Andy G said:
And not just on the statistics, but on the listener too.
This style of test is far removed from a real listening experience, but many seem to think that they provide the whole and only truth, and hold the results of such tests as actually meaning something concrete. And then they class themselves as being 'objective' ???
I have this funny feeling of deja vu here, but let us accept your criticisms of any tests done to date (we know they're wrong, cause they found no difference right?)
OK, knowing the faults scientifically as you do, tell us the way YOU would do a test that overcomes these fatal flaws.
Heck, let's even take a step back. Is it possible at all to test audio differences??
If it is, how/what is required to be done so that YOU would say, 'now that test was done correctly as far as I can determine'.
In other words, why don't you do something that might materially advance the discussion instead of carping from the sidelines?
I don't pretend to know the "technical" part of this issue but I just want to ask...why don't you favor the loudest of any cable?
Doesn't the loudest prove that the greatest signal is getting through? To whatever device you are powering?...the rest is a matter of conjecture, the other components strengths or weaknesses. The nuances of which some speak...it is merely conjecture and surely of the mind's making.
I've spent an embarrassing amount on cables, then, as I bought my first Mogami W2534 in length I was surpised by the result. To date I've not found a better cable and it can be bought for a pitance of the "high end" cables. I even use it to power mid/high drivers. I just think it's fantastic, simple, clean, basic cable. Clear and consise throughput.
Go to Redco.com, they've got a great price by the foot.
If you think the connectors are that important, buy the Vampire 800C RCA's. So, you build the best for $40-50 per cable. Gees, that's nothing in this rediculous cable world. Or, better yet, set your speakers with Speakon connectors. As for me, there will no longer be "speaker cables". Onboard amps, Jensen balanced transformers (yes I know they're expensive..still far less than the rediculous speaker/input cables), LR4 divided active networks and separately powered drivers have proven their superiority to me.
Then, I added a Bybee to my Digital cable (made of the Mogami W2534) and I was surprised by how pleasant they were. My own, homemade Mogami W2534, Bybee, Vampire cables are beyond all others that I've tried. Heck, I don't even know there ohm values. I stopped there. So much cheaper too. Louder, clearer, fuller, and easier to deal with cables than I've ever experienced.
There is indeed better out there. Likely, the best in the basis of electrical transmission. Simple, louder, purer, cleaner and impecaby precise. Always, of the most basic
Nobody can seem to explain the Bybee. Yet, I think it is so. They are just exceptional. Try it in your signal transmission. I haven't tried them on speakers. Some say they are incredible.
Bet not on the "better copper" therum. beleieve in what makes sense. And, when something like the Bybee comes along and proves it's validity...embrace it! I know I'm a believer. I still have no idea why.
Good luck guys.
T
Doesn't the loudest prove that the greatest signal is getting through? To whatever device you are powering?...the rest is a matter of conjecture, the other components strengths or weaknesses. The nuances of which some speak...it is merely conjecture and surely of the mind's making.
I've spent an embarrassing amount on cables, then, as I bought my first Mogami W2534 in length I was surpised by the result. To date I've not found a better cable and it can be bought for a pitance of the "high end" cables. I even use it to power mid/high drivers. I just think it's fantastic, simple, clean, basic cable. Clear and consise throughput.
Go to Redco.com, they've got a great price by the foot.
If you think the connectors are that important, buy the Vampire 800C RCA's. So, you build the best for $40-50 per cable. Gees, that's nothing in this rediculous cable world. Or, better yet, set your speakers with Speakon connectors. As for me, there will no longer be "speaker cables". Onboard amps, Jensen balanced transformers (yes I know they're expensive..still far less than the rediculous speaker/input cables), LR4 divided active networks and separately powered drivers have proven their superiority to me.
Then, I added a Bybee to my Digital cable (made of the Mogami W2534) and I was surprised by how pleasant they were. My own, homemade Mogami W2534, Bybee, Vampire cables are beyond all others that I've tried. Heck, I don't even know there ohm values. I stopped there. So much cheaper too. Louder, clearer, fuller, and easier to deal with cables than I've ever experienced.
There is indeed better out there. Likely, the best in the basis of electrical transmission. Simple, louder, purer, cleaner and impecaby precise. Always, of the most basic
Nobody can seem to explain the Bybee. Yet, I think it is so. They are just exceptional. Try it in your signal transmission. I haven't tried them on speakers. Some say they are incredible.
Bet not on the "better copper" therum. beleieve in what makes sense. And, when something like the Bybee comes along and proves it's validity...embrace it! I know I'm a believer. I still have no idea why.
Good luck guys.
T
Spiking tests are commonly administered to sort judges in wine competitions, for qualifying panelists in taste tests, and so forth. So yes, it's done all the time. While I haven't spiked pinot noir with cabernet, I have certainly done tests with viognier additions at the 1% level to syrah.
pano, please, please, don't conflate ABX with controlled testing. ABX is only one possible format, and less powerful (statistically) than others, e.g., triangle. ABX results are indeed valid, but there are other formats, equally valid, that may show better sensitivity to some phenomena. I'm still waiting for the critics of controlled testing to offer an alternative or indeed even a scintilla of evidence of why controlled testing works in every other area of sensory research but not high end audio.
http://www.stereophile.com/asweseeit/705awsi/
What John Atkinson thinks of blind testing.
Sort of half of no explanation why it ain't working...just a "feeling" that he was let down by his own "objectivist" stance.
Not considering that other factors might have contributed to his long term dislike, no - the blind test was wrong.
BTW - I have a 405, and I can't figure out what should be wrong with it. Ran it against Bryston 4BST on Kef's 104/2 - no diff at fairly high listening levels.
SY said:
No no, not at all. It just that the ABX test is so often mentioned. I would favor other methods, controlled of course, since the ABX seems to be so unliked by some.
Good to know of the spiking tests. So a good judge is able to sense a spike on the order of 1%? Of course spike it with 1% ammonia, and we could probably all tell!
A-K. My experience with the 405 was completely other.
But it was a really tough load - double stacked Quad ESL. But that's for another thread!
SY said:
Had some Bouree 1978 Latricières Chambertin a couple of weeks ago that was spiked with considerably more than 1%.
Scott, I remember being served one of the notorious Dr. Barolet burgs, a '29 Gevrey Chambertin. The Other Stupid took a swig and gave me a funny look. We both burst out laughing and said, "Wow, grenache really DOES age!" I'd guess Gigondas, but no way to find out for sure.
16hz lover said:
slight differences between the two loudspeakers?
or HUGE difference in placement within the room?
neededandwanted said:
okay,to refrain, i mean recap:
i set my preamp to mono, and wired the left channel with 16 gauge braided '100% copper' radioshack speaker wire, and the right channel was wired with 12 gauge solid core housing wire.
i then asked an uninterested 14 yr. old which side sounded better, and without hesitation the solid core side was chosen. i agreed wholeheartedly.
much clearer on the right side. more, tighter, faster, deeper, ....unhhhh.
some things are too sexy for science.
cable makes a difference, but it's exceedingly wrong to pay more than a few bucks for the 'right' wire.
and, p.s., my speaker channels/room position are ambidextrous, and the listening was done with the speaker about a foot and a half from the speaker, effectively negating the room's 'input'.
Duke, did you by any chance swap cables from left to right and try again?
If so, did the improvement follow the cable?
If so, did the improvement follow the cable?
Fair enough. Didn't mean to attack your experiment at all.
It is just that when upgrading a system, I always look at fixing the weakest link in the chain. So far the weakest link has never been speaker wire.
If someone told me that his left speaker sounded better than his right one, I would assume that the differences were:
[1] the room (huge factor)
[2] driver differences (bigger than anyone is ever willing to admit, except when they are hand-matching pairs)
[3] binding post / termination issues (always a point of SOME impedance mismatch, unless you are soldering directly to the voice coil, and then still is)
As a secondary point, being able to hear a difference is something beyond being able to tell which sounds better. Or better put, MANY inaccuracies will sound "better" for a short time. Add a bit of extra high end or low end equalization to one side of an AB comparison and ask people which sounds better and they will invariably pick the one less accurate.
I am not saying that wires don't matter. I DO think the interconnects on the low level signal path need help more than the speaker level path though.
Since you have experienced a difference with solid core wire, I invite you to try some 10 cents per foot Romex and see how that sounds.
I believe you do have a difference between your two wire sets. And who knows? Maybe one set sounds better at high volumes, and the other at low volumes.
I agree, wires can make a difference. Put it this way, if you set out to make BAD speaker cables could you? Sure.
For me, here's the REAL question for this thread:
If the difference between your two sets of wires is so signifigant that your wife and daughter can hear it, does anyone really believe it is unmeasureable?
What I really tire of is claims by manufacturers that the differences are huge, but still not able to be measured by any technology currently known to man. They spend many pages of fluff explaining how the human ear is an instrument unmatched by anything mankind has created, etc. while trumpeting the clarity you'll enjoy when you purchase $1000 worth of little ladders that hold your speaker cables off the floor.
Add to this the positive reviews of these magic beans by the magazines that count on them for advertising dollars and you have a perfect storm of nonsense, or non-science anyway.
We didn't come as far as we have by simply listening and not measuring.
It is just that when upgrading a system, I always look at fixing the weakest link in the chain. So far the weakest link has never been speaker wire.
If someone told me that his left speaker sounded better than his right one, I would assume that the differences were:
[1] the room (huge factor)
[2] driver differences (bigger than anyone is ever willing to admit, except when they are hand-matching pairs)
[3] binding post / termination issues (always a point of SOME impedance mismatch, unless you are soldering directly to the voice coil, and then still is)
As a secondary point, being able to hear a difference is something beyond being able to tell which sounds better. Or better put, MANY inaccuracies will sound "better" for a short time. Add a bit of extra high end or low end equalization to one side of an AB comparison and ask people which sounds better and they will invariably pick the one less accurate.
I am not saying that wires don't matter. I DO think the interconnects on the low level signal path need help more than the speaker level path though.
Since you have experienced a difference with solid core wire, I invite you to try some 10 cents per foot Romex and see how that sounds.
I believe you do have a difference between your two wire sets. And who knows? Maybe one set sounds better at high volumes, and the other at low volumes.
I agree, wires can make a difference. Put it this way, if you set out to make BAD speaker cables could you? Sure.
For me, here's the REAL question for this thread:
If the difference between your two sets of wires is so signifigant that your wife and daughter can hear it, does anyone really believe it is unmeasureable?
What I really tire of is claims by manufacturers that the differences are huge, but still not able to be measured by any technology currently known to man. They spend many pages of fluff explaining how the human ear is an instrument unmatched by anything mankind has created, etc. while trumpeting the clarity you'll enjoy when you purchase $1000 worth of little ladders that hold your speaker cables off the floor.
Add to this the positive reviews of these magic beans by the magazines that count on them for advertising dollars and you have a perfect storm of nonsense, or non-science anyway.
We didn't come as far as we have by simply listening and not measuring.
nope. didn't need to, really. but, what i did do, was wire them both with the 12 gauge, afterward, and checked to see they were now matched, which they were.
i also was using some leftover 12 guage for the initial test, so, i had to put the old speaker cable back on until i had a chance to go to the hardware store.
i was reading a thread here somewhere (not this one), and someone said solid core was good for speaker wire, which is something i never heard before, so i tried it, and was very pleased to hear a 'big' improvement. the degree of difference is something that would be hard to measure.
which brings me to another question. when people measure their speakers with a microphone, how do they compensate for the unique characteristics of the mic itself? wouldn't different mics respond differently?
and, do these types of measurements usually use white or pink noise as the source, or do they use sine wave sweeps and tones?
i also was using some leftover 12 guage for the initial test, so, i had to put the old speaker cable back on until i had a chance to go to the hardware store.
i was reading a thread here somewhere (not this one), and someone said solid core was good for speaker wire, which is something i never heard before, so i tried it, and was very pleased to hear a 'big' improvement. the degree of difference is something that would be hard to measure.
which brings me to another question. when people measure their speakers with a microphone, how do they compensate for the unique characteristics of the mic itself? wouldn't different mics respond differently?
and, do these types of measurements usually use white or pink noise as the source, or do they use sine wave sweeps and tones?
SY said:
As stated before, i understand why some think so, but i´m afraid it can´t work that way.
Unfortunately it is possible that someone just listening nearby while doing other things (the famous wife in the kitchen 🙂 ) is able to detect something that another person, while trying hard to listen carefully, does not notice.
This does depend on the state of awareness a person reaches.
You´re still assuming that someone is able to listen as good/reliable under a specific blind test condition as he does during his normal (sighted) listening.
That assumption might be right but is unverified. And there is strong evidence that the state of awareness is strongly affected by knowing to participate in a test and the more controlled a test is the more affection takes place.
Normal cure to that is to train under these controlled specific conditions to get used to it.
And finally to use a positive control to _know_ whether his listening is as good as it should be.
Another possibility would be to prevent the participant from knowing about his participation in a test.
Sometimes not easy to achieve but working quite well if possible.
planet10 said:
In every field using double blind tests there is a constant discussion about the quality of the results.
The blinding does only remove one bias mechanism, but does not lead to useful results by itself.
Wine competitions are normally preference tests and it is an intersting fact, that a certain wine does win a price at one competition and don´t get an award at another competition.
Robert T. Hodgson did an investigation on results of the California State Fair wine competition:
Robert T. Hodgson, An Examination of Judge Reliability at a major U.S. Wine Competition
Journal of Wine Economics, Volume 3, Issue 2, Fall 2008, Page 105-113
He found remarkably inconsistency in the scores of quite a lot judges if tasting the same wine several times.
As no controls (tasting the same wine several times is a negative control) were incorporated normally this were not obvious; this does not mean, that the blind tests were useless, but is just another example that blinding itself doesn´t garantuee useful (means valid, reliable and objective) results.
One might posit the hypothesis that being aware of participating in a test will increase awareness.
In our hunter/gatherer past stress in threatening surroundings increased the level of awareness - it had to, otherwise the mechanism would not have been conserved, but would have led to the extinction of those not able to function well and identify source of danger faster under stress.
Provided you agree that tests are/can be stressful situations.
My hypothesis is therefore that relaxed testing is not able to lead to finding those minor differences, as awareness is dulled.
@ audio-kraut,
there´s nothing wrong with your hypothesis, now do some test to verify it. 🙂
Just kidding- it´s all about using controls on sufficient sensitivity levels. If you´re argueing that a difference can´t be audible because otherwise it had to be heard in a double blind test, than you have to show that the test was sufficient sensitive. It is that simple.
If you didn´t supply the controls than of course we still can assume, can speculate (as we do constantly in all these threads around the world 🙂 ), but nothing more.
And normally, everyone who has conducted double blind tests himself, knows that the participants often don´t detect even quite large differences until they get used to the blind test scheme.
there´s nothing wrong with your hypothesis, now do some test to verify it. 🙂
Just kidding- it´s all about using controls on sufficient sensitivity levels. If you´re argueing that a difference can´t be audible because otherwise it had to be heard in a double blind test, than you have to show that the test was sufficient sensitive. It is that simple.
If you didn´t supply the controls than of course we still can assume, can speculate (as we do constantly in all these threads around the world 🙂 ), but nothing more.
And normally, everyone who has conducted double blind tests himself, knows that the participants often don´t detect even quite large differences until they get used to the blind test scheme.
My argument was aimed to those that claim that "blind" testing leads to skewed "negative" results confirming the null hypothesis due to stress, when it is the opposite that stress leads, and has to lead as viewed from an evolutionary point, to increased awareness.
I am sure you will find enough literature confirming this - after all, after our stressful existence in the Savannah chased by predators - we are still around arguing the point.
A "non switching test" where 50% of participants still hear a difference could be used to increase sensitivity - remove those 50% and test with the remainder.
The increase in the sensitivity imo can only be by statistical means, increasing the size of the sample.
The test after all is simple enough - different or not.
But my statistical skills are quite rusty - the last time I used them was 30 years ago....
I am sure you will find enough literature confirming this - after all, after our stressful existence in the Savannah chased by predators - we are still around arguing the point.
A "non switching test" where 50% of participants still hear a difference could be used to increase sensitivity - remove those 50% and test with the remainder.
The increase in the sensitivity imo can only be by statistical means, increasing the size of the sample.
The test after all is simple enough - different or not.
But my statistical skills are quite rusty - the last time I used them was 30 years ago....
jackinnj said:
Not guilty. Two co-workers were staring at an unexplainable result Friday. They had two 60GHz 1 meter interconnects ($400@) on the Q and Qbar outputs of a circuit and got 200pS skew on the scope. I noticed that they were guaranteed for length but the actual prop delay depended on the dielectric constant., sure enough they had been made a year appart from different batches of cable.
jlsem said:
Well, I did have to take my wife to the airport at 4AM Friday.
Maybe my choice of words was poor. I simply wanted to empathize with those confronted with a panel of skeptics trying to prove to them that they (the skeptics) actually do hear a difference. This is a lot different than a panel of those that believe in the difference trying to prove it, basically “on the skeptics turf by the skeptics rules”.
- Status
- Not open for further replies.