DAC blind test: NO audible difference whatsoever

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
just curious, if possible to do a blind test, can you tell whether you are driving a Ferraris or Toyota?


I thought about that, seriously i mean. :)

I would not recommend driving blinded ;) ...but as passenger, yes that would be possible (and probably fun to organize).


Here is what would happen:

1. You're blindfolded. Both car's engines are off. They're parked in a position you don't know.

2. Once you touch the door handle, you might be able to spot the difference. If it's a McLaren or a Lamborghini, the way the door opens will tell instantly. If you're not allowed to touch the handle, or the car, next: step 3.

3. Helped, you sit in the car on the passenger seat. You might, right there, feel the difference in the seat. Let's call that ''the butt feeling''. If you're butt-insensitive, next step 4.

4. You will touch the dashboard and might know instantly that alcantara is not available in a Toyota Corolla. If you are not allowed to touch, then next step 5.

5. The driver's will now start the engine. That's a big tell right there. A japanese 4-cylinders from Toyota cannot sound as a german flat-6, or an italian V10 or a turbocharged english V8. If you are not just blindfolded but also equipped with noise-cancelling stuff, then step 6.

6. At low speed, the vehicle will give you some tactile sensation. The suspension will be a big tell. If you don't feel it, then next step #7.

7. Driver will push the machine, while the Toyota will have limited dynamics, the exotic car will get you some 1.5 G at braking, cornering and maybe acceleration as well. A very very big tell, i would say. You might also feel some sickness, which is another big tell, especially if some vomit is implied.

8. At stoplights, in an exotic car, people will roll down their windows and talk to you, whether you're in the middle of a very important blind test, they won't care the slightest. In fact, if you wear blindfold as a passenger, in a Lamborghini, they might call the authorities and you'll end up hearing a police car, or two, chasing you. If the driver don't stop or you're still undecided whether it's a Toyota or an exotic, go to step 9.

9. At this point maybe you're on a highway flying at 330 kph OR in a Toyota at 190kph, you HAVE to know somehow, the easiest way would be to guess if the police car's sirens are getting louder or not. Tell immediately the driver as soon as you finally know the answer. If you don't: step 10.

10. You wake up at the hospital. It's the Toyota. You don't wake up. It's a Porsche 918 spyder.
 
Last edited:
Jokes aside, the DAC is mono-sensorial (can we say that?) opposed to a multi-sensorial pleasure-giving-machine such as a Ferrari.

The only job we ask from a DAC is related to some sonic result.

I don't think any aesthetic or tactil consideration can be. I wouldn't trust anyone who caress his DAC or stare at it for more than an hour.
 
I hope that the readers of this thread are able to see that one cannot logically conclude, based on the OP’s exercise, that there are no audible differences between the devices he tested.

The OP seems to be convinced of his conclusion but has declined to answer the critique of his approach. He has defended its validity with simple assertions, false analogies, ridicule, personal attacks, etc.

The OP doesn’t understand the distinction between: (1) being unable to discern a difference between A and B; and (2) concluding that A and B are identical. A key element of the test type used by the OP is understanding that rejecting a null hypothesis doesn’t equate to accepting, as fact, the opposite of the null hypothesis.
 
Last edited:
frugal-phile™
Joined 2001
Paid Member
The OP doesn’t understand the distinction between: (1) being unable to discern a difference between A and B; and (2) concluding that A and B are identical. A key element of the test type used by the OP is understanding that rejecting a null hypothesis doesn’t equate to accepting, as fact, the opposite of the null hypothesis.

+1

dave
 
I hope that the readers of this thread are able to see that one cannot logically conclude, based on the OP’s exercise, that there are no audible differences between the devices he tested.

The OP seems to be convinced of his conclusion but has declined to answer the critique of his approach. He has defended its validity with simple assertions, false analogies, ridicule, personal attacks, etc.

The OP doesn’t understand the distinction between: (1) being unable to discern a difference between A and B; and (2) concluding that A and B are identical. A key element of the test type used by the OP is understanding that rejecting a null hypothesis doesn’t equate to accepting, as fact, the opposite of the null hypothesis.

Nevertheless, this is 4 peoples opinion, with or without explanation, so take it as is.
 
mmerrill99 said:
It's human & understandable that many people ignore the pitfalls in perceptual testing because it delivers the results they want - the null results which are the inevitable outcome of ignoring these pitfalls.
But this test delivered the opposite of what the tester was expecting. One might also say that its human and understandable that many people ignore the results of perceptual testing because it delivers results they don't want - the null results which are the inevitable outcome of physics and psychoacoustics.

This focus on "training" is one of your blind spots - the training is only needed because this type of 'testing' is not 'normal' listening & in order to participate in such a test in an objective way, it is necessary to 'train' oneself to identify EXACT differences between two audio files & using this 'training' to then enter the blind test, only focusing on this EXACT difference. So have you done this in your listening or are you expecting some "obvious difference" to jump out & grab your attention?
The usual complaint about ABX is that it delivers bad (i.e. not what audiophiles expect) results because it is unlike ordinary listening. Your proposed solution is to make it even more unlike ordinary listening?

Markw4 said:
But, as a thought experiment, suppose you brought the DACs over here, and using my amp and speakers, I could repeatedly identify DAC differences using ABX testing. What would that mean for the truth of what you have thus far proved? Would it mean it's true now, but it might be false tomorrow that the DACs you tested have no audible differences? Would it actually ever have been true?

I think the problem I just tried to illustrate above is why some people might say that the ABX test you conducted did not prove the DACs have no audible differences. Rather, it failed to prove the DACs do have audible differences. Hopefully, it makes more sense to say it that way when viewed over time and with different experimenters.
If you could reliably hear difference it would not affect the truth which Jon has discovered. He found that four people in a certain environment on a certain day could not hear a difference (to the extent of his statistics). You might find that a different person in a different environment on a different day could hear a difference (to the extent of your statistics). However, he has not proved they are the same; you have not proved they are different. Can we drop the 'prove' word? You cannot prove anything with statistics, because randomness (if present) can always fool us.
 
Doppler9000 said:
I hope that the readers of this thread are able to see that one cannot logically conclude, based on the OP’s exercise, that there are no audible differences between the devices he tested.
True. But it is consistent with his results that there may be no audible differences between them. Do you accept that?

The OP seems to be convinced of his conclusion but has declined to answer the critique of his approach. He has defended its validity with simple assertions, false analogies, ridicule, personal attacks, etc.
As I understand it, his conclusion is that he and his friends could not hear a difference - which is not what he was expecting to find. The validity of his test has been attacked with simple assertions, false analogies, ridicule, personal attacks etc.

The OP doesn’t understand the distinction between: (1) being unable to discern a difference between A and B; and (2) concluding that A and B are identical.
We already know that A and B are not identical. They are different circuits so must have slightly different outputs. The conclusion is that they were indistinguishable by ears alone under those test conditions. Why can't people simply accept this test result as one valid data point in the quest for genuine human knowledge?
 
But this test delivered the opposite of what the tester was expecting.
Do you honestly believe that the o/p isn't on a crusade against what he considers snake-oil? Everything he has stated in his posts reveals this.
One might also say that its human and understandable that many people ignore the results of perceptual testing because it delivers results they don't want - the null results which are the inevitable outcome of physics and psychoacoustics.


The usual complaint about ABX is that it delivers bad (i.e. not what audiophiles expect) results because it is unlike ordinary listening. Your proposed solution is to make it even more unlike ordinary listening?
If you are happy with bad tests because they deliver the inevitable null results wanted - I hear what you're saying. Don't categorize what the o/p is doing as 'perceptual testing'. Yes, perceptual testing does require a different type of listening - all I'm saying is to follow the already established guidelines recommended for such testing. If the o/p or you are happy with not following these guidelines - I hear what you are saying.

If you could reliably hear difference it would not affect the truth which Jon has discovered. He found that four people in a certain environment on a certain day could not hear a difference (to the extent of his statistics). You might find that a different person in a different environment on a different day could hear a difference (to the extent of your statistics). However, he has not proved they are the same; you have not proved they are different. Can we drop the 'prove' word? You cannot prove anything with statistics, because randomness (if present) can always fool us.
Well he has tried to generalize from his results & I was pointing out the other group who are equally representative of audiophiles & who achieved a non-null result. Yet he doesn't seem to accept this alternative result even though he asked for just one example of a positive result which I provided
 
Last edited:
We already know that A and B are not identical. They are different circuits so must have slightly different outputs. The conclusion is that they were indistinguishable by ears alone under those test conditions. Why can't people simply accept this test result as one valid data point in the quest for genuine human knowledge?

The o/p seems to want to make changes to the test conditions to try to tease out differences but yet he rejects the whole area of the recommended guidelines for perceptual testing, preferring instead to tweak things which have little chance of revealing any differences - the problem lies in not recognizing what is involved in this type of listening test.

What he is doing is fine for answering some question but it's not addressing the question he purports that it is.
SY would point this out if he was here!!
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.