Can you do this blind? I would think the best method would be to play the two tracks in a random shuffle where you could get any combination of plays, including sequences of the same track consecutively.
If you can then identify the same track with a statistical level of significance,say 8 times out of 10 over time, then that would be a very convincing result.
I can I imagine I wonāt identify the same one. To me I think the whole thing is down to our brains and the process of trying to compare two things at different times as your relying on memory and this is flawed or at least in my case. The files contain exactly the same data I have no doubt about that. Using Audacity itās all a real-time process and not from memory and this reveals a different story as they sound the same. In Roon or any other music streaming app you canāt compare in real-time itās always from memory. And what I have noticed is the more you do the memory comaprison then itās not always differeny sometimes its more apparent others not.
I think youāve got it right there - itās all in our heads. I suspect the result of such a blind test over a long enough period would be 50/50.
This has generally been a fairly refreshing thread about this sort of topic. Usually they descend into insults and shouting very quickly but people have been willing to get involved in some proper analysis. Would like to see someone try the blind test though.
I still need to listen to the files you removed the padding and metadata from to see if this is any different. I doubt it but needs to be done all the same. But I guess this now comes with some sort of bias. Itās a funny old game this audio, and I feel is what the industry plays on with all these devices to fix the issues we never knew was there.
I wonder if someone with the files can use foobar2000 and the ABX Comparator component to perform an ABX test with the two files.
Thatās what I am doing, disc by disc.
Thatās effectively what I offered David (and CrystalGipsy and anyone else who wanted to play): 4 files, at least one of which is the Unitiserve rip and, at least one of which is the Melco rip (so 14 possibilities in all). The challenge was to distinguish which was which.
Think of this as 4 rounds of an ABX test. Obviously, 4 rounds is not enough to achieve true statistical significance. But it ought to be enough to dispel the notion that thereās some glaringly-obvious difference between the two.
Weāll see what David has to say ā¦
Reluctant to chip in on this heated debate, but I have compared a UnitiServe FLAC rip to a D100 FLAC (dBpoweramp) rip and they are audibly different and have different sizes. A spectral analysis shows they are also visually different, so I would question not the AoB, but the laser and error correction on the two ripping systems.
The odds of two CD rippers producing the exact same count of bytes is incredibly improbable, but not impossible. As the MD5 is showing the same values I would question if the correct files had been sent in the first place.
Yes, there is snake oil out there, but donāt discredit someone hearing an audible difference time and time again as there may be valid reasons for this.
As did Davidās original files:
% ls -l '14 - Nashville Cats 1.WAV' '14 Nashville Cats.wav'
-rw-r--r-- 1 distler 27891070 Dec 18 09:11 '14 - Nashville Cats 1.WAV'
-rw-r--r-- 1 distler 27648334 Dec 18 09:11 '14 Nashville Cats.wav
A bad rip is a bad rip. David sent me two good rips.
After trimming the metadata and the leading and trailing silence, the odds that two good rips contain exactly the same sequence of bits is 100%.
If you are willing to share the files you say are different (PM me), I would be happy to repeat the same analysis that I did with Davidās files.
If one of yours is a bad rip, we will quickly determine that ā¦
They will look different in spectral analysis if one has different padding or is different. These files had a silence padding. If I check both files spectral frequency display in Adobe Audition they look different unti I remove the padding and line them up so they are the same timing and then itās the same.
David did eventually write back. After confirming his choices, both on his iPad and his N10, he identified 1 out of 4 correctly.
Anyone else want to try, before we put this baby to bed?
Does that mean that Crystalgypsy didnāt hear the difference between the two files, using Roon. Are we both afflicted with this auditory disease?
Anyway, nothing lost. Iām hearing my music better than ever, and you need take no action.
Happy days for all!
This is why double blind or blind tests are the simple answer to resolving such mysteries. This is nothing new, as the experimental psychology literature long ago dealt with the issue of ādemandā effects in sighted tests.
p.s. I remain very impressed that the parties involved in this discussion provided files, took tests, etc. This is unheard of in similar debates on the many audiophile forums. Iāve personally never seen such a willingness to participate in such a way in a situation like this. Thanks to all!
Okay, I listened to @David_O_Higginsā files and I concluded that they worth both identical. I wasnāt familiar with the track music and after listening to the first track I thought I heard a difference even though my head said this could not be possible.
At first the opening vocal seemed brighter on the second file played. But later, after being certain I had no idea what file was playing I thought the first was brighter. However, this impression didnāt last after I was reasonably familiar with the music. They are indeed the same.
I conclude that we (humans) do not have very good short-term auditory memory. If we canāt trust ourselves to be objective we must look elsewhere for grounding.
The thing is, the question we are asking ourselves makes no sense. If I served up two plates of food from the same casserole would foodies be debating whether plate A tastes better than plate B?
There is no reliable evidence to argue that file A is better than file B, and whatās been suggested flies in the face of science and engineering.
I thank David for sharing the files. And no, I donāt think you are afflicted with an auditory disease. 
Believing what we hear, see or feel is a very human trait, but it doesnāt mean that we are correct.
Enjoy your music however you like!
On this topic (memory), I listened to a very interesting āHidden Brainā podcast today about the science of memory. Memory is a very complicated thing. People can even start believing that they actually did things they never did under certain circumstances. Fascinating stuff. Thereās been lots of related research on the problems with eye-witness testimony in the legal field.
I suspect most of us will hear an improvement is SQ anytime we purchase something to improve SQ. The more it cost, the better it will sound.
I agree with you here Martin. At first this is what I thought then the more you listen it changes. Having just tried all the files from Jaques I could not really identify any difference and have had to guess as to which come from the same source.
I didnāt tell you which file was which (and I wasnāt focussed on that myself), yet you heard a difference similar to that which prompted me into this madness, and now you donāt hear any difference. Quite remarkable.
Happy Christmas all.
David
.
I am really pleased with the way this went too. And I want, again, to commend David for his willingness to participate constructively.
I would like to think this could serve as a model for future such discussions (but Iām not holding my breath).