Proposed Amplifier Differentiation Experiment - Page 2 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 102Likes
Closed Thread
 
Thread Tools
post #31 of 227 Old 02-15-2016, 10:49 AM
FMW
AVS Forum Special Member
 
FMW's Avatar
 
Join Date: Jun 2007
Posts: 7,480
Mentioned: 13 Post(s)
Tagged: 1 Thread(s)
Quoted: 1654 Post(s)
Liked: 1710
OP, it has been done several times. If it helps the reality is that not all amplifiers sound the same but most of them do.
FMW is offline  
Sponsored Links
Advertisement
 
post #32 of 227 Old 02-16-2016, 06:14 AM
AVS Forum Special Member
 
kucharsk's Avatar
 
Join Date: Feb 2004
Location: Louisville, CO
Posts: 6,449
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1765 Post(s)
Liked: 1042
I'm glad you all feel so smug, but the bottom line is you enjoy your audio systems and I'll enjoy mine.

The components I've selected have all made distinct, audible improvements to my system and I will continue to select equipment that way.
Jady Jenkins likes this.
kucharsk is offline  
post #33 of 227 Old 02-16-2016, 08:06 AM
FMW
AVS Forum Special Member
 
FMW's Avatar
 
Join Date: Jun 2007
Posts: 7,480
Mentioned: 13 Post(s)
Tagged: 1 Thread(s)
Quoted: 1654 Post(s)
Liked: 1710
Quote:
Originally Posted by kucharsk View Post
I'm glad you all feel so smug, but the bottom line is you enjoy your audio systems and I'll enjoy mine.

The components I've selected have all made distinct, audible improvements to my system and I will continue to select equipment that way.

I don't think anybody is criticizing that. I think they are just trying to get you to understand why you hear some of the audible difference you hear. Obviously you don't accept it so your approach makes perfect sense under the circumstances.
FMW is offline  
Sponsored Links
Advertisement
 
post #34 of 227 Old 02-16-2016, 09:59 AM
 
LFEer's Avatar
 
Join Date: Jun 2013
Posts: 1,243
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 818 Post(s)
Liked: 532
Quote:
Originally Posted by kucharsk View Post
I'm glad you all feel so smug, but the bottom line is you enjoy your audio systems and I'll enjoy mine.

The components I've selected have all made distinct, audible improvements to my system and I will continue to select equipment that way.
What you are forgetting is that this is internet forum. What you or others do with your / their own sound system at home is your / their own business but what you are seeing is what typically happens on internet forums. Some post claims and others chime in or challenge. If you're not into such thing, then you know where not to go.
kucharsk likes this.
LFEer is offline  
post #35 of 227 Old 02-16-2016, 05:02 PM
AVS Forum Special Member
 
kucharsk's Avatar
 
Join Date: Feb 2004
Location: Louisville, CO
Posts: 6,449
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1765 Post(s)
Liked: 1042
Quote:
Originally Posted by FMW View Post
I don't think anybody is criticizing that. I think they are just trying to get you to understand why you hear some of the audible difference you hear. Obviously you don't accept it so your approach makes perfect sense under the circumstances.
I hear the audible differences because some equipment excels at musical reproduction more than others.

I get that many don't hear or don't buy that there can be differences, and I chalk that more up to the testing methodology.

I'm afraid there may well never be a resolution to that, and that's fine.

Many here believe the folks at high-end audio companies go home at night and rub their hands with glee over what over-priced crap they will get people to buy next, and I don't see that at all. I've talked to these people and know many of them personally and they do what they do because they believe in better sound. Perhaps all audiophiles are insane or merely deluded, but ultimately each person decides whether they hear improvements and whether they're worth their hard-earned money.

I now return you to your regular ABX discussion.
W4RLORD likes this.
kucharsk is offline  
post #36 of 227 Old 02-16-2016, 05:05 PM
AVS Forum Special Member
 
A9X-308's Avatar
 
Join Date: Mar 2008
Location: Australia
Posts: 8,321
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 1792 Post(s)
Liked: 1484
Quote:
Originally Posted by kucharsk View Post
Perhaps all audiophiles are insane or merely deluded
Not all, just the vast majority. I may find one of these this lifetime. I may also find Bigfoot.
A9X-308 is offline  
post #37 of 227 Old 02-16-2016, 06:35 PM
 
LFEer's Avatar
 
Join Date: Jun 2013
Posts: 1,243
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 818 Post(s)
Liked: 532
Quote:
Originally Posted by kucharsk View Post
I hear the audible differences because some equipment excels at musical reproduction more than others.
That's very true for speakers.
Quote:
I get that many don't hear or don't buy that there can be differences, and I chalk that more up to the testing methodology.
You've never done a level matched DBT, have you?
Quote:
I'm afraid there may well never be a resolution to that, and that's fine.
The resolution has been made. It's just that some don't accept or believe it. Some don't even know about it.
Quote:
Many here believe the folks at high-end audio companies go home at night and rub their hands with glee over what over-priced crap they will get people to buy next
What evidence is that based on?
LFEer is offline  
post #38 of 227 Old 02-16-2016, 07:52 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,626
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 607 Post(s)
Liked: 574
Quote:
Originally Posted by kucharsk View Post
Many here believe the folks at high-end audio companies go home at night and rub their hands with glee over what over-priced crap they will get people to buy next, and I don't see that at all.
No doubt some do and are just unscrupulous people who happen to have landed in the audio world as opposed to some other way to take people for money.

But I'd guess most aren't. I believe they probably share as a passion music and audio just like you and me, mean well, and are entirely convinced that what they hear is physical, and what they sell improves that. There's also the engineering aspect... and as a former engineer I completely get that. Sometimes we do things because they are conceptually elegant, or particularly clever solutions, or push the boundaries of what is achievable even if far removed from audible differences. But we also have to keep in mind that its easy to get married to your creations and become convinced that these aren't just tiny measurable differences we're chasing, but maybe just maybe audible ones too.

Hell, as objective as I am, I can fall to the same human tendancies. Recently have been evaluating multichannel amplifier solutions, and it's very very easy to start comparing harmonic and IM distortion levels like 0.005% vs 0.002% and thinking "dang, that's more than twice as good, and will only cost me a little bit more..." And you know what? Its as likely as not that I decide to spend more on a measurable difference I know full well isn't audible. But that's OK, because I do so fully informed, I understand what it is that appeals to me, and I know what to expect and what not to expect.

I don't think you personally mean harm by making recommendations based on your flawed perceptions (all of our perceptions are flawed, surprisingly so in fact). And I don't think you're dumb or necessarily foolish for not knowing any better, since the vast majority of audiophiles simply don't. You're just uninformed. I do start to wonder though after the information and explanations have been provided and rejected, just how foolish someone is thinking. Perhaps its human nature, and in discussions like this no one likes to be told they are wrong, especially from a bunch of internet nobodies like us who probably don't do it with the tact that we should.
beaveav and Greenwood Ave like this.
Bigus is offline  
post #39 of 227 Old 02-17-2016, 12:11 AM
AVS Forum Special Member
 
kucharsk's Avatar
 
Join Date: Feb 2004
Location: Louisville, CO
Posts: 6,449
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1765 Post(s)
Liked: 1042
Ah, the "you audiophiles are just too uneducated to know everything sounds the same" argument.

Well if that's the case there's no need for this entire thread, as all amps sound the same by default.
W4RLORD likes this.
kucharsk is offline  
post #40 of 227 Old 02-17-2016, 01:36 AM
FMW
AVS Forum Special Member
 
FMW's Avatar
 
Join Date: Jun 2007
Posts: 7,480
Mentioned: 13 Post(s)
Tagged: 1 Thread(s)
Quoted: 1654 Post(s)
Liked: 1710
Quote:
Originally Posted by kucharsk View Post
I hear the audible differences because some equipment excels at musical reproduction more than others.
For sure. That is always the case with speakers and room acoustics. If you are talking about cables or DAC's, for instance, then no you hear audible differences because of hearing bias. Your statement is too broad.

Quote:
I get that many don't hear or don't buy that there can be differences, and I chalk that more up to the testing methodology.
I can't imagine why. Logic would tell you that if you don't hear a difference when bias is removed, then the difference is likely due to bias. To argue the opposite is simply not logical nor is it correct.

Quote:
I'm afraid there may well never be a resolution to that, and that's fine.
These things were resolved long ago. Audiophiles, though, don't accept the resolutions because it appears to conflict with the hobby, attack hearing ability or ego. Some differences are audible and some are not. To say that all is audible simply explains that you haven't experienced scientifically valid comparisons.

Quote:
Many here believe the folks at high-end audio companies go home at night and rub their hands with glee over what over-priced crap they will get people to buy next, and I don't see that at all. I've talked to these people and know many of them personally and they do what they do because they believe in better sound. Perhaps all audiophiles are insane or merely deluded, but ultimately each person decides whether they hear improvements and whether they're worth their hard-earned money.
No, I think high end audio manufacturers simply suffer from the same hearing bias you and I and everyone else experiences. I think they probably mean well and, like you, don't do bias free comparisons. I was an audiophile for many decades and I wasted all kinds of hard earned money on things that didn't make a difference. After a couple of years of bias controlled testing, I simply know better now. So no, I don't think audiophiles are deluded. They are simply unaware of their hearing bias.

Quote:
I now return you to your regular ABX discussion.
Thanks. I sent it back to you.
FMW is offline  
post #41 of 227 Old 02-17-2016, 05:26 AM
AVS Forum Special Member
 
kucharsk's Avatar
 
Join Date: Feb 2004
Location: Louisville, CO
Posts: 6,449
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1765 Post(s)
Liked: 1042
Quote:
Originally Posted by FMW View Post
No, I think high end audio manufacturers simply suffer from the same hearing bias you and I and everyone else experiences. I think they probably mean well and, like you, don't do bias free comparisons. I was an audiophile for many decades and I wasted all kinds of hard earned money on things that didn't make a difference. After a couple of years of bias controlled testing, I simply know better now. So no, I don't think audiophiles are deluded. They are simply unaware of their hearing bias.
Personally I think the biggest bias is the belief that DBT has anything to do with the way we actually experience audio reproduction, but I'll just bow out now.
kucharsk is offline  
post #42 of 227 Old 02-17-2016, 05:55 AM
AVS Forum Special Member
 
Gooddoc's Avatar
 
Join Date: Oct 2004
Posts: 8,729
Mentioned: 182 Post(s)
Tagged: 0 Thread(s)
Quoted: 3736 Post(s)
Liked: 3336
Quote:
Originally Posted by FMW View Post
For sure. That is always the case with speakers and room acoustics. If you are talking about cables or DAC's, for instance, then no you hear audible differences because of hearing bias. Your statement is too broad.



I can't imagine why. Logic would tell you that if you don't hear a difference when bias is removed, then the difference is likely due to bias. To argue the opposite is simply not logical nor is it correct.



These things were resolved long ago. Audiophiles, though, don't accept the resolutions because it appears to conflict with the hobby, attack hearing ability or ego. Some differences are audible and some are not. To say that all is audible simply explains that you haven't experienced scientifically valid comparisons.



No, I think high end audio manufacturers simply suffer from the same hearing bias you and I and everyone else experiences. I think they probably mean well and, like you, don't do bias free comparisons. I was an audiophile for many decades and I wasted all kinds of hard earned money on things that didn't make a difference. After a couple of years of bias controlled testing, I simply know better now. So no, I don't think audiophiles are deluded. They are simply unaware of their hearing bias.



Thanks. I sent it back to you.
Quote:
Originally Posted by kucharsk View Post
Personally I think the biggest bias is the belief that DBT has anything to do with the way we actually experience audio reproduction, but I'll just bow out now.
There's no doubt the DBT is valid for audio. The difficult part is that the study needs to be designed so that evidence of the alternative hypothesis can be statistically validated, otherwise the null hypothesis will simply be confirmed. This does not make the null hypothesis valid, it is simply a preordained result of the study design and lack of power.

Some, if not many, audio studies fall into this category, but the results are paraded around here as though there is definitive proof of a statistical power too great be questioned, and those that do question are mercilessly brow beaten.

Many large studies have reached the wrong conclusion, and the typical audio study is very small compared to those, for example, in medicine.

I'm am 100% about science, but rigid defense of dogma without great evidence to support it, even if it's the only evidence, is not very science like. It's opinion with some evidence to support it, nothing more.
Gooddoc is offline  
post #43 of 227 Old 02-17-2016, 06:43 AM
AVS Forum Special Member
 
Join Date: Jan 2007
Posts: 2,630
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1331 Post(s)
Liked: 816
Quote:
Originally Posted by kucharsk View Post
Personally I think the biggest bias is the belief that DBT has anything to do with the way we actually experience audio reproduction, but I'll just bow out now.
It's a Hit and Run!!!!
beaveav likes this.
jaddie is offline  
post #44 of 227 Old 02-17-2016, 08:46 AM
 
LFEer's Avatar
 
Join Date: Jun 2013
Posts: 1,243
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 818 Post(s)
Liked: 532
Quote:
Originally Posted by kucharsk View Post
Personally I think the biggest bias is the belief that DBT has anything to do with the way we actually experience audio reproduction, but I'll just bow out now.
Why don't you at least try DBT yourself once before you bow out.
LFEer is offline  
post #45 of 227 Old 02-17-2016, 08:48 AM
 
LFEer's Avatar
 
Join Date: Jun 2013
Posts: 1,243
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 818 Post(s)
Liked: 532
Quote:
Originally Posted by jaddie View Post
It's a Hit and Run!!!!
OMG, so much explanation with so few words! You win the efficiency award!!!
LFEer is offline  
post #46 of 227 Old 02-17-2016, 10:45 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5770
I wonder if the OP now realizes the futility of his original intent.
Gooddoc and lovinthehd like this.
Bill Fitzmaurice is offline  
post #47 of 227 Old 02-17-2016, 06:43 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,626
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 607 Post(s)
Liked: 574
Quote:
Originally Posted by kucharsk View Post
Ah, the "you audiophiles are just too uneducated to know everything sounds the same" argument."
Ignoring the "everything" hyperbole, yes, most simply are uninformed. A smaller fraction I would guess are just stubborn, or wilfully ignorant I suppose. The former is fine... there are a great many things in life of which I'm simply ignorant, and some I probably "think" I know about and just don't realize how ignorant I am. Hopefully I won't be stubborn when and if someone tries to educate me, but I certainly can't give a guarantee as that is human nature. I try to keep an open mind as much as possible. Some apparently keep it shut light a bank vault.

Quote:
Well if that's the case there's no need for this entire thread, as all amps sound the same by default.
Not all, but there's only one way to tell which ones.
Bigus is offline  
post #48 of 227 Old 02-17-2016, 06:46 PM
 
lovinthehd's Avatar
 
Join Date: Dec 2007
Location: OROR
Posts: 16,231
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 4588 Post(s)
Liked: 4795
Quote:
Originally Posted by Bigus View Post


Not all, but there's only one way to tell which ones.
I've been waiting for this definitive list for my whole life!
lovinthehd is offline  
post #49 of 227 Old 02-17-2016, 07:05 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,626
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 607 Post(s)
Liked: 574
And BTW, there is nothing inherent about bias controlled testing that would make it different from how you normally experience audio.

Sure, there are some specific protocols like using an ABX box that introduce another device in the chain you can (and did) claim degrades the audio signal like recording a DVD to a VHS tape. But you don't have to use an ABX box to do ABX testing, it just saves time and effort.

And some protocols force you to listen in a way that isn't the same as you normally would. Again using ABX as the comparison, you probably wouldn't alternate back and forth like that normally. Although both logic and experiments suggest that as our audio memory is relatively short and degrades with time, the rapid switching is to your advantage, giving you the greatest possible chance to detect a difference.

But ABX isn't the only way to do controlled testing. Indeed, you can design a test where you compare two components just as you would normally, swapping them in and out of your system at your leisure, listening as long as you want to each with whatever music you choose, except that your bias has been removed. It just takes a hecknof a lot longer to do such a test to produce statistically significant population results. But getting results valid to you is a bit easier. There are many possible ways, but the most obvious is probably to disguise the two products' identities, perhaps by putting each in an identical enclosure. You listen as you wish, you just don't know which is which.

The ONLY thing that is different about that test and the way you normally listen is your knowledge of which component is which. And if you want to argue that changes the way you listen, then go ahead... you're preaching to the choir!

But sometimes the most interesting tests are when the subject doesn't know he is being tested. This is used occasionally in medicine and other fields, when there is concern that knowledge of the test changes your psychological state, senses, etc. in a way that might invalidate the test. eg, you think controlled testing somehow ruins the way you listen to music, thus if you are in a test, ANY test, you're convinced it will make everything sound the same even if you previously thought otherwise, and you'll just then say everything sounds the same. You can address that by introducing comparisons in the mix that genuinely should be reliably detected and identified to gauge the honesty of the test subject, or just don't let them know they're being tested, or lie about the true nature of the test.

That has been done several times for audio but not rigourously so and not with statistical power. But you can guess at the results anyway.
Bigus is offline  
post #50 of 227 Old 02-17-2016, 11:22 PM
AVS Forum Special Member
 
A9X-308's Avatar
 
Join Date: Mar 2008
Location: Australia
Posts: 8,321
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 1792 Post(s)
Liked: 1484
Quote:
Originally Posted by Bigus View Post
Sure, there are some specific protocols like using an ABX box that introduce another device in the chain you can (and did) claim degrades the audio signal like recording a DVD to a VHS tape.
He can claim it, but won't be able to back it up in any way. It's a hyperbolic claim with no basis in fact.
A9X-308 is offline  
post #51 of 227 Old 02-17-2016, 11:40 PM
Member
 
Join Date: Oct 2014
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 58 Post(s)
Liked: 16
I do not understand the reasoning for the creation of this thread, amplifiers are already tested in things like THD+N and channel seperation etc.

Video Pioneer PDP-LX5090 (calibrated)
Audio Benchmark AHB2, Chord Hugo 2, KEF LS50 (Frosted Black), Custom Design FS104
Cables QED Reference HDMI, QED Reference Optical Quartz, QED Reference USB A-B, QED Supremus, QED Signature Audio 40
Sources PlayStation 3, PlayStation 4, MacBook Pro, AirPort Express
W4RLORD is offline  
post #52 of 227 Old 02-18-2016, 12:27 AM
AVS Forum Special Member
 
Join Date: Jan 2007
Posts: 2,630
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1331 Post(s)
Liked: 816
Quote:
Originally Posted by Bigus View Post
And BTW, there is nothing inherent about bias controlled testing that would make it different from how you normally experience audio.

Sure, there are some specific protocols like using an ABX box that introduce another device in the chain you can (and did) claim degrades the audio signal like recording a DVD to a VHS tape. But you don't have to use an ABX box to do ABX testing, it just saves time and effort.

And some protocols force you to listen in a way that isn't the same as you normally would. Again using ABX as the comparison, you probably wouldn't alternate back and forth like that normally. Although both logic and experiments suggest that as our audio memory is relatively short and degrades with time, the rapid switching is to your advantage, giving you the greatest possible chance to detect a difference.
It's called "echoic memory", it's very short, especially for subtle qualitative impressions, longer for information content. You pretty much have to switch instantaneously for the highest resolution ABX.
Quote:
Originally Posted by Bigus View Post
But ABX isn't the only way to do controlled testing. Indeed, you can design a test where you compare two components just as you would normally, swapping them in and out of your system at your leisure, listening as long as you want to each with whatever music you choose, except that your bias has been removed.
You can't' remove bias in that sort of test. It's neither single, nor double blind. And the transition is far to slow to be useful. It becomes a highly biased impression.
Quote:
Originally Posted by Bigus View Post
It just takes a hecknof a lot longer to do such a test to produce statistically significant population results. But getting results valid to you is a bit easier. There are many possible ways, but the most obvious is probably to disguise the two products' identities, perhaps by putting each in an identical enclosure. You listen as you wish, you just don't know which is which.
You can't actually maintain the blind criteria, even if you put things in identical cabinets. You know which you're listening to, even if you don't know what you're listening to. There is a basic left vs right bias, top vs bottom, etc. So you can't even do controlled single blind. No amount of statistics will remedy the fact that the data is biased. Even a bias relating to how light hits one unit vs the other.

In a double-blind ABX, you don't know what A or B is, but even the A vs B bias is removed because the task is to match A or B to X.
Quote:
Originally Posted by Bigus View Post

The ONLY thing that is different about that test and the way you normally listen is your knowledge of which component is which. And if you want to argue that changes the way you listen, then go ahead... you're preaching to the choir!
That is a very important difference. The data collected that way is definitely biased, though you may not know exactly how. Bias does, however, invalidate the results.
Quote:
Originally Posted by Bigus View Post
But sometimes the most interesting tests are when the subject doesn't know he is being tested. This is used occasionally in medicine and other fields, when there is concern that knowledge of the test changes your psychological state, senses, etc. in a way that might invalidate the test. eg, you think controlled testing somehow ruins the way you listen to music, thus if you are in a test, ANY test, you're convinced it will make everything sound the same even if you previously thought otherwise, and you'll just then say everything sounds the same.
Using an uninformed test subject can be done in medicine, but usually isn't because there are too many uncontrolled factors. It's why medications and treatments are compared against placebos.
Quote:
Originally Posted by Bigus View Post
You can address that by introducing comparisons in the mix that genuinely should be reliably detected and identified to gauge the honesty of the test subject, or just don't let them know they're being tested, or lie about the true nature of the test.
It's actually done the other way. You present a deliberate null condition, where A and B are identical. It's used to confirm there is no bias. Results should show no reliable match to X. You can't really introduce a detectable change because there are no solid metrics for consistent audibility of things like distortion.
Quote:
Originally Posted by Bigus View Post
That has been done several times for audio but not rigourously so and not with statistical power. But you can guess at the results anyway.
Introducing an intentionally audible change is only useful if you're trying to quantify the specific audibility of that change. Otherwise, it's not going to prove much.
jaddie is offline  
post #53 of 227 Old 02-18-2016, 08:01 AM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,626
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 607 Post(s)
Liked: 574
Quote:
Originally Posted by jaddie View Post
You can't' remove bias in that sort of test.
Sure you can, just takes work. If you really are concerned about the listener being biased by which enclosure he "believes/guesses" holds which component, or by scratches on one enclosure vs the other etc., simply have someone intermittently swap internals of the two enclosures according to a random schedule. The listener is completely blinded to which component he is listening to but retains the ability to swap them at his leisure. It's single blinded without a doubt, and so long as the person overseeing the test isn't and has no contact with the person swapping the internals of the enclosures, it's double blinded as well. Although unlikely, it's possible that some differences in enclosures (scratch or whatever) might lead to a repeatable favoring of that enclosure regardless of internals, but the data would show that and it is clearly distinct from an audible differences of internal components. In the worst case, you could invalidate the test claiming that the enclosure preference completely masked any possible component preference, but I think you're really stretching there. After all, it isn't difficult to make two enclosures visible identical, weigh the same after components added, and otherwise impossible to tell apart by any reasonable means the listener has available. If he's doing the swapping, starting wtih two boxes presented to him (internally swapped or not) at the beginning of each listening session for him to connect in whatever order he wishes, it's very hard to see how any left right, up down, etc. bias could possibly arise. And to be frank, so long as the enclosures are physically indistinguishable to him, and he starts each leisure listening session from a default position (both boxes sitting on a table for him to connect in his system), there's no reason to swap internals... only randomly vary which box sits where on the table at the beginning of each session. It takes work, but it's far from impossible.

Quote:
And the transition is far to slow to be useful.
Depends on what you mean by useful. Since I agree with you about the length of auditory memory, I too believe this sort of test would seriously reduce the likelihood of detecting audible differences. And even if it could, it would take so long to build statistical significance as to become impractical.

But for demonstrating to a single individual the power of perceptual bias, and how differences that were "obvious" disappear when bias is removed, it would be highly useful.

Quote:
Using an uninformed test subject can be done in medicine, but usually isn't because there are too many uncontrolled factors.
I think my use of "occasionally" covered that.

Quote:
It's actually done the other way. You present a deliberate null condition, where A and B are identical. It's used to confirm there is no bias. Results should show no reliable match to X. You can't really introduce a detectable change because there are no solid metrics for consistent audibility of things like distortion.
You can do both, and each has its own purpose. And of course you can introduce detectable differences just as you allude to later...

Quote:
Introducing an intentionally audible change is only useful if you're trying to quantify the specific audibility of that change. Otherwise, it's not going to prove much.
Sure it does. It proves that the test subject isn't biased for the objectivist position, convinced that there are no detectable differences in controlled conditions, thus randomly guessing A or B even when audible differences really do exist. Or, that his hearing is terrible! Any data for a subject where results for such intentionally introduced samples approaches random guessing while being clearly non-random for the rest of the population should be excluded based on either of the above possible reasons.


Last edited by Bigus; 02-18-2016 at 08:10 AM.
Bigus is offline  
post #54 of 227 Old 02-18-2016, 08:37 AM
 
lovinthehd's Avatar
 
Join Date: Dec 2007
Location: OROR
Posts: 16,231
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 4588 Post(s)
Liked: 4795
Quote:
Originally Posted by W4RLORD View Post
I do not understand the reasoning for the creation of this thread, amplifiers are already tested in things like THD+N and channel seperation etc.
That isn't the type of testing/comparison being discussed. Did you read the OP?
lovinthehd is offline  
post #55 of 227 Old 02-18-2016, 09:36 AM
Member
 
Join Date: Oct 2014
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 58 Post(s)
Liked: 16
Quote:
Originally Posted by lovinthehd View Post
That isn't the type of testing/comparison being discussed. Did you read the OP?
Maybe it's different for stereo amplifiers, but I sure as hell can tell the difference between them.

Video Pioneer PDP-LX5090 (calibrated)
Audio Benchmark AHB2, Chord Hugo 2, KEF LS50 (Frosted Black), Custom Design FS104
Cables QED Reference HDMI, QED Reference Optical Quartz, QED Reference USB A-B, QED Supremus, QED Signature Audio 40
Sources PlayStation 3, PlayStation 4, MacBook Pro, AirPort Express
W4RLORD is offline  
post #56 of 227 Old 02-18-2016, 09:38 AM
 
lovinthehd's Avatar
 
Join Date: Dec 2007
Location: OROR
Posts: 16,231
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 4588 Post(s)
Liked: 4795
Quote:
Originally Posted by W4RLORD View Post
Maybe it's different for stereo amplifiers, but I sure as hell can tell the difference between them.
Sure you can, so would you like to participate in such a test to prove that ability to others for the sake of the amplifier listening culture?
lovinthehd is offline  
post #57 of 227 Old 02-18-2016, 10:05 AM
AVS Forum Special Member
 
Join Date: Jan 2007
Posts: 2,630
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1331 Post(s)
Liked: 816
Quote:
Originally Posted by Bigus View Post

Sure it does. It proves that the test subject isn't biased for the objectivist position, convinced that there are no detectable differences in controlled conditions, thus randomly guessing A or B even when audible differences really do exist. Or, that his hearing is terrible! Any data for a subject where results for such intentionally introduced samples approaches random guessing while being clearly non-random for the rest of the population should be excluded based on either of the above possible reasons.
The problem with that is selecting what the "audible difference" is. The audibility of any form of distortion (including FR changes) is a multi-dimensional issue. You can't just say, "I'll inject 3% THD" because the resulting harmonic distribution radically affects audibility, as does the nonlinearity that causes the distortion. You can't just say, "I'll modify FR by XdB" because the audibility of a response change varies with degree of change, bandwidth of the change, spectra of the stimulus, and specific area of the total response modified. And so one. Every distortion you add cannot be simply specified. You end up testing the audibility of a specific distortion, and that's about it. The proof of non-bias must be a null result, identical A and B, and 50% X results.
jaddie is offline  
post #58 of 227 Old 02-18-2016, 11:23 AM
Member
 
Join Date: Oct 2014
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 58 Post(s)
Liked: 16
Quote:
Originally Posted by lovinthehd View Post
Sure you can, so would you like to participate in such a test to prove that ability to others for the sake of the amplifier listening culture?
I have listened to many stereo amplifiers, with many different speakers and sources so no need to test. The biggest difference that is easily spotted imo is how well they render bass. Many amplifiers have a roll-off and IMD is much more noticable when playing bass notes. Some amplifiers internal power supply can also affect low-end performance too. The worse the amp, the more 'boomy' the bass (though some people like this, as per tube amps), cleaner, less-distorted amps render bass very tightly and sound much more natural.

Video Pioneer PDP-LX5090 (calibrated)
Audio Benchmark AHB2, Chord Hugo 2, KEF LS50 (Frosted Black), Custom Design FS104
Cables QED Reference HDMI, QED Reference Optical Quartz, QED Reference USB A-B, QED Supremus, QED Signature Audio 40
Sources PlayStation 3, PlayStation 4, MacBook Pro, AirPort Express
W4RLORD is offline  
post #59 of 227 Old 02-18-2016, 11:35 AM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,626
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 607 Post(s)
Liked: 574
Quote:
Originally Posted by jaddie View Post
The problem with that is selecting what the "audible difference" is....

You end up testing the audibility of a specific distortion, and that's about it.
You're missing the forest for the trees I think. If you add samples to the test that have varying amounts of distortion, say THD average over the musical selection of 1%, 5%, 10% in addition to the actual devices under test, then yes, you test the audibility of that specific distortion along with testing the audibility of differences of the two devices under test. You could do the same with varying FR aberrations. But that serves an important function as I explained, but will do so again.

We would expect that such easily measured differences would at some threshold become audible for most listeners. We don't need to know what the threshold is beforehand, only have some reasonable expectation so that we tailor the samples appropriately. In a population of listeners, hearing acuity will vary. Perhaps we get 100% that consistently differentiate the 10% distortion sample, 80% for the 5% THD, and only 10% seem to be able to distinguish 1% THD (and, of course, 0% can tell 0.05 and 0.01% apart ). Then all is well. But now imagine that within our population, not 100% can distinguish the 10% distortion, maybe its just 99%. And that 1% have results equivalent to random guessing on all samples, including all distortion levels and the two devices under test. For whatever the reason (either hearing impaired, or bias against the possibility of audible differences), these listeners should be excluded from the data.

You could of course introduce numerous variations and essentially test for audibility of a specific type of distortion, but that is a somewhat different intent and test design. We don't need thorough coverage of the possible distortions, just a few crude levels to establish some reference level of listener hearing or cooperation. These altered samples are introduced along with our two devices under test, and unknowingly to the participants, but are not the primary focus of the test.

This protocol IMO should be followed in all DBT tests of this nature, ABX or otherwise, and I've suggested as much in the past but unfortunately don't believe this is something that has been done.

Quote:
The proof of non-bias must be a null result, identical A and B, and 50% X results.
That only tests for one form of bias, it doesn't prove other bias (specifically the one I've addressed above) doesn't exist. I could take such a test, randomly guess A or B on everything without bothering to listen at all, pass your null test, and still be 100% biased and my results absolutely invalid.

Bigus is offline  
post #60 of 227 Old 02-18-2016, 11:50 AM
Member
 
Join Date: Oct 2014
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 58 Post(s)
Liked: 16
I can understand the need for a A/B tests on equpment that is similar in specs to one another, to test if there's any audible difference between the two (so you can therefore go with the cheapest option and have relatively the same sound quality, aka result). With higher-end equpment though it's absolutely a mute point. If you cannot tell the differnce between an amp like a Denon PMA520AE and a Sugden A21al Series 2 for example then you must be hearing impaired (or not wanting to listen for differences).

Video Pioneer PDP-LX5090 (calibrated)
Audio Benchmark AHB2, Chord Hugo 2, KEF LS50 (Frosted Black), Custom Design FS104
Cables QED Reference HDMI, QED Reference Optical Quartz, QED Reference USB A-B, QED Supremus, QED Signature Audio 40
Sources PlayStation 3, PlayStation 4, MacBook Pro, AirPort Express

Last edited by W4RLORD; 02-18-2016 at 01:03 PM.
W4RLORD is offline  
Sponsored Links
Advertisement
 
Closed Thread Audio Theory, Setup, and Chat

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off