As I have said before, because of the digital edits...these "tests" are absolutely bunk.

You are wasting your time even listening to the files. Both should be 100% raw dumps.

So since I don't touch "A", [the pristine, better product's signal] whatsoever, I guess your concern is that inferior product "B" with the supposedly "bad sound stage"/ "bad dynamic range" (or whatever) gets, um, "magically repaired so it sounds just like the more expensive product" via the level change I apply, including of course the noise/distortion my volume alteration injects, and that's the only reason the two are indistinguishable in an ABX test . . . hmm. . . .if that's indeed true--which I highly doubt but I'll play along for now--then this means instead of buying expensive products like "A" all we have to do is buy inexpensive products like "B" and then apply the same tiny volume tweak I apply in software. BINGO: $200 products now sound just as good as $20,000 products but you save a ton of dough.

This is making a mountain out of a mole-hill. You want to use science to administer the rule of variable control but fail to realize the science says there exists a variable (level of audio) that must be controlled (somehow) for the results of the test to be valid. We have many ways to implement said level control, however that same science has "proven" that digital level control can best analog level control (especially in this domain where the level changes may be too precise for a typical analog volume ladder to compensate for). So...which is it? Or do you always get to have your cake and eat, lol?

Regarding all the other "crap" -- seriously, we need to test all those things for the results to be valid? By what part of the scientific method is that required? The variables under test are the two files and the audible difference thereof. The variable to control, to avoid false-positives (and, likely, based on other known (soft) science regarding human statistical preference, the "win" going to the louder file), is the level. The method to control said variable was chosen as digital modification to the file. I really don't see what the problem here is, except it doesn't jive with your agenda!

Of course, your previous beef was that it "modified the digital bits"...I suppose after debunking that you must back-pedal and find something else to whinge about?

Seriously, logic = fail in this world -- I don't know what to do anymore....

edit: to remove my..."personal issues" -- and to say:

The methods to control the variables must not introduce any more which are not then controlled. However, in that regard, I'd also posit that digital level control introduces fewer variables than an analog resistor ladder or another attenuating control was put in place (for only the singular, louder, file) given that digital volume leveling is "lossless" (unless you clip or dip under the noise floor...which isn't different from analog volume).

So since I don't touch "A", [the pristine, better product's signal] whatsoever, I guess your concern is that inferior product "B" with the supposedly "bad sound stage"/ "bad dynamic range" (or whatever) gets, um, "magically repaired so it sounds just like the more expensive product" via the level change I apply, including of course the noise/distortion my volume alteration injects, and that's the only reason the two are indistinguishable in an ABX test . . . hmm. . . .if that's indeed true--which I highly doubt but I'll play along for now--then this means instead of buying expensive products like "A" all we have to do is buy inexpensive products like "B" and then apply the same tiny volume tweak I apply in software. BINGO: $200 products now sound just as good as $20,000 products but you save a ton of dough.

Exactly -- if you follow his logical fallacy through to the end, it still doesn't work in his favor, !

My concern is an invalid test because of any software modification, period. Don't look beyond the scope into what you think I'm afraid I will "hear".

Obviously I cannot comprehend all the "jumps" that are being made in here as to why it's ok to edit raw files in software to try and skew the outcome of the experiment.

Yes, it's different and I'm astounded that people aren't picking up on this trick. You are editing the raw files themselves.

And I've still not seen any comment on my request of the delta's between the two raw files. I remember doing that experiment with 128kb mp3's vs 256kb mp3's vs raw .wav files back in the day. It was interesting to see what got "dumped".

The length that some folks will go to to try and justify any modification of a file is beyond me (especially by someone with a bias). I'm not back-peddling on anything. I'm just tired of the goalpost moving in here by the "scientists". Either follow the methods or don't...you cannot pick and choose to suit your fancy.

Please explain, in layman's terms, my logical fallacy. Obviously I cannot comprehend all the "jumps" that are being made in here as to why it's ok to edit raw files in software to try and skew the outcome of the experiment. If we're doing this, we need to see all the different variations of this experiment (as I've already stated).

I read this as you don't trust him (fine, no real reason to), but you should trust the method (assuming it was performed "honestly").

As I've mentioned previously, I can film the entire production from start to finish, showing the entire process with no edits, and the video will include the SHA1 hashes of the files (~digital fingerprints) so people who download the test can verify the file they saw me create in the video is indeed the one they just downloaded for the test.