In Reply to: RE: That clears up a lot! posted by Tony Lauck on September 2, 2009 at 09:10:59:
Tony:
You claim to know great deal about ABX tests for someone with no experience whatsoever.
Let’s start with the math. Please discuss this with a statistician you trust. It’s perfectly possible to prove that a subject isn’t hearing anything with this type of test and the subsequent analysis, to any degree of confidence you like. Because of the possibility of Type II error (which basically occurs when a subject is hearing differences but fails to record the correct answer for some reason), the number of trials required is much larger. Since it is very difficult to get any subject to do 25 or more trials with the same material, these tests are primarily used to verify positive results. That’s why we did not claim to have proven the negative, though the absence of positive results after so many trials still reveals something useful to those not already committed to dismissing the results (and the experiment, and the report, and the professional research scientists who reviewed it) as worthless.
You posit an elaborate theory about human perception, even citing evolution, and claim that whether or not the theory is valid, the fact that you can conceive of it in the first place is enough to invalidate our work. Your theory rests on all kind of strange assumptions, including that the subject might be deliberately trying to sabotage the experiment. “The listeners might simply be guessing,†you say, as though this proves your point. If the differences between the two sources are not apparent to the subject, guessing is what he must do when marking down either A or B as a source. The whole point of the test is to answer the question, “Is the subject able to hear a difference, or is he just guessing?â€
You then justify your refusal to expose yourself to this kind of test on the basis of more unwarranted claims, e.g. “the more you do these tests, the more everything sounds the same.†This is the exact opposite of what happens. I’ll grant you that if the differences between two sources are undetectable, continuing to guess which you are listening to can be frustrating. But one’s perception isn’t dulled. Any real difference, however subtle, still emerges reliably, as you would realize if you tried it yourself.
Finally, you cite your experience with Clark Johnsen’s system versus mine, as though it had anything to do with this issue. It doesn’t. You’ve told this story here before, so let me take this chance to address it.
I never used ABX tests to pick my equipment or set up my system, and I have no idea where you got this idea. And even if I had, how much you enjoyed listening to it would prove nothing. The two systems were designed for entirely different purposes.
Clark’s system (at least at the time we are discussing, which is when I heard it and measured it) was very fat in the range from about 70-200 Hz, and a very polite, recessed upper midrange. Its lush sound was very attractive on a wide range of recordings, with the added benefit in the midrange of correcting the inaccurate EQ he was applying to a lot of the 78s he played. (At the time I’m talking about he used a GE variable-reluctance cartridge and a preamp with no adjustable turnover or roll-off.)
Like you, I found the lush, attractive sound if his system very enjoyable on a large variety of recordings, but it was euphonic, and too kind to his source material to differentiate good recordings from bad in the way I need. I don’t wonder that you enjoyed it more than mine; on many recordings, I did too.
But unlike Clark, I use my system for mastering. If there’s too much 3.1k by half a dB, or if I’m using the wrong EQ for a 78 someone has brought me to transfer, I need to know that right away. With my system, I do, because the recording sounds wrong, which is to say it’s not as much fun to listen to. So, relatively few recordings sound really right in my studio. You may not understand why I would choose a system like that, but I’m sure many others on this forum do get it. -- EBM
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- RE: That clears up a lot! - EBradMeyer 09/3/0903:29:26 09/3/09 (10)
- RE: That clears up a lot! - Tony Lauck 06:26:37 09/3/09 (9)
- RE: That clears up a lot! - EBradMeyer 15:10:27 09/3/09 (8)
- RE: That clears up a lot! - Tony Lauck 17:25:53 09/3/09 (7)
- RE: That clears up a lot! - EBradMeyer 02:52:30 09/4/09 (6)
- RE: That clears up a lot! - chris.redmond2@bushinternet.com 11:06:28 09/4/09 (5)
- RE: That clears up a lot! - EBradMeyer 14:02:04 09/5/09 (3)
- RE: That clears up a lot! - chris.redmond2@bushinternet.com 02:44:39 09/6/09 (2)
- RE: That clears up a lot! - rick_m 09:25:01 09/6/09 (1)
- RE: That clears up a lot! - chris.redmond2@bushinternet.com 12:33:25 09/6/09 (0)
- ABX box definitely better than Ritalin nt - Tony Lauck 17:03:17 09/4/09 (0)