[skamp]

Knowing what A and B are does not in any way make an ABX test invalid

Knowing what to listen for gave you an advantage that you wouldn't have had, if you hadn't analyzed the files first. And for the same reason, you didn't choose which files to ABX by accident. Not exactly 100% blind, is it.

TBH I don't think you're a very good sport. Every time there's a blind test on Head-Fi, you go straight to analyzing the files, going so far as to reveal some clues to others (spoiling the fun for everyone else), that you didn't even gather by just listening to the files. I wish you would keep your findings to yourself.

I'm much more impressed by miceblue's result, assuming he didn't do anything special prior to ABXing.

 

[skamp]

And since you seem to have figured out which is which by analyzing the files, whatever preferences you may have are now tainted with expectation bias. Good job.

 

[stv014]

You can consider my results invalid if you want, I do not really care. The purpose of an ABX test is simple and clear: to determine whether one can hear a difference between two sounds or not, the only thing that is required to be a secret is the identity of X in each trial, that has to be found out solely by listening. People do not normally buy and own random gear blindly anyway, so knowing what A and B are is realistic and fair.
 
By the way, here is my guess what the files are:
 

Spoiler: Warning%3A%20Spoiler!

OK, only the md5 sum for now:
 
d11e4840d392a73166e0e3308c2e5c04

 

[stv014]

Quote:

And since you seem to have figured out which is which by analyzing the files, whatever preferences you may have are now tainted with expectation bias. Good job.

 
I do not recall making any comments on which files I prefer, nor have plans to do so, the differences are not large enough to make one clearly "better" than another. But if I wanted, I could easily randomize the files again to a different order of letters.

 

[skamp]

By the way, here is my guess what the files are:

Guessing which is which by thoroughly analyzing the files and cross-referencing them with my own RMAA measurements is completely besides the point, surely you can understand that. It has no value whatsoever in a (tentatively) blind listening test. It's good that you have such skills, but in this particular context, you're just showing off and spoiling the test for everyone else. You do this systematically with every listening test, and it's getting really aggravating.

 

[miceblue]

Quote:

Quote:

Knowing what A and B are does not in any way make an ABX test invalid

Knowing what to listen for gave you an advantage that you wouldn't have had, if you hadn't analyzed the files first. And for the same reason, you didn't choose which files to ABX by accident. Not exactly 100% blind, is it.

TBH I don't think you're a very good sport. Every time there's a blind test on Head-Fi, you go straight to analyzing the files, going so far as to reveal some clues to others (spoiling the fun for everyone else), that you didn't even gather by just listening to the files. I wish you would keep your findings to yourself.

I'm much more impressed by miceblue's result, assuming he didn't do anything special prior to ABXing.

Nope, I literally just download the files, find a spot that I like to listen to and won't mind repeating for a while, and just do the ABX test using Foobar.

 

[stv014]

Quote:

and spoiling the test for everyone else.

 
Exactly how ? Your RMAA measurements of all the devices tested have already been publicly available before, and are even linked in your signature. I did not include any information that allows for finding out which FLAC file is which device, except for those who also analyze the files themselves anyway.

 

[skamp]

You're being obtuse. I didn't guarantee that no-one could cheat at this test, and just because someone COULD, is not a valid reason for doing it yourself and describing every step of your process. 95% of everything you've done didn't involve any listening, as shown by your posts.

I'm fed up with this and I'm tempted to ask a moderator to lock this thread.

 

[Mattimis]

Its a blind test which is the entire point, determining which source a and b are destroys the value of a blind test entirely, because it is no longer subjective.
(no matter how much you defend it)
 
The fact that the data is available to determine which is which should be an after thought for you to enjoy once you have done the blind test to compare your results.

 

[skamp]

Analyzing the files as thoroughly as you have done, before actually listening to them, defeats the point of a blind listening test. Also, processing files like you do in your own thread in order to mask differences and the effect of the ADC will undoubtedly raise suspicions on whether previously audible differences where erased or masked by all the processing.

I don't doubt your skills, but in both cases, they're not serving the initial purpose.

 

[stv014]

Repeated one more (and last, I am done with this thread) time:
 
Quote:

The purpose of an ABX test is simple and clear: to determine whether one can hear a difference between two sounds or not, the only thing that is required to be a secret is the identity of X in each trial, that has to be found out solely by listening.

 
It is amazing how this can be so hard for some to understand.