Results can be tabulated for individuals and entire group.panomaniac said:
You have to be careful about mining the data for that handful of the 100 who just happen to get lucky. A two stage design, where the "goldenest" ears are given a second test, would be a good way of getting around that.
Sorry Dumbass, I should have explained it more clearly. At its roots the contention is that top end equipment is capable of revealing differences between cables. I add 'top end' because most would agree Nordost rotary telephone handset cables will be limited in effectiveness by the carbon element. Logically the passive components used in top end equipment don’t obscure those differences. The two examples I used are internal cross connect wire and protection relays, in two different ways.
In the first interconnects are compared blindly to internal cross connect. If no differences are heard either the effect of the interconnect is at or below that of the internal wiring of the device, meaning of no consequence, or the internal cross connect of a device which is capable of revealing cable differences obscures interconnect differences, a logical contradiction. The amp/speaker example uses a similar approach but added an active device to the comparison, making potential differences even easier to catch and failure harder to justify.
ABX testing the ABX comparator is still open to the objection, a stretch for sure, the comparator is too coarse to reveal its own flaws. Comparing the DUT against its internal components leaves no place for logical escape I can see, nor further regress. The same argument I applied to the comparator, when applied to the internal components of high end electronics, is an admission they are too coarse to reveal differences in cables and therefore cables don't alter a system's sound. I thought of anyone you'ld be especially aflutter about that. BTW, I find the certainty with which you think you have me pegged as entertaining as it is misguided.
macgyver10, I agree about the semantics of 'live'. PA events are live performances but you’re really, literally, hearing the output of speakers and not live people and instruments. I also agree with your higher weighting of dynamics, moving your head six inches causes all kinds of frequency response changes. I listen to a great deal of solo classical guitar and piano so PA sound has no relevance. That the majority of recent popular recordings are junk doesn’t really either. The ubiquity of fast food isn’t an argument against good cooking. I find systems which do a good job on material with an honest acoustical reference also seem to work very well playing movies and questionably recorded material so side benefits exist. I’m not convinced by those who claim their systems are so accurate marginal material is now intolerable and suspect instead the cumulative combination of bad material and system flaws just beneath audibility push things over the edge. Some of the most extreme tweaks, Herb Reichert comes to mind, claim the exact opposite of their designs. I’m also not as convinced of the divide between accurate reproduction and better (in the sense of more enjoyable and engaging) reproduction. It’s not my experience.
In the first interconnects are compared blindly to internal cross connect. If no differences are heard either the effect of the interconnect is at or below that of the internal wiring of the device, meaning of no consequence, or the internal cross connect of a device which is capable of revealing cable differences obscures interconnect differences, a logical contradiction. The amp/speaker example uses a similar approach but added an active device to the comparison, making potential differences even easier to catch and failure harder to justify.
ABX testing the ABX comparator is still open to the objection, a stretch for sure, the comparator is too coarse to reveal its own flaws. Comparing the DUT against its internal components leaves no place for logical escape I can see, nor further regress. The same argument I applied to the comparator, when applied to the internal components of high end electronics, is an admission they are too coarse to reveal differences in cables and therefore cables don't alter a system's sound. I thought of anyone you'ld be especially aflutter about that. BTW, I find the certainty with which you think you have me pegged as entertaining as it is misguided.
macgyver10, I agree about the semantics of 'live'. PA events are live performances but you’re really, literally, hearing the output of speakers and not live people and instruments. I also agree with your higher weighting of dynamics, moving your head six inches causes all kinds of frequency response changes. I listen to a great deal of solo classical guitar and piano so PA sound has no relevance. That the majority of recent popular recordings are junk doesn’t really either. The ubiquity of fast food isn’t an argument against good cooking. I find systems which do a good job on material with an honest acoustical reference also seem to work very well playing movies and questionably recorded material so side benefits exist. I’m not convinced by those who claim their systems are so accurate marginal material is now intolerable and suspect instead the cumulative combination of bad material and system flaws just beneath audibility push things over the edge. Some of the most extreme tweaks, Herb Reichert comes to mind, claim the exact opposite of their designs. I’m also not as convinced of the divide between accurate reproduction and better (in the sense of more enjoyable and engaging) reproduction. It’s not my experience.
That is odd, since I think there's no more than a millimeter's distance between our views and I've gotten no flak.
In any case, the way the ABX tests generally work (I assume you've read Lipshitz's and Nousaine's papers) is that the subject does the test unblind with the AB switch and confirms that he/she can indeed distinguish A from B. Once that's done, the randomization of X occurs. This is similar to the way dowsers are tested- if you don't have a copy of Randi's "Flim-Flam," a seriously good and entertaining book, run right out etc. In any case, to invoke "box error" in these situations, one must posit that the A and B switching always seem to be more transparent than the identical circuitry in the X leg...
rdf said:
NOW we're getting somewhere. I was worried that you guys had gotten stuck in the trees and couldn't find your way out of the forest.
rdf said:
Right on! Though I do understand what they mean, I always smile and say to myself "bad crossover."That has been my experience with really top notch systems. Bad recordings are still bad, but they are rarley "unlistenable." A great system will not only pull magic out of great recordings, it will pull real music out of bad ones. Tricks and flaws become obvious, but the music is still there and can be delightful.
That has been my experience excactly. Odd that so many people don't find this. Must be the crossovers again.
rdf said:
Well, I'll admit to being a bit of a hypocrite here. I'll still always design my components, and system based on trying to get the most "accurate" results. By that I mean flattest possible FR, attempt to keep x-over influences out of the key midrange frequencies, etc. etc. (I primarily concern myself with loudspeaker design because I like the acoustic aspect of their design). I'll also concern myself about attempting to keep FR flat both on and off axis, designing my x-overs for BOTH their on-axis and 30deg off axis response. As well as keeping enclosures non-resonant, with front baffles (hopefully) designed to reduce diffraction effects and time-align drivers. I like the design approaches of Thiel and B&W and Wilson for example.
I guess what I'm really getting at is that my design philosophy has shifted from what would classically be considered "accurate" for a loudspeaker such as; extremely flat FR, high order x-overs to reduce inter-driver overlap etc. to a more sensitivity and dynamics centric approach.
I find that in speaker design, particularly, these goals can be quite mutually exclusive because often the drivers with the smoothest response (ie Dynaudio) have relatively low sensitivity (85dB/1W/1M), whereas the really sensitive drivers can be quite choppy in FR, but give you 100dB or higher sensitivity. Then I'll generally try and use the natural envelope of the driver to augment a lower order x-over. I can often achieve a 3rd order response, with only a 1 or 2 order x-over. I'm attempting to resistively pad nothing if possible, and not lose a single watt to heat unnessesarily.
What this all means is that different design philosophies result in making different comprimises, and I've yet to come across a system that can really do the entire range of music genres equally well.
There's a place for electrostats...and horns...and domes and Heil AMT's perhaps even Mangers, but stuff them all in the same box, and you'll end up with a dog's breakfast.
Currently (purely philosophically) my leanings are towards the high sensitivity dynamic very wide range variety of loudspeaker...that is as "accurate" as those types of drivers will allow.
SY said:
Had to think about that one. It doesn't appear to have been done, or explicitly stated, in the BAS tests or the Tiefenbrun run. It would also seem impractical and poor procedure (due to the peer/expectation pressure they discuss in the IQ example) in the group tests commonly published. I agree though if a solo subject claimed they could clearly hear a difference unblind and failed the blind portion they were deluding themselves in the first half. (Or lost it under pressure in the blind one. I know it's weak, just being complete.) In isolation it's somewhat a null result still consistent with the ABX kit obscuring differences. Everyone would just agree the subject imagined them and disagree on the root cause. Throwing light on it requires a meta survey that includes subjects who declined to proceed on the grounds they couldn't hear a difference in the unblind portion, with ideally further testing of those subjects using a different protocol. Consent does make a much stronger argument in favour of ABX tests conducted with all other extraneous factors removed though.
macgyver10, I find it funny that coming from somewhat opposite approaches we arrive at the same general preference in loudspeakers. A pair of Apogee Centaur Minors sit in storage while Tannoy Monitor Gold 12s do service as my daily listen.
rdf said:
The philosophy may differ in some details, but ultimately we both really just want to hear some great music that moves us.
In that regard, it's not too surprising that we'd end up in the same place!
I haven't had the opportunity to hear either of those speakers...although I think a friend made some new cabinets for the Tannoy 12 ( concentric tweeter? large displacement ported enclosure?)
The ribbon would probably max out in SPL close to where the Tannoy starts a 1 or 2 watts I would imagine?
Just two matters at this time.
We researchers where I worked had funny lunch-time de-stressors. At a time a coin was tossed over 150 times, and it was recorded (a) the number of times heads or tails came up, and (b) the predictions of everybody (4 of us) as to what will come up. Not unexpectedly the coin predictions reached close to 50/50 in the end. But sometimes the same side will come up 7 out of 8 times in succession. Likewise, sometimes I guessed right with similar accuracy and at other times wrong some 7 out of 8 times.
The lesson is that if any of those increments were to have been taken alone, it could have been interpreted that I can accurately predict the situation 7 out of 8 times - I must have some supernatural powers! It is clear from this that if one wants any degree of accuracy quite a large number of samples needs to be conducted. I am sure that this is not new to anyone (another member has already mentioned this), but it does not seem to be observed in some blind tests I have seen.
Then a lot of valuable comment has been given here regarding tests, but for me it is simply a case of someone reporting a "definite difference" between cables, only to loose that ability in blind tests. That surely already shows something?
The further point about switching times between DUTs is certainly valid - but under which conditions were these claimed "definite/dramatic/fantastic differences" experienced in the first place? Cables were then "changed", presumably by hand and without a rush. Sometimes they were taken home and listened to for some time at a time. Again, if the difference/improvement was experienced with the certainty often proclaimed, why does one suddenly need laborotary techniques to establish it when people do not know what they are listening to? To me this qualifies the hearing faculty as unreliable even under such conditions, which makes it the main variable. Refining test procedures to be orders better is necessary under academic conditions, not to prove that I do not need expensive cables or that there is indeed a worthwhile audible difference.
One does not dispute the standards mentioned here by many, but if such a degree of refinement is required to differentiate, are we not concentrating on the minute and ignoring the substantial (room acoustics, variation in hearing conditions, etc.) as pointed out here by several members in the past?
For my money, if there is a clear difference between what a listener finds when listening under known vs. unknown conditions (I mean different products) and the situation is further overwhelmingly supported by basic electronic principles, then the situation is clear to me. I spend my further energy in other more meaningful directions.
Regards.
PS: Regarding claims of exotic cables, I would humbly submit that I am still ahead. Nobody has mentioned of how cables must wear out with all those electrons racing round the corners at almost the speed of light. One needs to replace those worn cables every so often! Watch the advertisements and look out for my comment: "I told you so!"
We researchers where I worked had funny lunch-time de-stressors. At a time a coin was tossed over 150 times, and it was recorded (a) the number of times heads or tails came up, and (b) the predictions of everybody (4 of us) as to what will come up. Not unexpectedly the coin predictions reached close to 50/50 in the end. But sometimes the same side will come up 7 out of 8 times in succession. Likewise, sometimes I guessed right with similar accuracy and at other times wrong some 7 out of 8 times.
The lesson is that if any of those increments were to have been taken alone, it could have been interpreted that I can accurately predict the situation 7 out of 8 times - I must have some supernatural powers! It is clear from this that if one wants any degree of accuracy quite a large number of samples needs to be conducted. I am sure that this is not new to anyone (another member has already mentioned this), but it does not seem to be observed in some blind tests I have seen.
Then a lot of valuable comment has been given here regarding tests, but for me it is simply a case of someone reporting a "definite difference" between cables, only to loose that ability in blind tests. That surely already shows something?
The further point about switching times between DUTs is certainly valid - but under which conditions were these claimed "definite/dramatic/fantastic differences" experienced in the first place? Cables were then "changed", presumably by hand and without a rush. Sometimes they were taken home and listened to for some time at a time. Again, if the difference/improvement was experienced with the certainty often proclaimed, why does one suddenly need laborotary techniques to establish it when people do not know what they are listening to? To me this qualifies the hearing faculty as unreliable even under such conditions, which makes it the main variable. Refining test procedures to be orders better is necessary under academic conditions, not to prove that I do not need expensive cables or that there is indeed a worthwhile audible difference.
One does not dispute the standards mentioned here by many, but if such a degree of refinement is required to differentiate, are we not concentrating on the minute and ignoring the substantial (room acoustics, variation in hearing conditions, etc.) as pointed out here by several members in the past?
For my money, if there is a clear difference between what a listener finds when listening under known vs. unknown conditions (I mean different products) and the situation is further overwhelmingly supported by basic electronic principles, then the situation is clear to me. I spend my further energy in other more meaningful directions.
Regards.
PS: Regarding claims of exotic cables, I would humbly submit that I am still ahead. Nobody has mentioned of how cables must wear out with all those electrons racing round the corners at almost the speed of light. One needs to replace those worn cables every so often! Watch the advertisements and look out for my comment: "I told you so!"
Johan - you just "unintentionally" revealed the next marketing strategy of the cable companies.
Instead convincing new customers - which is rather expensive as to the ad budget - convince the ones you have to buy more and more often - good cables just wear out. To keep the advantage - like tires, every so many listening hours you gotta exchange. Maybe they even give you a core fee for the old one.
Instead convincing new customers - which is rather expensive as to the ad budget - convince the ones you have to buy more and more often - good cables just wear out. To keep the advantage - like tires, every so many listening hours you gotta exchange. Maybe they even give you a core fee for the old one.
Hurry folks, this week only
2 CABLE RE-CORES FOR THE PRICE OF ONE !!
That's right we will re-core your old tired cables with brand new copper / silver / gold / manganese / cores...and the best part you get the other cable re-cored free.
Quasi Cable Re-Corers Pty Ltd will give you your cables back sounding as good as the first day you bought them. Money bacj guarantee if they don't.
Send AMEX / VISA / or cheque rignt now and get your terminals re-connected for half price...THAT'S RIGHT HALF PRICE.
Call now...operators standing by.
2 CABLE RE-CORES FOR THE PRICE OF ONE !!
That's right we will re-core your old tired cables with brand new copper / silver / gold / manganese / cores...and the best part you get the other cable re-cored free.
Quasi Cable Re-Corers Pty Ltd will give you your cables back sounding as good as the first day you bought them. Money bacj guarantee if they don't.
Send AMEX / VISA / or cheque rignt now and get your terminals re-connected for half price...THAT'S RIGHT HALF PRICE.
Call now...operators standing by.
Sorry guys, prior art.
http://www.diyaudio.com/forums/showthread.php?postid=1039377#post1039377
My legal department will be calling you shortly.
http://www.diyaudio.com/forums/showthread.php?postid=1039377#post1039377
My legal department will be calling you shortly.
rdf said:
Ah...not so Mr RDF, please refer to a number posts starting from http://www.diyaudio.com/forums/showthread.php?postid=797965#post797965 post #440 where you you find some collbration between my self, poobah, anatech and others about fillied cables concepts. Discussions progresses to company structures and who got particular secretaries.
But I'm sure your legal department will pick that up and avise you accordingaly.
Best Regards
Quasi
CEO - Cable Fillers & Technologies Limited
now things get serious...just don't let the kiddies play with those...
But RDF - we are not developing the concept of refillable cables - that would be uneconomical from the profit margin view - no, we advocate replacement of cables worn through by long listening sessions. The secret is - how to repaint the trade ins and resell them as new and upgraded - that's where the cash flows...
- Status
- Not open for further replies.