Quote:
My point is simple: You don't want to discuss DBT, you want to shove it down people's throats even though you don't understand it, nor science and scientific method. As you, and others have this attitude, DBT discussion has been restricted in the forums.
You seem, however, to want to try and make my posts to somehow be about cables, which they are not. My posts are about science. Without a proper understanding of science, discussing it even is near pointless.
Quote:
It appears to me that your posts in this thread are an example of a kind of 'smoke-screen' BS that believers of all sorts of magic use to defend their unenviable position in rational discussions. Vague general remarks ('you don't understand science properly' etc), evasion of direct confrontation vis-a-vis the facts relating to subject-matter, and of course, relying on 'endless discussions' argument (which actually become endless because someone tries to defend something magic, where any rational conclusion is impossible by definition) in order to give artificial credence to their position ('teach the controversy!', where there is actually none), at best, or ban, censor or marginalize more critical attitude towards those things, at worst, as we see on Head-fi.
The reality is, however, that there is nothing ambiguous, complex or controversial about the fact: If you really hear the difference with your ears, you should hear the difference with the same ears when your eyes are shut (ceteris paribus). If you don't, well, then the difference was not about hearing. It is as simple as that, you don't need to talk about philosophy of science or statistics to realize that.
Quote:
Dexter -- I think the point many of us have been trying to make is that formal audio DBT's are hard to do properly, very tricky to analyze correctly, and do not often establish general conclusions. In other words, many published studies have rather limited impact on our hobby and our day-to-day audio decisions.
Please understand that I love DBTs, believe in them, and undersand them well -- I am a statistican with world-class credentials (posted previously).
Most of us would agree with you that a difference you cannot detect blindly does not exist except for placebo. I highly advocate single-subject blind testing on yourself, getting a friend or spouse to help out. Buying audio without audition or return privs is just dumb. When I try something new, I work hard to test it blindly in my own room. If I don't hear a difference blind, or as blind as I can be (you sometimes have to settle for non-blind A/B'ing, but I try hard to set the bias in favor of "new and shiny" aside), I don't pay the money, unless I buy it for looks etc. as previously discussed.
This is very different from discussions of published DBTs, which were really gumming up several forums, or worse, discussion like this thread -- arguing about DBTs! So we made Sound Science, a great pun, and we are all happy there. Head-Fi has not banned DBTs, as many other posters have pointed out, we just said: do it in one forum only.
Just as Tyll is working on better ways to measure headphones, people should be working on better ways to run DBTs. Today's formal DBTs are surely not a panacea -- of course they seem at first to be the only scientific way forward, but on closer, more sophisticated analysis, many flaws (well discussed elsewhere) surface: response bias, forced choice, limited listening time, group effects, small sample sizes, inappropriate pooling of individual results, no adjustment for time and fatigue factors, insufficient alternatives, unrealistic settings, abuse of significance testing, lack of power against small but real non-null alternatives, etc.
Oh fun, another one of _these_!
So I actually think that Currawong and Wavoman are both right here, and both make good points. Since everyone here seems to like Wavoman's, I'll just focus on Currawong's.
Currawong's point, as I take it, is to say that when DBT is introduced, it quickly becomes established as the ultimate, sine-qua-non of judgment about any given item, as if there is a a coherent and universally-agree-upon methodology for conducting DBT. This turns many people off DBT, because as Wavoman states, there isn't. Now if we want to make small statements like, when testing, you should listen to each blind, or if you can't tell a big difference between items, you should say so, or something else, no one would argue. But DBT proponents, and many on the Sound Science Forums I've read, even here, go far beyond such commonsense propositions.
They believe - yes, "believe" is the correct term here - in Science with a capital S. Science "proves" things, establishes "facts" that are simply no longer disputable.
Academically speaking, that's just not correct. But even without getting into the dreaded "philosophical discussion", I think most people will recognize that this sort of thing simply does not obtain in daily life. Science is great - we have much to thank it for. But the problem I believe Currawong is bringing up is a real and valid one: it's the worship of Science by those who are not themselves scientists, and don't really understand how complicated science actually is. It's almost never a matter of having some Big Question that we need Science to answer, and then conducting some tests, and then getting The Answer and then no longer needing to discuss anything anymore. Just think about it, if that was the case, why would scientists disagree? Why would there be paradigm shifts every couple of decades where everything that was "established knowledge" within a given sub-field has to get thrown out based on a new model? But perhaps most of all, you're never going to have anything resembling this hagiographic conception of "Science" when it comes to human beings. Because human beings are INFINITELY more complicated than anything else we've ever encountered. And real scientists are not like those hollywood or storybook scientists who are these Leonardo Da Vinci generalists who seem to know everything about everything. Real science only makes progress by dividing everything up into such infinitely tiny spheres of consideration that virtually anything that anyone ever "proves" about anything is almost entirely meaningless in terms of definitively answering any "real world" question, unless and until we put together with tons of other individual "proofs" which by that very fact of addition becomes contestable, such that by the time you build up a grand theory of anything (esp anything human) based on science, it's no longer really "science" at all, but has a large component of "faith" in the sense I believe Currawong was trying to suggest. (we could also say, guesswork, or inductive logic.)
Economics thinks of itself as a science. Yeah, right. We've seen how that turned out. There are tons of "social sciences" like political science, sociology, etc. Marxism supposedly provides the most scientific and rational account of human socioeconomic development. It's not exactly universally accepted. Now even if you're going to throw all of those out, and say, those aren't "real" science, ok, but what about biology? Neuroscience? I can't tell you how many ridiculous articles clog up the world's papers every week discussing some new neuroscientific investigation. The investigations themselves, outside of a contextualizing discussion relating them to human sociology, culture, etc., are almost entirely meaningless, and the scientists themselves are smart enough to know never to make any real claims for them, but since that wouldn't sell papers, we get endless science reporters who tell us how "X is related to Y batch of cells" or "Z chromosome" as if that means anything, and nine times out of 10, you'll find it all traced back to the "amygdala, the primitive center of the brain that is the seat of emotion."
The only "fact" is, the human animal, for all our insane amount of research, simply gets more and more complex the further in we go, and while it's certainly possible to say a great many things that we weren't able to before, there are precious few answers to any of the big questions that people most want to know about. How does this related to audio? Well, one thing is that you've got electrical engineers on one side, who know very well how to build circuits. But on the other, you don't have anywhere near a perfect understanding of how human hearing works, whether for musical reproduction or just in general. Because before the 19th century, people thought we got sense data through our eyes or our ears or our hands, and then all of a sudden, everything went haywire, because the 19th century physiologists realized that more than 90% of our sensory perception was actually happening in our brain, which we know practically nothing about. Any philosophy has been trying to recover ever since.
As I said before to Crazy Carl in the Sparrow thread, I've read a lot of good reviews on this site where the reviewer will stress that the differences between A and B are incredibly, incredibly minor. Now if someone chooses to ignore that and go ahead and buy the more expensive one, that's their own fault. I think the biggest problem on the forum is actually reviewers having the moral hazard wavoman talked about, and I don't know an easy way to deal with that.
But you're never going to convince people that DBT should be the only legitimate way of analyzing gear simply by repeating the words "fact," "science," "rational," etc., over and over and using all caps. Because, even if they're not egotistical maniacs who have $100k of gear and need to justify their own purchases, and even if they're not making money off of the products you're questioning, just as regular human beings they're smart enough to know that just because you listen to A and listen to B and can't immediately discern a difference between them doesn't mean there isn't a difference to be found. To many people, all coffee and all wine taste the same. They learn there are differences only after much training and time. And it's much easier to say, well look, you can save a lot of money by _not learning_ those differences!
As I think ProgRockMan stated earlier, DBT ultimately proves _far too much_ because it quickly leads to Crazy Carl's position that everything basically sounds the same. Which I think is true - everything does, at some level of generalization, sound _basically_ the same. And listening to a great song with friends out of a crappy boom box or an old car stereo while having an amazing time is going to "sound" great. Even though the SQ, in some abstract sense, would be incredibly poor.
I think the better reviews around head-fi already make use of blind testing, but I think it would be difficult to chide them for not doing it in an absolutely standardized way, since there's no accepted standard by which it would be done. And given our present understanding of the human brain and thus the human sensory system, I doubt there could be.
