would want to contact that individual to do further testing and see if they could replicate the results. His solution?
No individual = no further scrutiny can occur. Case closed.
He also used other steps to ensure no further examinations could occur. He heavily stressed in the paper the fact that the gear he used to test the listeners had to be
extremely good. [Technically, he never actually ran the tests also on lesser gear, as far as I can tell, so this was really just conjecture on his part.] The loudspeakers alone were $80,000, at least in their "special edition" version (if I recall correctly) and most other audio researchers don't have such speakers on hand. So any attempt to replicate his findings which failed to do so, using only say $70,000 speakers, he could easily dismiss with: "Well they used junky gear and I
warned them not to!".
There were numerous other questionable things about the paper which I am omitting for brevity.