Hi,
ditton said:
But I think it reasonable to put the focus on the audio-psychology of the judges - would be good if there were body of knowldege ...
I always tell this little episode of mine, back from the days when I used to arrange the occasional DB Test.
An ex BBC Engineer and virulent cable unbeliever got so much on my nerves with his prejudiced, uninformed bable about DBT and how cables make no difference that I included him in an experiment.
I posited that we would test the differences between cables. I had three others in the group who generally where openminded, that is they accepted cables can make a difference as a matter of fact but where not particulary gung ho about cables.
What I in fact tested on the test group was the polarity reversal of one stereo channel. Now I do not think ANYONE would argue about this being a clearly audible change?
Well, when the listeners where "blind" to the change the three "controls" scored a "perfect" DB Test (10/10 correct identifications) while our "disbeliever" scored perfectly random.
Seeing the test sheets he originally accused me of fraud.
I then told him what he had ACTUALLY been listening to. We have not on very cordial terms ever since....
I guess my point is threefold.
When we test "gear" we test it's impact on music replay, which is two layers of abstraction. Sort of like testing the impact shapes of glasses have on the taste of Wine.
Secondly, if the test is not truely blind (meaning the subject is unaware of what is being tested and is thus capable to switch of her or his internal "realiaty generation engine" which tends to make her or his perception of reality agree with her or his prejudices) the expectations and/or prejudices act as strong randomising agent.
If I believe strongly there is no difference I will not hear it even if it present, if I believe strongly that there is a difference I will hear one where non exist. Either one will radomise results to a considerable degree.
There are other issues too, often people will rank familiar items higher than those that are in a pseudo objective sense actually better but less familar.
As a result all subjective testing (and DB Testing of the impact of gear on music is still PURELY subjective) of music is a major minefield.
Ciao T