I wonder how you would classify someone who posts derogatory remarks, with no intention of stating or proving a point.
That's Mooly's job, and he will blame it on spell checker and let it all slide.
Right, so you are saying that DUT & LISTENER should be characterised. I would also suggest that the playback equipment, environment, testing procedure & other factors may also need to be characterised in order to qualify the suitability of the listening test for it's desired purpose.
And,in the search for truth, I wonder has Marce (not picking on him but he is the most recent example) actually participated in such controlled testing in order that his bias might be identified? Or is he happy to ignore his likely bias & decide that his biased listening results (even if blind) are more reliable than other biased listening results?
This doesn't just apply to marce.
Agreed, with a couple of reservations. (1) Not everything is line level, for example a ribbon mic preamp might have 75 db of gain available.
(2) Some people have noticed that when opamps are swapped in various <broken implementation> circuits, there is some resulting characteristic to circuit distortion that tends to follow each opamp around from circuit to circuit. I don't know why.
A very easy to image and extreme example is guitar distortion pedals. If you have several distortion pedals and try a particular opamp in each pedal, you notice there is some characteristic sound that follows the opamp around. If you try it with a different opamp, a different sound seems to follow it around from pedal to pedal. Of course, these circuits are designed to "misuse" the opamps.
But then if you take another circuit, still not well designed, but not so crazy as a distortion pedal, say, maybe a chorus pedal, or a parametric EQ, something like that, you also usually hear a sound characteristic to each opamp.
While all these sound changes I have described are caused by very significant circuit design issues, the observation that some characteristic sound seems to follow the opamp around leads people to assume the opamp itself is the primary cause of the sound differences, which may not be a proper conclusion.
The problem of determining cause and effect in complex systems is something humans are not intuitively very good at. Way too much to go into here. But if someone changes only one variable of many and sees a change at the output of a system, there is some tendency to conclude one knows the primary controlling variable. I will stop here.
Last edited:
Yes, the overall test should be well controlled. Well, as best can be done at least! And documented thoroughly, with an open acknowledgement to the limitations of the test at hand. Those forces limit one's ability to generalize the results of the test.
So, no, testing phono preamps where the amp/speakers is a boombox and the location is under your nearest freeway isn't going to help much of anyone. 🙂
So, no, testing phono preamps where the amp/speakers is a boombox and the location is under your nearest freeway isn't going to help much of anyone. 🙂
Mark, those examples are not "broken", they just rely on exercising the non-linear area of the opamps adjacent to their defined operating limits. The differences arise from the fact that the opamp designers don't have to follow any particular/common guidelines on how exactly the opamp shall behave beyond those operating limits. So everyone goes his own way as he sees fit. Thus a large variation.
Probably out of fear that numerology might be exposed as not having more substance than a hot air balloon. Better have a deaf audience that can't hear anything, or biased people that would deny they heard any difference. There's a trick to shame the later though. Those caught red handed once will subsequently decline to take any further listening tests, declaring all of them irrelevant and not worth their time etc.
More of it later.
Those are very leading questions for someone swearing not to pick on Marce. 😉
Former: I don't see that at all--given the very two people who've openly claimed to not hear differences since you've joined the thread have also stated clearly that it points to needing better controlling. I'm sure there are plenty of members out there that are biased towards null and don't want greater standards to control their own biases in a test, but I'll let them out themselves.
Latter: that is a bit of misrepresentation, as the said individual has used the "your equipment isn't good enough" as a cudgel to beat opposing viewpoints rather than work with the limitations. Doesn't exactly inspire confidence.
More than marce is implicated - don't make him the fall guy or even the martyr figure in all of this 😉
No misrepresentation, just following your own logic - if equipment isn't good enough (as you already stated... boombox, etc) to reveal differences, I don't see that this isn't a major limitation that trumps all others?
I do that daily. I can tell from a different floor whether my better half has turned on the mini system, the big system, the HT system or the tabletop radio. I don't even need comparative AB, start any and I can tell which it is. Guaranteed > 99% accuracy. 😀
Last edited:
Yeah, odd that Jay has been very quiet since he demonstrated he could get 8/8 right on a DBT using 2 bit identical files...
If you look at all the test files posted here, from the Sousa test to the 'wire or banana' test the people who believe they can over ride their brains when doing sighted listening are the last to step up to offer an opinion, chosing to either bit peek, or wait until after the results are known or just carp on about how the test is bad. The very best was someone complaining that a true life performance with just a singer, his guitar and a mic with zero additional processing had 'too poor production values' to allow him to judge it.
Those who are dismissed as 'T&M' are the ones who are willing to try the test and learn something from it. It's odd, and in a way sad.
So from the last 16 years of posts on this forum its clear which group are closed minded.
I would also note that a couple of posters on this thread actually understand blinded trials and the science and stats behind them from fields outside audio. They know what they are talking about rather than just being another internet random blowhard. but S/N is low so knowing which ones know their onions can be hard if you don't do your research.
Well aware. 😉
Think there's a typo, but I'll take your gist. One has to take each case one-by-one (dangers of meta-analysis) If one documents the equipment, then we just have to do the best we can. For some, no amount of equipment/room/etc will be sufficient, for others, the opposite. (Which is why most larger, better controlled studies tend to yield null results when their respective smaller studies show a result).
Doing science is fuzzy. Octuply so when it's humans involved. (3-2-1 wait for that statement to be misrepresented)
- Status
- Not open for further replies.