Why haven't subjectivists and objectivists met to do a live ABX test?

[Geert]

I might be wrong but Harman’s aim is to identify user preferences.

And they try to assess why people prefer something so their engineers understand what to focus on. These tests are documented in Toole's and Olive's papers.

All we’re trying to do here is establish whether it’s possible to identify differences in well measuring amps.

But that was not the topic of discussion you responded to yourself.

 

[rsoffer]

Absolutely 100%! I can't help feeling that most of the subjectivist moaners simply don't understand how science works. Scientists want to find ways to break current theories, because that means we get to create new ones. If you're a physicist, then discovering a phenomenon that truly broke General Relativity would be the highlight of your career, and physicists have dedicated their entire lives to finding such a thing.
But science demands a very high standard of truth, one which is seemingly far beyond the reach of the average audio pundit.

100%.

As it stands now, unless you’ve done the testing yourself you’re bound to two existences:

1.) You choose to believe all well measuring amps sound the same. You stick with your THX amp, while audiophiles bombard you with videos about how terrible it sounds and why the new discrete amp on the block sounds way better. Maybe they are right, you’re missing out on all of this interesting and amazing sounding gear.

2.) You choose to believe there are subtle audible differences between all amps. You spend hours listening and reading reviews of gear, you buy that new amp. Now another amp comes along, you’re intrigued. Time to upgrade. There’s still that voice in the back of your head that says, what if this is all a giant waste of time?

 

[Geert]

I can't help feeling that most of the subjectivist moaners simply don't understand how science works.

Subjectivist want science to follow the circle below counter clockwise. They already have a conclusion, science only needs to come up with an explanation.

 

[ahofer]

You spend hours listening and reading reviews of gear, you buy that new amp. Now another amp comes along, you’re intrigued. Time to upgrade.

You expect to hear a difference; you do; you take it home and in a few weeks your biases aren't pushing the same way; you become dissatisfied; maybe there is something better.

Bias is non-stationary.

 

[Blumlein 88]

Having just listened to 10 seconds of music it seems like it may be long enough to decide which is preferable, but probably won't allow for a detailed exposition on the differences. Are these tests usually just 'which do you prefer' or do they expect finer detail?


There's a first time for everything!

Before you can have a preference you must detect a difference. Short tests of 10 seconds or less are the most discerning method of determining if you hear a difference. If you cannot hear a difference then obviously there is no basis for a preference. That does not prevent one from using longer times for preference ratings, but those regularly fail. All we need if for someone to detect differences via preference reliably. Many contend longer is better, but it hasn't worked out that way when tested. Going from memory which may be faulty, the test I mentioned earlier which lasted for a month involved 3% distortion. Preferences were not reliable. Within minutes, using short term testing, recognition was near 100%.

 

[Blumlein 88]

I believe the tests are usually to identify X.

That being said, I feel the simplest most fair test would be to know in advance which amp you are identifying, and keep that constant the entire test. For example, amp A is the THX amp. Amp B is the Asgard. Identify which is the THX amp every time on random switching of the cables.

Why do you need to know to detect a difference? Even so, yes blind testing can be done that way. You know what A is and you know what B is. Given X do you know which is which? When I do a Foobar ABX on something I know what the two alternatives are.

A duo-trio test you know what the reference is. You then listen to A or B. One is the reference and one isn't. You pick which it is with the ability to compare to reference. Pretty close to what you are describing. Another version is a triangle test. You have three versions. Two are identical one is different. Choose the odd man out. These methods are also employed for preference testing. The 8 generation digital to analog copy files I posted as a triangle test.

 

[rsoffer]

Why do you need to know to detect a difference? Even so, yes blind testing can be done that way. You know what A is and you know what B is. Given X do you know which is which? When I do a Foobar ABX on something I know what the two alternatives are.

A duo-trio test you know what the reference is. You then listen to A or B. One is the reference and one isn't. You pick which it is with the ability to compare to reference. Pretty close to what you are describing. Another version is a triangle test. You have three versions. Two are identical one is different. Choose the odd man out. These methods are also employed for preference testing. The 8 generation digital to analog copy files I posted as a triangle test.

I feel as though changing X throughout the test makes it more difficult and confusing. Knowing what X will be in advance and keeping that static through the test means you can spend a few minutes before the test calibrating your ears to know what X sounds like. Then all you need to do is consistently tell the difference and identify which is which when A and B are randomized.

 

[Blumlein 88]

I feel as though changing X throughout the test makes it more difficult and confusing. Knowing what X will be in advance and keeping that static through the test means you can spend a few minutes before the test calibrating your ears to know what X sounds like. Then all you need to do is consistently tell the difference and identify which is which when A and B are randomized.

I am not sure I understand. If you know what X is, and swap A and B, you really are only renaming what is being swapped. For instance in a duo-trio test you know what the reference is at all times. You even know what the alternative is. You then get an unknown and match it as either reference or not. And can go back to reference for comparison any time you wish. You can make whichever you want the reference. So how does this not meet the manner in which you wish to do the test?

 

[Overseas]

The idea of the 2 camps doing same test is great.
The problem is, what stops the obj. camp to pretend they hear no difference whatever is played?!

 

[Blumlein 88]

The idea of the 2 camps doing same test is great.
The problem is, what stops the obj. camp to pretend they hear no difference whatever is played?!

 

[Overseas]

You probably think you know which camp I am, based on my post. Well, you cannot guess. Unless you really believe any non-technical person is a 'subjectivist'.

 

[Overseas]

Subjectivist want science to follow the circle below counter clockwise. They already have a conclusion, science only needs to come up with an explanation.
View attachment 250468

Looks like YOU already have a CONCLUSION about the other camp, whatever that is. I really cannot follow you because all I want is: 1) best sound in my budget (wtf is 'best', please enlighten me objectively) and 2) to hear at least ONCE the MOST PERFECT sound on PLANET just to have a reference.
So, which camp am I?!

 

[rsoffer]

I am not sure I understand. If you know what X is, and swap A and B, you really are only renaming what is being swapped. For instance in a duo-trio test you know what the reference is at all times. You even know what the alternative is. You then get an unknown and match it as either reference or not. And can go back to reference for comparison any time you wish. You can make whichever you want the reference. So how does this not meet the manner in which you wish to do the test?

Yeah, I guess you're right. It's basically the same test, just more intuitive naming IMO but shouldn't matter.

 

[rsoffer]

The idea of the 2 camps doing same test is great.
The problem is, what stops the obj. camp to pretend they hear no difference whatever is played?!

Wasn't suggesting both camps do the test. Only those who believe they hear differences will take the test, which will be administered and monitored by the "other camp" or an unbiased third party.

 

[Overseas]

Yeap, so the 'subjectivists' are presumed guilty. It is a good start

 

[rsoffer]

Yeap, so the 'subjectivists' are presumed guilty. It is a good start

That's how science works, you assume the "null" case.

 

[Overseas]

Science is an absolute ideal.
'Objectivists' are prone to human error.
It is a good, safe thing that 'objectivists' are not threatened by any test. No claim, no error. Their only claim is the other side is in error. Seems to me we are on the philosophical and not scientific field.
Otherwise, I totally agree to blind tests.

 

[Geert]

Looks like YOU already have a CONCLUSION about the other camp, whatever that is.

People claiming to hear all kinds of differences without any proof is not a conclusion, it's an observation. There are thousands of posts proving the claims.

all I want is: 1) best sound in my budget (wtf is 'best', please enlighten me objectively) and 2) to hear at least ONCE the MOST PERFECT sound on PLANET just to have a reference.
So, which camp am I?!

You didn't make any statement about how you value the scientific approach to audio so how would I know? Your two objectives are off topic so I won't respond to that.

 

[rsoffer]

Science is an absolute ideal.
'Objectivists' are prone to human error.
It is a good, safe thing that 'objectivists' are not threatened by any test. No claim, no error. Their only claim is the other side is in error. Seems to me we are on the philosophical and not scientific field.
Otherwise, I totally agree to blind tests.

I'm not sure what you want. One camp claims they can hear differences, the other believes they cannot. What is the point of testing the camp which cannot hear the differences?

The point is to prove its possible to hear the differences. It only requires one person doing it reliably and consistently. What would you propose?

 

[Geert]

'Objectivists' are prone to human error.

But far less than people only believing what they hear and refusing to accept scientific research and methodologies.

It is a good, safe thing that 'objectivists' are not threatened by any test. No claim, no error.

The 'claims' of objectivist can be found in thousands of scientific studies and in applied science areas like math and engineering. No where else is being wrong so obvious.