Why 24 bit audio and anything over 48k is not only worthless, but bad for music.

[colonelkernel8]

I surmise the participants have better trained ears than me if they can pull off meaningful differences blind wheas I would lose more than 1/2 of my cognitive ability to tell them apart if I participate blind

You lose half your cognitive ability every time you close your eyes? Wild.

 

[71 dB]

“Audio purists and industry should welcome these findings,” said Reiss. “Our study finds high-resolution audio has a small but important advantage in its quality of reproduction over standard audio content. Trained listeners could distinguish between the two formats around sixty percent of the time.”

So they were able to distinguish a difference a tiny bit better than chance would predict, how is that "important" ?

How indeed! To me it would be important if listeners could distinguish between the two formats 99-100 % of the time. I'd say 99 % is enough, because after all we are human beings and prone to making mistakes from time to time... ...however 90 % success rate is already too low for the differences to be "important." Such differences are miniscule, challenging to hear and therefore almost irrelevant.

 

[eq1849]

Haven't those papers been discredited here a million times already?

If I recall correctly, someone claimed they discredited the research by equivocating micro noise bursts with 400 seconds of Bach.

Did I say that, no I didn't.

I said that you pick and choose whatever supports your feelings and are happy to flip flop between science, blind testing and whatever else suits you, cherry picking the bits that support your beliefs and shunning the rest.

Psychology is a science and it explains a great deal of things in audio perception but you can't accept it because to do so shoots down your obsession with gear and the notion that everything makes an audible difference. Instead you invent an alternate reality that aligns with your perception and feelings, like sight is needed to hear properly for example.

You are not interested in learning anything, only seeking out piecemeal data points that you can mentally manipulate to give you some level of confirmation for your existing beliefs and feeling.

when you take the blind listening tests, how long do you listen?

Also, do you wait awhile between tests, or take another immediately?

 

[colonelkernel8]

If I recall correctly, someone claimed they discredited the research by equivocating micro noise bursts with 400 seconds of Bach.

I think there are far more legitimate claims than that.

 

[eq1849]

I think there are far more legitimate claims than that.

By all means let’s hear your take..

 

[colonelkernel8]

By all means let’s hear your take..

Just to be absolutely clear you're referring to the "HiRes" paper from 2017 right? Not any of Milind N. Kunchur's papers?

 

[BS5711]

I agree 100% on that. I thought it was pretty obvious to everyone hence why I didn’t reply on that statement. The way we perceive things is of course has to align with our subjective beliefs not just audio but everything!

What about the “deeply contaminated” part, that was the bit that again you ignored.

If your perceptions are indeed deeply contaminated you are building a belief system around data points that are unreliable at best and entirely imaginary at worst.

“ All that we perceive might be deeply contaminated by our subjective beliefs on the physical world.”

 

[danadam]

There's a reason why volume should be matched to 0.1 dB or better

 

[theveterans]

What about the “deeply contaminated” part, that was the bit that again you ignored.

If your perceptions are indeed deeply contaminated you are building a belief system around data points that are unreliable at best and entirely imaginary at worst.

“ All that we perceive might be deeply contaminated by our subjective beliefs on the physical world.”

I didn't ignore the Deeply contaminated part. I've done blind test before and even then, it's still contaminated with random results (i.e. I can get 18/20 correct responses on a given day then goes to 9/20 correct the following day). To me, blind test is at best moot but to somebody, because of random results, conclusion is inaudible

There's a reason why volume should be matched to 0.1 dB or better

This would've been better represented by graphs. I can't picture what's happening

 

[BS5711]

I didn't ignore the Deeply contaminated part. I've done blind test before and even then, it's still contaminated with random results (i.e. I can get 18/20 correct responses on a given day then goes to 9/20 correct the following day). To me, blind test is at best moot but to somebody, because of random results, conclusion is inaudible

I guess that comes down to how we view the value of the results.

If, overall, across several series of the same test, you cannot demonstrably show a clear indication one way or the other the results indicate that, overall, there is nothing that you can reliably differentiate.

If you repeated the same test across multiple sessions and overall indicated say 80 out of 100 that would be pretty solid evidence, if across that same 100 repetitions you got 50 that is clearly rather less compelling even if on one day you got 18/20.

Again, I believe you take from the tests what suits your feelings rather than trust what the the test demonstrates and use that as a data point, good, bad or inconclusive.

If you did in fact have say 80 out of 100 then we are wasting time with this conversation because I would agree with you there is something that you are differentiating.

 

[theveterans]

I guess that comes down to how we view the value of the results.

If, overall, across several series of the same test, you cannot demonstrably show a clear indication one way or the other the results indicate that, overall, there is nothing that you can reliably differentiate.

If you repeated the same test across multiple sessions and overall indicated say 80 out of 100 that would be pretty solid evidence, if across that same 100 repetitions you got 50 that is clearly rather less compelling even if on one day you got 18/20.

Again, I believe you take from the tests what suits your feelings rather than trust what the the test demonstrates and use that as a data point, good, bad or inconclusive.

If you did in fact have say 80 out of 100 then we are wasting time with this conversation because I would agree with you there is something that you are differentiating.

Found my old test back then. It was the Goldensound DAC filter back then, but it was 15/20 was the best outcome I got, but most of the time, it's just random:

foo_abx 2.2 report
foobar2000 v2.6.4
0124-05-09 21:32:38

File A: Test A (High Performance Filter).wav
SHA1: d626785e576b21b988a3ff3c59f85d3de27ed86d
File B: Test B (Normal Filter).wav
SHA1: 6cefd9bc846b7ba69d2bb06a869596cb740a4c0e

Output:
Core Audio : Yggdrasil+ [exclusive], 16-bit
Crossfading: NO

21:32:38 : Test started.
21:37:33 : Test restarted.
21:37:33 : 00/01
21:38:58 : Test restarted.
21:38:58 : 00/02
21:40:03 : Test restarted.
21:40:03 : 00/03
21:41:07 : Test restarted.
21:41:07 : 01/04
21:42:17 : Test restarted.
21:42:17 : 02/05
21:43:44 : Test restarted.
21:43:44 : 03/06
21:45:42 : Test restarted.
21:45:42 : 04/07
21:47:57 : Test restarted.
21:47:57 : 05/08
21:49:03 : Test restarted.
21:49:03 : 06/09
21:50:02 : Test restarted.
21:50:02 : 07/10
21:53:35 : Test restarted.
21:53:35 : 08/11
21:54:56 : Test restarted.
21:54:56 : 08/12
21:56:37 : Test restarted.
21:56:37 : 09/13
21:59:44 : Test restarted.
21:59:44 : 10/14
22:00:46 : Test restarted.
22:00:46 : 11/15
22:02:38 : Test restarted.
22:02:38 : 12/16
22:03:44 : Test restarted.
22:03:44 : 13/17
22:05:17 : Test restarted.
22:05:17 : 14/18
22:06:31 : Test restarted.
22:06:31 : 14/19
22:08:16 : Test restarted.
22:08:16 : 15/20
22:08:16 : Test finished.

 ----------
Total: 15/20
p-value: 0.0207 (2.07%)

 -- signature --
c5e68ef4cc31b72cdd0fb36b601026280d83ec28

 

[BS5711]

Found my old test back then. It was the Goldensound DAC filter back then, but it was 15/20 was the best outcome I got, but most of the time, it's just random:

If most of the time it is just random then doesn't that demonstrate that you cannot actually differentiate them ?

To me it clearly does and I would put some value on that outcome not look for a reason why the outcome might be flawed, then just get on and enjoy nice gear and nice music regardless.

 

[theveterans]

If most of the time it is just random then doesn't that demonstrate that you cannot actually differentiate them ?

It’s incredibly more subtle than any A/Bing I’ve ever encountered. I usually find myself becoming unable to reliably tell as the test progresses. The switching of DAC filters from internal DAC to a software upsampling like HQPlayer is more readily audible than Goldensound’s filter samples on his website

I’ll do the test later tonight after work to see how I fare this time. Will report back the foobar comparator results

 

[bigshot]

I've done blind test before and even then, it's still contaminated with random results (i.e. I can get 18/20 correct responses on a given day then goes to 9/20 correct the following day).

Random results are random results. You can’t cherry pick the best numbers out of the bunch. All of the tests count equal. If you can’t hear a difference your picks will be all over the place, some close, some far.

Do 100 tests and average them. All of them.

 

[BS5711]

It’s incredibly more subtle than any A/Bing I’ve ever encountered. I usually find myself becoming unable to reliably tell as the test progresses. The switching of DAC filters from internal DAC to a software upsampling like HQPlayer is more readily audible than Goldensound’s filter samples on his website

Can you reliably pass a blind A/B comparison between software upsampling and DAC upsampling ?

If yes then there is obviously some genuine audible difference.

I am not debating that differences with audio electronics could or do exist, I really don't care, only the test method used to determine audibility.

If you can pass a robust blind test reliably and reasonably repeatably then that is solid evidence of audibility. If you can't and then look for reasons to discount a legitimate test as a viable datapoint that is where our paths deviate and you start looking for what I see as irrational reasons why blind listening is different to sighted listening.

For me blind and sighted listening are hugely different because blind removes a factor (sight obviously) that creates faulty perception while you rationalise that the difference in blind versus sighted listening means that sight is inherently necessary for proper hearing. To me that is utterly flawed logic to avoid accepting that a lot of what you HEAR is actually only what you PERCEIVE and not due to equipment differences at all

But we have been over that round and round, it has got a bit old, you know my thoughts, I know they don't matter to you, lets call it a day