Not sure what you are interesting in knowing?
Brains are in large part pattern recognition machines. Some patterns are innate, such as certain visual characteristics of spiders and snakes. Some patterns for behavior recognition of other humans also appear to be innate.
Brains are sort of pre-organized to some extent prior to birth, and they are sufficiently plastic to self-organize in many ways after that. Much of how we hear music can be traced to early-life exposure, such as the propensity to develop perfect pitch. Other patterns are learned to a large extent in later stages of brain development, such as preferences for certain types of music.
Myself, I hear shortcomings in data converters to a degree other people I know don't. But, some of them hear other things I don't, or am not very good at.
So, what matters to you?
I'm just trying to point out that there is a strong difference in between having perfect pitch or the ability to discern your mom's voice and the ability to perceive shortcomings in DACs.
To discern patterns, we need inputs. And it would be ridiculous to think that our physical auditory system is infinite in resolution. So my question is simply: do we have indications, beyond sighted listening, that the very small differences in between DACs can actually be perceived as inputs ?
Which is not a question in most other situations.
edit: but as you said, the information might simply be unavailable at this point.
To discern patterns, we need inputs. And it would be ridiculous to think that our physical auditory system is infinite in resolution. So my question is simply: do we have indications, beyond sighted listening, that the very small differences in between DACs can actually be perceived as inputs ?
Which is not a question in most other situations.
edit: but as you said, the information might simply be unavailable at this point.
Last edited:
Indications, yes. For those who like ABX examples, differences in terms of intersample overs are detectable: Intersample Overs in CD Recordings - Benchmark Media Systems, Inc.
EDIT: Actually, I'm trying to find the link to where they talk specifically about ABX, but they do use ABX at Benchmark and my recollection is that intersample overs are detectable with ABX. Just have to find it.
Aside from that, I don't know of other published ABX results. Some people do appear able to hear many or most DAC differences blind. They may not use ABX hardware or software, but if differences are small they do some kind of blind checking with a human helper or by other means. They are people who want to make double sure they aren't making any mistakes. However, that's probably a small group. There is probably a much larger group that is very dependent on sighted listening. For small differences, hard to say when they hear something for real and when it may not be so real. And the biggest group of all is probably people who don't hear a difference and don't care.
If you want to hear for yourself and you're located somewhere near me, I have some DACs I could show you. It's not hard to hear actually, although it usually takes people new to it some hints as to what differences to listen for. Usually, the easiest thing to notice is how cleanly recorded cymbals sound. Actually, a lot more is different that just that, but cymbals are typically best to start with. Thing is, when most people listen to music, they don't notice at all how cymbals sound. But then, most people aren't interested in hi-fi. Good bass for movies maybe.
Last edited:
Maybe a little related, here is a warning about something to avoid when doing AB comparisons: The PLL in your D/A Can Taint A/B Listening Tests! - Benchmark Media Systems, Inc.
What makes it especially interesting is they think they may have discovered a new psychoacoustic illusion.
What makes it especially interesting is they think they may have discovered a new psychoacoustic illusion.
A History of Bias? Pt 1
JonBocani has said the he is unbiased and has no agenda with regard to audibility differences.
Is he an open-minded seeker of truth?
Does he believe the humans have sufficiently acute hearing that his "testing" isn't a foregone conclusion?
Does he appear to have biases that would make one suspicious that he is results-driven?
Could it be that JonBocani believes that high end audio is just a scam?
JonBocani, in his own words:
A post referring to ca 2012 or before.
From 2015
He ignored this critique of the ABX approach…
From 2016
Regarding a "test" her was planning:
From:
“World' Best Midranges - SHOCKING Results & Conclusions.”
From:
“The Ultimate Blind Test - The whole speakers/system.”
In this thread, JonBocani writes about his expectations regarding reactions to the results of his “test”. Buffoons will be shocked to find that there are no audible differences, he posits, before he even started the test!
From 2017
JonBocani has said the he is unbiased and has no agenda with regard to audibility differences.
Is he an open-minded seeker of truth?
Does he believe the humans have sufficiently acute hearing that his "testing" isn't a foregone conclusion?
Does he appear to have biases that would make one suspicious that he is results-driven?
Could it be that JonBocani believes that high end audio is just a scam?
JonBocani, in his own words:
A post referring to ca 2012 or before.
From 2015
JonBocani said:
He ignored this critique of the ABX approach…
DPH said:
From 2016
Regarding a "test" her was planning:
From:
“World' Best Midranges - SHOCKING Results & Conclusions.”
JonBocani said:
From:
“The Ultimate Blind Test - The whole speakers/system.”
In this thread, JonBocani writes about his expectations regarding reactions to the results of his “test”. Buffoons will be shocked to find that there are no audible differences, he posits, before he even started the test!
JonBocani said:
From 2017
JonBocani said:
Last edited:
Thanks, yes this is interesting but seems also somewhat self-contradictory:
In the top of he piece, John Siau states:
Which I understand as meaning that only switching between a signal source such as a sine wave will reveal this phase shift but using music as a source, won't.
Yet he later seems to contradict himself & refers to hearing differences between music :
You asked for papers about summary statistics in auditory perception & I pointed the way to find these.
Now you switch to a totally different angle/question - if I can paraphrase - how can any new knowledge about how auditory perception works matter when differences between DACs are below audibility (according to you). Have I paraphrased you correctly?
If you can prove that differences between DACs are below audibility your question has validity but until then it's just your opinion which has been shown to be false by the Benchmark example for starters.
I think he may have been trying to say that when the pitch shift occurs, it goes by quickly enough and is small enough in magnitude that no pitch shift event is noticeable by conscious awareness. But, after the pitch shift event has concluded, conscious awareness then mis-perceives the apparent brightness of the immediately following program material.
You shouldn´t be so sure about that, as the ability to discern mom´s voice points directly to the long term storage of auditory input.
Btw, that categorization is an important factor while transferring content from working memory to long term storage should be an undisputed fact.
Isn´t that sort of circular running? The usual understanding is based on a certain type of model about the physiological working mechanisms of our hearing sense. The normal thresholds of hearing were gathered in (usually small sample) tests and from these results people are concluding if a difference between two DUTs should be audible according to this model, if the measured differences are below or above the mentioned thresholds.
To check if these model based premises are correct we are doing perceptual evaluation tests, but that only makes sense if we are really testing the model properties.
See for example the recommended length of music samples; the recommended range varies from <5s - 25s , but you´ll have a hard time to justify this range when comparing it to the various models of auditory memory.
The next point is to consider the influence of any test situation; as said before psychophysical tests are done for roughly 150 years and science knows therefore a lot, but this knowledge base is routinely ignored by the people doing tests like this one we are discussing.
That´s the reason why i questioned DF96´s usage of the term "ears-only test" because we _know_ that still a lot of bias mechanism are at work if people are participating in controlled listening tests including the "blind" property.
As a carefully working experimenter you have to consider all of these effects to get correct results. The golden rule of tests is to "block out what you can and randomize what you can´t block out".
So now back to your question - sorry for the long sidestep - there is indeed some evidence beside the sighted anecdotical; whenever people are doing controlled tests which include training of the participants under the specific test conditions and/or are using larger sample sizes there routinely pops up evidence for audibility that is not compatible to the model currently in use.
But then of course people who don´t like these results are questioning the methodology, something that DF96 forgot to mention, it works in both ways.
As there is no perfect experiment there is always room for doubts....
Last edited:
My location is publicly available on my profile. I'm actually more in Brussels these days. I've no idea where you are though ? Maybe you prefer to go through PM for that ?
I don't understand how a whole set of ABX tests, which differ only in the strength of the difference, cannot be used to establish the threshold of audibility for that particular difference in that particular test setup. You seem to be proposing an infinite set of tests, with each test establishing the sensitivity of the previous test. Turtles all the way down?mmerrill99 said:
I note your unwillingness to actually answer Earl Grey's questions. Your credibility decreases day by day.
Don't be silly. I was making a perfectly plain and sensible point about the way in which civilisation, books and science mean that it is valid for someone to comment on something which he has not personally experienced. There are people on this forum who appear to disbelieve this; there may be a few of them in this thread.Doppler9000 said:
So you accept that the results may be true, even if arrived at by a method you regard as invalid? If the results had been opposite (i.e. there was a clearly noticeable difference between the two DACs) then you would still be here criticising the method?
I am pleased to have lightened your day. You were the one who asked for near-ideal components in order to test any DACs.kinsei said:
Does JB have any comment to make? He certainly looks like he is on a mission, so unless he has a very good explanation to offer I retract my defence of him (on this point) in my earlier posts. So there are at least two missions in this thread. Could people please declare their interests?Doppler9000 said:
I regard high-end audio as a mistake, but whether it is a scam depends on whether the people selling it know it is a mistake. Sadly, in many cases instead of offering evidence to support their case they merely pour FUD over the opposition's evidence while writing nice stories to encourage their friends.
Of course I realise that bias still exists in a blind test. My concern is that many people still seem to fondly believe that they can be unbiased in a sighted test! Some of them even say so on this very forum.Jakob2 said:
That´s a description of a positive control on different sensitivity niveaus. 😉
Only difficulty is that something like that wasn´t done in this test and actually i can´t recollect any other "ABX" (testing effects we usally discuss in this forum) in which a sufficient positive control was used.
But you seem in this regard to rely on similar "weak" evidence provided that it suits your belief. Bias is always at work....
See for example your comment on the non detectable difference between a tube amplifier and a solidstate amplifier. I´m quite confident that you haven´t seen a comprehensive description of this experiment and in that case it is obviously impossible to draw any conclusion.
I´d assume that it might have been within the famous "Clark amplifier challenge" but as - according to Clark´s own description of the results - it is extremely unlikely that it had happen, the conclusion is justified that there were some wrongdoing during the test or in collecting the data. As posted before the calculated probability for the test results he described was p < 1 x 10*(-6) , so to question it is justified.
So you of course realise it, but still preferred to use a term that is simply incorrect and misleading? Although an established term for "unsighted" listening tests already exists?
Maybe it is a matter of bias/belief ? 🙂
Again you fail to understand simple logic - how can an infinite number of tests (which I'm not suggesting, btw) of unknown sensitivity tell us anything about threshold of audibility?
I see Jakob2 has already replied making the same point
He has rejected my answer preferring instead to change it into something I didn't say so I'm not interested in wasting my time dealing with this - make of that what you want
Last edited:
kinsei said:
I must have misunderstood you.kinsei said:
He asked the emperor to show us his clothes. The emperor decided to respond by debating the meaning of "clothes".mmerrill99 said: