DAC audible differences are finally proven with ABX?

[PristineSound]

How could we prove dishonesty here?

Barring a confession, the evidence could only ever be circumstantial. So it would have to be one of those 'beyond a reasonable doubt' cases.

You are correct in that (1) ASR is not a court of law, who the hell do we think we are going around adjudicating cases, it's almost laughable (2) there is no way we can prove anyone is being dishonest, we can prove that the test results are simply not repeatable.

However, if there is nothing wrong with the ABX device and the test results are not repeating, there is no one to stop anyone else to conclude dishonesty on Josh's part. If there is nothing wrong with the ABX device and the test results cannot be repeated, then I for one, will conclude dishonesty.

 

[Plcamp]

However, if there is nothing wrong with the ABX device and the test results are not repeating, there is no one to stop anyone else to conclude dishonesty on Josh's part.

If you ever watched the British series “Fake or Fortune” you regularly see that artwork cannot be definitively - no doubt at all - certainly by a particular artist. The art world solves that by having an appointed expert or organization of experts make the call (for any given artist), “genuine” or “not”, considering all available evidence. It’s the only method to arbitrate.

ASR defaults to being that role for such claims as Josh. Because no other can claim the objectivity and expertise that is brought to bear right here.

And rejection of a claim does not automagically condemn the claimant as ‘dishonest’.

ASR should not shy away from that authority. It should be asserted loudly and broadly and for the benefit of everyone.

If that’s deemed arrogant, then so be it.

 

[fpitas]

A lot of experiments in science are not reproducible. Sometimes it's dishonesty, or seeking publicity etc. But sometimes it's just experimental error.

 

[pablolie]

It doesn't have to be outright, willful dishonesty. It can simply be a lack of ability in setting up an environment that truly doesn't give away other clues. But then, it's dishonest to present the experiment as such.

 

[AndyBell]

And rejection of a claim does not automagically condemn the claimant as ‘dishonest’.

ASR should not shy away from that authority. It should be asserted loudly and broadly and for the benefit of everyone.

Only Josh knows if he’s being honest.

We can begin to doubt it if the box is transparent and he cannot repeat the result in controlled conditions, assuming he’s willing to repeat the test.

But bias is so powerful that it is theoretically possible he conned himself into thinking the test was objective when it was inadvertently designed to provide the ‘desired’ result. In that case he’s being honest but not objective.

I doubt he’ll agree to repeat the test under stringent controls and I also doubt anyone will be able to replicate it.

Where does that leave us? With a bunch of possibilities and probabilities but no definitive conclusion about his test.

Somehow I think obtaining the box from him was a mistake - maybe we should have taken the position that unless he repeats it under strict, supervised, conditions then we’d file it under ‘anecdotal’ and move on to other things…

 

[krabapple]

You are correct in that (1) ASR is not a court of law, who the hell do we think we are going around adjudicating cases, it's almost laughable (2) there is no way we can prove anyone is being dishonest, we can prove that the test results are simply not repeatable.

Oh, you'll never prove that, sorry. Not to true believers in oreilles d'or.*
They will just say: "None of youse guys could repeat it. Maybe someone else can."
They aren't strictly wrong, they simply don't give a fig about the probabllity factor.


*know what? I could not find an image of 'golden ears' on Google Images. I refuse to ask AI to make one.

 

[krabapple]

If you ever watched the British series “Fake or Fortune” you regularly see that artwork cannot be definitively - no doubt at all - certainly by a particular artist. The art world solves that by having an appointed expert or organization of experts make the call (for any given artist), “genuine” or “not”, considering all available evidence. It’s the only method to arbitrate.

ASR defaults to being that role for such claims as Josh. Because no other can claim the objectivity and expertise that is brought to bear right here.

And rejection of a claim does not automagically condemn the claimant as ‘dishonest’.

ASR should not shy away from that authority. It should be asserted loudly and broadly and for the benefit of everyone.

If that’s deemed arrogant, then so be it.

I hope we can render a verdict of not guilcup

 

[krabapple]

Somehow I think obtaining the box from him was a mistake - maybe we should have taken the position that unless he repeats it under strict, supervised, conditions then we’d file it under ‘anecdotal’ and move on to other things…

I believe some pest here on this thread already called for 'proctored tests'

 

[danadam]

*know what? I could not find an image of 'golden ears' on Google Images.

I searched "golden ear meme" and "audiophile golden ear meme" and found these:
, ,

 

[krabapple]

Nice work, soldier!

 

[rjp]

If there is nothing wrong with the ABX device and the test results cannot be repeated, then I for one, will conclude dishonesty.

That’s so sad
Is there no possibility he made a mistake?

 

[pablolie]

That’s so sad
Is there no possibility he made a mistake?

Quite a few apologists here. He made $3k out of the "mistake", after previously stating he could not possibly part from the precious device for even a week to get it tested... come on...

 

[Chr1]

Perhaps it was all an elaborate, cynical money making ruse.

He could well be on a beach somewhere, cocktail in hand, laughing his ass off at all the chaos he's caused here.

Keyser "Josh" Soze.
Covert audio grifter extraordinaire.

..

 

[krabapple]

Of course it could be a mistake. That would be one possible conclusion if the ABX box turns out to have a 'tell'.

 

[PristineSound]

That’s so sad
Is there no possibility he made a mistake?

If you read his test method, it's a tall order for a mistake to occurred.

While anything is possible, it does not mean they are probable. At the end of the day, even the most diligent judicial system, the jury has to make verdict based available facts.

 

[pma]

@amirm , are there any news about the abx box? Has it arrived in your site?

 

[danadam]

Maybe someone finds that useful: https://danadam.codeberg.page/binomial_calculator/

 

[Curvature]

Reminds me of the research into ultrasonic audibility using dubious methods.

 

[krabapple]

Maybe someone finds that useful: https://danadam.codeberg.page/binomial_calculator/

View attachment 555564

This table of p-values for a given number of successes in a given number of binary trials (up to 100 trials) says 10/10 right has a p=0.001 As does google's AI. Ditto a coin flip calculator for 10 heads or 10 tails in 10 coin flips. So they must use a different calculator/assumptions that that one.

 

[olieb]

This table of p-values for a given number of successes in a given number of binary trials (up to 100 trials) says 10/10 right has a p=0.001 As does google's AI. Ditto a coin flip calculator for 10 heads or 10 tails in 10 coin flips. So they must use a different calculator/assumptions that that one.

Both results are correct.
p=0.001 = 0.1% ≈ 0.0977% ≈1/1024