SY said:
Normally proper testing should include a exhaustive measurement of the DUTs. And sufficient "sensitivity" simply means, that the listener panel has to reach the sensitivity region in which the measured differences of the DUTs really are.
Just for example, if the measured differences in frequency response level were ~0.2dB, it wouldn´t help to show, that the listeners could discriminate level differences of ~2-3dB.
Another example regarding cd-players; the abx-website did claim in their historical data section, that audible difference between a phillips cd100 and a later generation Sony ES707?? was verified on 0.05 level.
Leaving aside, that no real data was presented, no controls etc. etc. then this could be a useful positive control. The listener panel in a (modern) cd-player double blind test should be able to discriminate the older player from a newer one.
The negative control would be to compare two identical cd-players.
So, if i got the meaning of your response right in regard to my question, your answer is a "no", but you think, that some papers of the authors mentioned by you, would qualify for double blind tests of sufficient sensitivity. Or did i misunderstand you?
AFAIR Davis did no listening tests, Greiner war working (mostly) on loudspeaker issues; could you be more specific about which part of their work could qualify?
Lipshitz´s article on phase audibility could be another control starter, but it points mainly to listening tests done by the BAS and published in their magazine, which is not accessible without difficulties.
BTW, did you follow the publications back then about the "tiefenbrunn tests"? There was some room for disussions about background hiss and switching relay sound. These issues raise of course some questions about the "sensitivity" .
AFAIR Lipshitz was a strong advocat of frequency response leve accuracy better than 0.1dB, because otherwise audible differences would occur, but i´ve never seen any double blind verification of this thesis.
If it exists that could be a good positive control as well.
Wishes
Lipshitz discussed the level-matching critereon in several of his papers. It's necessary but not sufficient.
Completely agree on the need for negative controls; in the tests I designed professionally, these were always incorporated, as was subject training. In my work doing wine judging, test panel sensitivity was always included; to judge at the good competitions, I needed to pass DBTs on volatility, alcohol level, residual sugar, TCA level, and brettanomyces levels. Again, putting controls onto sensory testing is necessary for the work to be good, but having controls is not sufficient to make work be good. I eagerly await the first test (after 30 years of unsupported claims) from the purveyors of fancy wires that any difference whatever can be reliably heard by anyone between two wires that meet basic engineering criteria.
The BAS Stuff is a mixed bag- there's some fine work in there (Meyers's CD/SACD, for example) and other stuff which seemed less rigorous. Sturgeon's Law and all that.
Completely agree on the need for negative controls; in the tests I designed professionally, these were always incorporated, as was subject training. In my work doing wine judging, test panel sensitivity was always included; to judge at the good competitions, I needed to pass DBTs on volatility, alcohol level, residual sugar, TCA level, and brettanomyces levels. Again, putting controls onto sensory testing is necessary for the work to be good, but having controls is not sufficient to make work be good. I eagerly await the first test (after 30 years of unsupported claims) from the purveyors of fancy wires that any difference whatever can be reliably heard by anyone between two wires that meet basic engineering criteria.
The BAS Stuff is a mixed bag- there's some fine work in there (Meyers's CD/SACD, for example) and other stuff which seemed less rigorous. Sturgeon's Law and all that.
Yeah, I saw that, if it's the one about the relay "make" mechanics sounding different than the relay "break". Cuing is a real bear to try to keep out of a good experiment.
Greiner did a few papers on cables, mostly showing how pathological examples interacted with speaker impedances to cause readily audible frequency response errors. He was the guy who got me to spend a year devastating my own beliefs.
Yes, i meant the "relay sound issue" , sometimes the devil is in the details. 
But it happens more often than i´d have thought- another example has been described by Bart Locanthi.
Both examples do raise some questions about the level of sensitivity reached in the tests. Of course there a ways to argue why it more likely that something like these "noise" remains undetected, but it is not really convincing.
Of course, controls are not everything, but they provide something of a basic safety ground.
The positive control does provide evidence for a sufficient sensitiviy level in the test and the negative control ensures, that no other factors are the reason for detected differences.
I remember that sometimes the negative control was only a pretended switch between the DUTs; while that may work to detect a possible bias from the panel it doesn´t help to reveal possible problems in the technical setup, like for example switching noise.
Yes Greiner did some work on cables, but i think he never conducted any double blind tests on this topic.
BTW, the Meyer/Moran on SACD vs. 16Bit/44.1kHz left me with mixed feelings; basically due to the lack of controls and because of not adressing previous work of others with diverging results.
Regarding your training routine on tasting dbt, did you do any similar training on audio dbts?
Wishes
But it happens more often than i´d have thought- another example has been described by Bart Locanthi.
Both examples do raise some questions about the level of sensitivity reached in the tests. Of course there a ways to argue why it more likely that something like these "noise" remains undetected, but it is not really convincing.
Of course, controls are not everything, but they provide something of a basic safety ground.
The positive control does provide evidence for a sufficient sensitiviy level in the test and the negative control ensures, that no other factors are the reason for detected differences.
I remember that sometimes the negative control was only a pretended switch between the DUTs; while that may work to detect a possible bias from the panel it doesn´t help to reveal possible problems in the technical setup, like for example switching noise.
Yes Greiner did some work on cables, but i think he never conducted any double blind tests on this topic.
BTW, the Meyer/Moran on SACD vs. 16Bit/44.1kHz left me with mixed feelings; basically due to the lack of controls and because of not adressing previous work of others with diverging results.
Regarding your training routine on tasting dbt, did you do any similar training on audio dbts?
Wishes
So, you think not choosing id better than choosing?
This isn't the case with me, as I hear noticeable differences with each and every cable I use, or test.
Furthermore, a certain cable sound better connecting my CDP to my pre, while a different cable sounds better connecting my receiver to my pre.
Is this why you cannot accept that some others do hear differences between cables?
How do you choose transducers?
Jakob2 said:Yes, i meant the "relay sound issue" , sometimes the devil is in the details.
Stanley Lipshitz was able to score 100% on the Tiefenbrunn test because he discovered that in one situation (with a SONY PCM-1 in the chain) the background hiss was slightly higher. No other panel member got that and scored basically the same as random guesses.
Jan Didden
Andre Visser said:
So there it is proven that some can detect differences that other will miss.
It is proven that some may use clues that have nothing to do with the difference, if any, in the music, yes
.BTW, in the end Lipschitz revealed his 'trick' and the other also noticed it.
Jan Didden
SY said:
I can't recall that I was talking about your system.
My comment was aimed at the suggestion, made by many, that it is the equipment's fault if cables make a difference. My experience is that I hear differences with the better sounding systems and nothing with the lucky-packet systems.
janneman said:
Same here with the relay sound. When A to B and B to A are different and distinct, it is easy to learn them. The noise issue would, I presume, make DBT between LP and CD difficult to say the least.
I find it disappointing when all too often non-tech oriented folks take an aggresive disinterest in some fairly simple engineering explanations that turn up along the way. Like SY's example of cables causing easily detected frequency response abberations, or cables that make equipment unstable, or "directional" cables that turn out to have hidden "secret" conponents in one or the other termination or have the shield connected at only one end, etc.
Happy Hanukka there Joshua!
<- (Hannukka bush)janneman said:It is proven that some may use clues that have nothing to do with the difference, if any, in the music, yes
BTW, in the end Lipschitz revealed his 'trick' and the other also noticed it.
Jan Didden
I'm sure if I show you my 'trick' to recognise cable differences, you will hear it too.
André
No.
I await some evidence that these differences can actually be heard absent simple engineering issues (e.g., wire too small for a given current, highly capacitive wires causing frequency response errors...), beyond, "Because I said so." I am completely open-minded on the subject, but after thirty years of null results and and evasions, I don't really spend much of my experimentation time worrying about such things. If there's a good piece of work showing the contrary, that will change my priorities.
The Kabbalah.
janneman said:
That was afair a bit later; the "realy sound issue" occured in the first tiefenbrunn-test. Quite late at the end of the session Lipshitz noticed the different relay sound.
scott wurcer said:
Same here with the relay sound. When A to B and B to A are different and distinct, it is easy to learn them. The noise issue would, I presume, make DBT between LP and CD difficult to say the least.
I find it disappointing when all too often non-tech oriented folks take an aggresive disinterest in some fairly simple engineering explanations that turn up along the way. Like SY's example of cables causing easily detected frequency response abberations, or cables that make equipment unstable, or "directional" cables that turn out to have hidden "secret" conponents in one or the other termination or have the shield connected at only one end, etc.
Happy Hanukka there Joshua!
But on the other hand it is amazing, that so often a double blind test confirmed the nill hypothesis, despite the fact that differences existed, which otherwise were confirmed as audible.
Sorry to be repetitive in this point, but is there any other rescue from this than the usage of (positive) controls?
Wishes
SY said:
The cables I test are all commercial ones, well engineered and acknowledged as such by professionals and reviewers. So, the differences I hear aren’t due to engineering issues.
What I get from your post is that you failed to hear differences between different well engineered cables and that you rely on published work.
However, are you sure that others cannot hear such difference, due to actual sound differences between cables?
- Status
- Not open for further replies.