If I could figure out how to get the time away from work, I'd set the test up for him and administer it, just because he's willing and interested in doing things right. Realistically, though, that may be tough to do (I've been home exactly 48 hours in the last three weeks and am gone all this week, too).
Hey Tubeguy!
I was thinking about your offer the other day. Sadly my personal auto situation isn't much better - been driving around with the engine light on for a week or 2 ...or 3 🙂
Another thing that I have wanted in order to do a test in addition to SY's protocol is at least 1 set of cables with very close LCR measurements but different construction and neither of us owns an LCR bridge. Or am I wrong about that and we only need a cheap meter?
Anyway right now is a busy time of year but maybe after all of the drama of the season dies down I can borrow a relatives vehicle and make a day of it.
Key
I was thinking about your offer the other day. Sadly my personal auto situation isn't much better - been driving around with the engine light on for a week or 2 ...or 3 🙂
Another thing that I have wanted in order to do a test in addition to SY's protocol is at least 1 set of cables with very close LCR measurements but different construction and neither of us owns an LCR bridge. Or am I wrong about that and we only need a cheap meter?
Anyway right now is a busy time of year but maybe after all of the drama of the season dies down I can borrow a relatives vehicle and make a day of it.
Key
What were the measured LCR of the wires supposedly tested? How were they measured? TG1954 has (surprise) a tube amplifier which could certainly turn into a tone control given enough variance in the wire parameters.
Wouldn't it be best to ascertain what wires exactly were to be tested, prior to any protocol setup?
Wouldn't it be best to ascertain what wires exactly were to be tested, prior to any protocol setup?
I will try to organise one with friends,based on SY's suggestions.Although I beieve the ones we already done were not much different,this time we'll try to follow SY's even closer.If I remember well,TG1954's was very close too.
All after X-mas though.
I too am looking forward to a technical answer from the smart ones, but from one of the dumb ones (me) I point out that any difference in cables in that situation is completely explained by standard EE theory.
I really think NO-ONE has ever said there will n9t be differences between cables, it is clear that there can be.
The argument really revolves around the explanation for these differences, standard existing knowledge or 'previously undiscovered knowledge, hence on the cutting edge of scientific knowledge'.
That is why you may often see references to well constructed, non pathological etc etc.
All my teachers sent me to the principal's office. Especially Grade 8 typing. Too many hormones, not enough F,D,S,A J,K,L ;
Here is a link to a thread on this forum where I describe my system (its post number 5 of the thread). Warning: there are no turntables, tube amps or esoteric cables in my system. For some that automatically disqualifies me from being acceptably HIFI, but, such is life.🙂
http://www.diyaudio.com/forums/pc-based/153446-digital-room-correction-project.html
Last edited:
Well I guess I was hoping to get a little of everything in. Tubeguy wanted to use his preferred system of tube amplification with a set of full range speakers. I guess I can see how this is setting us up for the whole critical damping issue coming into play. I wanted to also bring either a stereo or quad set of active 2 way monitors and test out line level interconnects.
Could we test for both basically though since even with LCR variations the cable effect is thought to be so small? Basically do the casual test of Ubercool hi tech speaker cables vs non pathological then later try to tighten the controls with the cables electrical parameters. Maybe have us switch so that we try to listen on unfamiliar systems. I dunno it is a bit of a mess at this point haha but we could clean it up I think.
Could we test for both basically though since even with LCR variations the cable effect is thought to be so small? Basically do the casual test of Ubercool hi tech speaker cables vs non pathological then later try to tighten the controls with the cables electrical parameters. Maybe have us switch so that we try to listen on unfamiliar systems. I dunno it is a bit of a mess at this point haha but we could clean it up I think.
Cool. Maybe you guys can figure this out and I will just try my best to comply. Tube guy said he could get his hands on some cool cables. I can do my best to make some matching ICs and Cable sets with what I can get my hands on - I do have a garage full of weird different wire spools I just don't exactly have any trustworthy test equipment to even know where to start.
The cables were manufactured by HMS and they did a special version with identical optics so that they were not distinguishable by VL or participants (page 16, 17 in the pdf).
Sturm used two CD-players marked with L and R, so that the switching person (called VL = Versuchsleiter) knows about the one to start in the various trials, but didn´t know which cable was connected to which cd-player.
From the 6 trials, one was used as the negative control (only a pretended switch), in the remaining 5 the presentation order was changed in an unregular way, but the presentation scheme was the same for all groups.
In run "B" for each group the order of presentation was reversed compared to run "A".
A maximum of 10 participants was allowed for each group, the listeners were not "boxed" so they could see what the neighbour marks on the sheet, but were instructed not to do so in the beginning.
Wishes
So there was no 'X'? The subjects were aware of everytime the source CD player was switched?
jd
It's apparently a sighted paired-preference rather than ABX, so there's no X and all sorts of cues (though it won't be biased for one TYPE of cable as long as that's hidden). And they know how their neighbors voted as well, and I've personally seen the group dynamics in those situations (ETF shootouts are a classic for this).
OK, it wasn't my German, it's just not a well-controlled test. Interesting attempt, and I thank Jakob for sending me the paper and trying to clear up the translation. If I had someone who could make me the "optically identical" cables, I would be much more inclined to do single participant double blind triangle comparisons, with the test subject controlling the switching. MUCH more powerful methodology.
That is the important point and is basically the same as all other people who have to rely on their senses to get their work done, try to do.
And exactly that, what everybody else does in his normal life. 🙂
I think you express it yourself that nobody could say that cables were inaudible in every case in your post #9729.
And that not only wrt to the normal cable interaction with the equipment, but also wrt to RFI or EMC in general.
The burden of proof lies always at the claimant.
Unfortunately the evidence for inaudibility isn´t that strong, as the conclusion is based purely on psychoacoustic and on mediocre (at best) test attempts.
One should think, that this must have been clear for every test, but if you read the documentation of DBTs you´ll find that it is nearly always not so.
It is rare that participants in DBTs got used to the test conditions.
Differences were above the hearing thresholds and confirmed to be audible in general in other tests.
But the more interesting point i had in my mind was the difficulties the listeners had to detect "sameness" ; discrimination tests routinely present the same DUT to the listener, and he has to detect that the presentation was the same.
You´ll find in all these cases that the participants (to 70-75%) will answer with "different" although "same" would have been the correct answer.
Even confirmed by Tom Nousaine.
So, i´d think the conclusion, after knowing this fact, _must_ be to avoid test protocols in which listeners must detect "sameness" (apart from the case where it is used as a control), but as you can see in SY´s test proposal that´s not true. 🙂
Or one would think, if you use the "sameness" test, listeners would be trained under these conditions and there would be a control that shows that listeners were able to detect something under these test conditions, but no, that´s not the case either.
The reason for the above mentioned (somewhat sad case) is simply that the experimentators do have a totally different test goal, and it is not to show that a difference can be detected, but more if the difference is as "earthshaking" as in advance claimed.
Because such an earthshaking difference should be detected in any case.
Might be true (at least the difference between music on and off doesn´t need training under test conditions 🙂 ), but the examples from the viscoglab give reason for the question what amount of a difference might remain undetected in these cases.
Wishes
Last edited:
Yes, Sturm avoided the question for "sameness" (apart from trial 3 which served as a negative control) and opted instead for a paired preference test scheme (forced choice test) quite common in sensory testing. 🙂
Wishes
Last edited:
Could you elaborate a bit about the "all sorts of cues" at work in this specific test? And maybe why any cue would not be averaged out?
At least we know that the presentation order wasn´t known and changed, we know that 112 participants did the test (which made at least 12 different groups of listeners over three days; that´s the minimum afair it wasn´t the maximum of 10 persons in each run, so the number of groups was in fact higher, but is not reported)
I´m sorry, but you must be more specific in that point; which aspects should have been more controlled, because the effect would have favoured one answer over the other?
Yes, it was a nice surprise that at least somebody did something newer, and a bit more relied on the experience in cognitive psychology and avoiding some of the confirmed weaknesses of the usual attempt. (but of course not without some problems- just to indulge my usual fence sitting habit 🙂 )
Nice testing method as it requires a smaller number of trials to reach the same SL, but is it really more powerful wrt to the test goal?
Edit: I see you mentioned the single participant situation; so a smaller number of trials is something to favour, but i´d question the "much more powerful" part as it normally would again rely on the ability of the participant to detect that two of the three were the same?!
Wishes
Last edited:
Yes, if what one wants to determine is "Is there a difference?" the triangle method is far more common in sensory research. Once difference thresholds are established, paired-preference comes into play to see "which one do people like better?" The problem with starting farther down the food chain is that even with nothing switched (and the presenter being unaware of that, which didn't seem to be the case for the negative controls), if the crowd decides earlier on that it likes "A," it will tend to vote for A (we saw that at the transformer shootout this year).
hi ya jacob,
not that it is important in any way, but the points we went over earlier about even loudspeakers needing to be blind tested (if we are to be consistent in all areas) reminded me of a thread I started quite a while back.
It was basically pondering the very same question...how do WE as diyers who think our own speakers are 'fantastic' know how they stack up against the opposition.
I personally (whilst being quite apprehensive really) would find i=t a fascinating experience to do exactly as we were discussing...put our own speakers up against others in a true blind test..
http://www.diyaudio.com/forums/multi-way/129510-how-immune-we-audiphile-wankosa.html
Not quite. I am afraid that I feel *most* people DO think they are unaffected by bias! But apart from that minor sticking point,. yes you are correct.
Poor language, heat of the moment, ..many explanations for the seeming contradictions. I don't think even the most hard headed anti cable guy wold actually argue that cables CANNOT ever be audible would they??? Do I think it is as easy or obvious for people to hear cables as they make out? Nope, not at all. I think that most would in fact NOT be able to identify it blind.
But that does not say that cables cannot have an audible effect. Fully understood by existing EE theory. there was a particular audio critic issue (maybe over two issues?) that showed this very fact.
The point is, it is completely and fully understood (or so I believe) and is not a case of new phenomena, breakthroughs or quantum tunneling.
And NEVER a justification for outrageous cable prices!! hahaha.
But yes, you are correct. I have been guilty of sloppy language indeed. The truth of the matter is that for me cables take on a position of *no* importance in audio. That is my real stance.
That leads to positions like if people concentrate on cables, then by default they are looking in the wrong areas (if audio quality is what we are after), stuff like that.
Oh well, sounds like we are starting to flog dead horses here. I reckon the evidence is pretty conclusive..if there were any truth to the position of cables being a significant factor in audio it would have been shown, even slightly, in 'studies' done.
You and I disagree on how well the tests weer done, and how well they need to be done,. but to me if cable audibility is so fragile that it disappears at the first hurdle, it simply points to it's true importance.
Funny, I seem to have the idea that it is reasonably common??? I often see things like 'the person did it sighted a few times, felt sure that he could identify the cables' yada yada. I could be wrong.
Dunno. Unless there are evidences for these 'other tests' rather than 'everybody knows' we are kinda back to cables...'everybody knows they sound different' until a test is done.
that does not mean you are incorrect, just that it is not clear in these quotes of yours.
Yeah, that is interesting. Is it a damnation of dbts per se tho?? Or simply an indication of expectation, natural traits of hearing differences even tho none exist???
MAYBE IT EVEN GOES BACK TO THE TRAINING POINTS WE HAVE BEEN DISCUSSING? (sorry, caps lock) What I mean is, we wondered recently about 'the person needs to know what he is being tested on'. that makes sense.
BUT, equally (especially in light of what you just raised) maybe all we need to do is suggest something...and the person will hear it! that was shown with Jon Dunlavy holding up cables...and people heard changes.
Tell people we are testing 'analog to digital conversion' and even if truly transparent, people will hear degradation (and you could probably break it down, vinyl guys definitely WILL hear degradation).
DON'T tell them we are doing analog to digital conversion, then which of them will hear it even if it exists??
Interesting stuff.
So, from your perspective, what would you like to see changed/added to any test that SY might be invloved in, ar any test were his protocol might be used. what would you do to make it more robust?
Bconnor, thanks for your link. wow! great stuff. As I said, I though it might be 'my kind of system'! It is...well except for the computer bits that is..I hate computers and am really dumb with them, but apart from that I am impressed.
Consider me invited around next time I am in sydney ok??
not that it is important in any way, but the points we went over earlier about even loudspeakers needing to be blind tested (if we are to be consistent in all areas) reminded me of a thread I started quite a while back.
It was basically pondering the very same question...how do WE as diyers who think our own speakers are 'fantastic' know how they stack up against the opposition.
I personally (whilst being quite apprehensive really) would find i=t a fascinating experience to do exactly as we were discussing...put our own speakers up against others in a true blind test..
http://www.diyaudio.com/forums/multi-way/129510-how-immune-we-audiphile-wankosa.html
Not quite. I am afraid that I feel *most* people DO think they are unaffected by bias! But apart from that minor sticking point,. yes you are correct.
Poor language, heat of the moment, ..many explanations for the seeming contradictions. I don't think even the most hard headed anti cable guy wold actually argue that cables CANNOT ever be audible would they??? Do I think it is as easy or obvious for people to hear cables as they make out? Nope, not at all. I think that most would in fact NOT be able to identify it blind.
But that does not say that cables cannot have an audible effect. Fully understood by existing EE theory. there was a particular audio critic issue (maybe over two issues?) that showed this very fact.
The point is, it is completely and fully understood (or so I believe) and is not a case of new phenomena, breakthroughs or quantum tunneling.
And NEVER a justification for outrageous cable prices!! hahaha.
But yes, you are correct. I have been guilty of sloppy language indeed. The truth of the matter is that for me cables take on a position of *no* importance in audio. That is my real stance.
That leads to positions like if people concentrate on cables, then by default they are looking in the wrong areas (if audio quality is what we are after), stuff like that.
Oh well, sounds like we are starting to flog dead horses here. I reckon the evidence is pretty conclusive..if there were any truth to the position of cables being a significant factor in audio it would have been shown, even slightly, in 'studies' done.
You and I disagree on how well the tests weer done, and how well they need to be done,. but to me if cable audibility is so fragile that it disappears at the first hurdle, it simply points to it's true importance.
Funny, I seem to have the idea that it is reasonably common??? I often see things like 'the person did it sighted a few times, felt sure that he could identify the cables' yada yada. I could be wrong.
Dunno. Unless there are evidences for these 'other tests' rather than 'everybody knows' we are kinda back to cables...'everybody knows they sound different' until a test is done.
that does not mean you are incorrect, just that it is not clear in these quotes of yours.
Yeah, that is interesting. Is it a damnation of dbts per se tho?? Or simply an indication of expectation, natural traits of hearing differences even tho none exist???
MAYBE IT EVEN GOES BACK TO THE TRAINING POINTS WE HAVE BEEN DISCUSSING? (sorry, caps lock) What I mean is, we wondered recently about 'the person needs to know what he is being tested on'. that makes sense.
BUT, equally (especially in light of what you just raised) maybe all we need to do is suggest something...and the person will hear it! that was shown with Jon Dunlavy holding up cables...and people heard changes.
Tell people we are testing 'analog to digital conversion' and even if truly transparent, people will hear degradation (and you could probably break it down, vinyl guys definitely WILL hear degradation).
DON'T tell them we are doing analog to digital conversion, then which of them will hear it even if it exists??
Interesting stuff.
So, from your perspective, what would you like to see changed/added to any test that SY might be invloved in, ar any test were his protocol might be used. what would you do to make it more robust?
Bconnor, thanks for your link. wow! great stuff. As I said, I though it might be 'my kind of system'! It is...well except for the computer bits that is..I hate computers and am really dumb with them, but apart from that I am impressed.
Consider me invited around next time I am in sydney ok??
- Status
- Not open for further replies.