Could someone help me to think through my ABX result using Bayesian reasoning?

[manisandher]

Altman DAC? Isn’t that the one touting a Jitter Scrambling Decorrelator thingy? Should that not get rid of any dependency on the incoming signal? Did you have it enabled?

Disabled. I don't like what it does to the sound. (I actually don't much like the sound of this DAC full stop. Picked it up cheap and thought it'd be interesting to play around with. Came with some interesting modifications - a number of Paul Hynes regulators on board, and a Paul Hynes linear 12v PSU.)

 

[Raindog123]

We took analogue captures of the DAC outputting a 10kHz sine. Here is Mans's analysis

Any lower frequencies, by any chance?

 

[KSTR]

FWIW:

View attachment 171986

All equipment fed off single dedicated AC mains circuit.

Altmann DAC? No more questions, Your Honour ;-)

 

[manisandher]

Altmann DAC? No more questions, Your Honour ;-)

Haha... I'll accept your address though

I want to make it absolutely clear that it wasn't my regular DAC back then, and that we didn't use it because of its amazing resolution and sonics. (As I said, I bought it cheap, and it came with some interesting modifications.) We used it for the test because, at the time, I couldn't think of any way of capturing the digital input to the DAC other than via spdif. And the Altmann was the only DAC I had to hand with an spdif input.

So, "no more questions" from your side. But let me ask you one from mine. Any idea how a DAC connected via BNC spdif, isolated from the audio PC via a pulse transformer, might be influenced by a different load on the PC (if indeed that's the mechanism at play)?

 

[KSTR]

EMI would be one guess. Pulse x-formers lose isolation above 10Mhz or so. Plus i've seen x-former coupled outputs which still had galvanic or capacitive connection between grounds

Jitter change is a second guess but less likely IMHO. Would need direct signal analysis on the SPDIF signal to verify.

 

[mansr]

Wasn't there an Altmann Attraction used for part of this? I know Mani has(or had) one.

Is that the bare PCB nailed to wooden plank? If so, then yes.

 

[mansr]

So, "no more questions" from your side. But let me ask you one from mine. Any idea how a DAC connected via BNC spdif, isolated from the audio PC via a pulse transformer, might be influenced by a different load on the PC (if indeed that's the mechanism at play)?

The 10-metre single-ended connection between DAC and amp could conceivably be picking up interference.

 

[solderdude]

Is that the bare PCB nailed to wooden plank? If so, then yes.

But with special regulators...

 

[JRS]

There are a small number of people who are able to contract their tensor tympani voluntarily. Audiometric studies show this has dramatic effects on the transfer function for both air and bone conduction, see Wickens et al. 2017 for instance. But most studies of this system are concerned with tensor tympani myoclonus, which may be a cause of some forms of tinnitus.

Fantastic! Can we all learn this or is like wiggling your ears are curling ones tongue? It was interesting to see the profound changes in bass-lower midrange sensitivity--up to a whopping 25dB! And to actually pull the eardrum in against significant "negative" pressure, testimony to its power. Many thanks for the link.

Now back to the original thread.

 

[manisandher]

Now back to the original thread.

By all means feel free to continue this discussion. I'm sure it's of more interest to people than the Altmann DAC... albeit specially modified .

 

[krabapple]

As MarkS says.. you need more data. This means more ABX testing, properly conducted. You could do this over several days and quite when you think you had enough. It needs to be verified independently though. So a lot of train tickets will have to be paid...

Or -- measurements that showed a difference that is in the realm of known audibility.

Which would simply mean: the setup that was intended to produce identical outputs...didn't. End of magical story. (Though we still have Mani's curious inability to 'clearly' hear what he claims he clearly can hear. Has he ever even addressed this huge red flag?)

Think about it. If there was a real heard difference here, either it is something already measurable and known to be audible (and thus, an experimental method error), or something quite new to science.

That's how unlikely Mani's result is, on its face.

I hope people appreciate that.

 

[manisandher]

(Though we still have Mani's curious inability to 'clearly' hear what he claims he clearly can hear. Has he ever even addressed this huge red flag?)

 

[krabapple]

After 15 seconds you will forget what A is and what B is, and the whole test becomes useless.

OP had only access to X, never to A or B once the test began, which in my opinion, is pointless.

It's already been pointed out (*cough* by me *cough*) that his first two sets of trials amounted to tests of audio memory. A crippled ABX test. It was more akin to an A/B test where the answer to the first two trials was known. (As such, it would have been more informative if the task had simply been for Mani to say whether each X sounded different from, or the same as, the X before it.)

I asked why anyone would do this when intending to do an ABX test. Mansr said because Mani wanted to. It was a waste of time per ABX testing, but already showed that something was wrong with the claim being tested.

A difference 'clearly' heard in routine listening should not disappear even in a quasi A/B series like Mani did.

 

[krabapple]

Pardon, all I saw was you acknowledging that you understood solderdude when he made the same point. But if you've addressed it further, please link to it, I can't claim to have read every post here (yet).

This disparity between your fundamental 'prior' -- your claim that you've always clearly heard a difference between nominally bit-perfect audio presentations -- and the difficulty you encountered doing it under controlled conditions, is a proverbial elephant in the room. So it deserves address.

It also illustrates why it's important to not just report ABX successes/failures, but also describe how the testing 'went'. Was it easy or hard?

It is very common for audiophiles to claim they easily, often immediately, hear X 'sighted' -- whether it be cables, speaker break-in, DAC differences, lossy audio. "Even ,y wife could hear it!" It's been the case for as long as I've been participating in the hobby (40 years). Claiming hearing ability like this is de rigueur among 'high end' journalists writing for Stereophile, The Absolute Sound, Sixmoons, etc. If real, 'aceing' an ABX should be child's play for such people.

So it irks me if possible audible differences unearthed after shall we say, considerable listener effort under controlled conditions (indicating that they are, at most, subtle differences, hard to hear even under optimal conditions) are touted as the evidence for such claims. I'd call that shifting the goalposts to claim a win.

 

[manisandher]

This disparity between your fundamental 'prior' -- your claim that you've always clearly heard a difference between nominally bit-perfect audio presentations -- and the difficulty you encountered doing it under controlled conditions, is a proverbial elephant in the room. So it deserves address.

For the last time... I COULD HEAR A DIFFERENCE IN ALL THE TESTS! But in the first two, I couldn't assign A or B correctly, because I had no reference any more. Really, how difficult is this to understand?

It is very common for audiophiles to claim they easily, often immediately, hear X 'sighted' -- whether it be cables, speaker break-in, DAC differences, lossy audio. "Even ,y wife could hear it!" It's been the case for as long as I've been participating in the hobby (40 years). Claiming hearing ability like this is de rigueur among 'high end' journalists writing for Stereophile, The Absolute Sound, Sixmoons, etc. If real, 'aceing' an ABX should be child's play for such people.

WTF?

So it irks me...

Oh, you're irked are you?

... if possible audible differences unearthed after shall we say, considerable listener effort under controlled conditions (indicating that they are, at most, subtle differences, hard to hear even under optimal conditions) are touted as the evidence for such claims. I'd call that shifting the goalposts to claim a win.

I disclosed everything in the OP, and have been totally forward and honest throughout the rest of this thread. And I certainly haven't shifted any goalposts - I even accepted that the results of the first two test have to be included with the third.

You know what, I've got too much going on in my life right now to be dealing with this crap. I'll not be posting in this thread any more.

 

[mansr]

 

[AudioStudies]

Something tells me you two won't be exchanging Christmas cards . . .

 

[DimitryZ]

For the last time... I COULD HEAR A DIFFERENCE IN ALL THE TESTS! But in the first two, I couldn't assign A or B correctly, because I had no reference any more. Really, how difficult is this to understand?



WTF?



Oh, you're irked are you?



I disclosed everything in the OP, and have been totally forward and honest throughout the rest of this thread. And I certainly haven't shifted any goalposts - I even accepted that the results of the first two test have to be included with the third.

You know what, I've got too much going on in my life right now to be dealing with this crap. I'll not be posting in this thread any more.

Hello.

I don't understand this: you say that the difference between A and B was well known to you - you had a way of identifying the sonic signature of each. So, in the first two trials, once you heard the original A and B upfront, you should have made a note that A is one setting and B another (that if your setup had identity of A and B hidden). After that, you should have been reliably able to identify all subsequent Xs - after all, the difference has already been well understood in your prior experience.

Since you weren't able to do this in the first two trials, it suggests that the difference was subtle enough to become entirely opaque to you in a matter of a few minutes, such that your aural anchor became fully ineffective.

But once the tests were restructured to give you A and B reference back every 30 seconds, you did very well. This suggests that your prior experience had a very short shelf life, but with constant re-referencing, you are able to hear a difference. I tend to think that 9/10 is rare enough that it should be taken seriously.

The other thing that I am not clear on is what is your computer source? You mentioned that the data came to the DAC as SPDIF, which is unusual output mode for most modern computers, USB being much more common.

 

[krabapple]

For the last time... I COULD HEAR A DIFFERENCE IN ALL THE TESTS! But in the first two, I couldn't assign A or B correctly, because I had no reference any more. Really, how difficult is this to understand?

//
" For many years, I'd been certain that I could hear clear differences between bit-identical playback means" [emphasis yours]

" we'd conducted 2 non-ABX tests before the ABX, where I scored no better than guessing
- even here, I was certain I could hear differences, but had no reference against which to assess, so simply guessed A or B in my responses"

"
- I scored 9/10 in the only ABX we conducted
- I got #9 incorrect - we were 15 minutes into ABX at that point, and I was feeling fatigued (we'd conducted 2 non-ABX tests beforehand - see below)"

"It took 18 minutes... of hard concentration"

//

Does not compute.

Ta.

 

[tuga]

The 10-metre single-ended connection between DAC and amp could conceivably be picking up interference.

Would it pick up interference with one playback setting but not the other?