It should get equal critical time. In fact I read far more of it in the "forums", far more.
Pano i could differentiate the fruit files and that as we know was on flawed gear.
the old school studio practice of listening on a "bad" speaker for mono sum and tonal change serves to hone one's listening skills (i miss my Auratone 5c's) and the way it was explained to me years ago by a studio engineer was purposefully listening through a speaker that's not flat( i term it "non linear" but i've been told that's incorrect) serves to either exaggerate differences or homogenize them, can't say it's factually correct but i know it works!
the old school studio practice of listening on a "bad" speaker for mono sum and tonal change serves to hone one's listening skills (i miss my Auratone 5c's) and the way it was explained to me years ago by a studio engineer was purposefully listening through a speaker that's not flat( i term it "non linear" but i've been told that's incorrect) serves to either exaggerate differences or homogenize them, can't say it's factually correct but i know it works!
Yep, I was hoping to do the same with cheap earbuds or laptop speakers. Didn't work for me. 🙁
Exactly.
Agreed. Sighted listening is good for finding small differences, however it may easily get biased and we need tons of self-control not to be beaten by our own biases. DBT is a good feedback for ourselves. And ABX as well. I definitely spend more time with "sighted" listening, but need blind testing not to fool myself.
Yes. And that's something I've learned from blind testing, to be more honest with my sighted tests. Being honest with myself - do I really hear a difference, or do I just imagine it?
And that is the point.
Let's step back a bit from the assumption that ABX is the reality & our other listening is full of bias & misleading us. Why do we think that? Have we proven to ourselves that ABX is actually showing us what is truly audible or is it interfering with our facility to differentiate differences? For our own ABX tests, what controls do we have inside the ABX test that shows us this is the case?
Without such internal checks on the test itself, we are making all kinds of unwarranted assumptions about its validity.
I know this is a sticky area but the Sine wav signals posted by PMA must give people some indication of this factor at play? And this sine wav test is just a single dimensional difference as Jakob2 has pointed out & what we are doing on a ABX test of music is a multidimensional test.
I maintain that unless you can reliably spot the difference multiple times in sighted listening by isolating a specific factor in the sample (like the bass attack) then you are almost 100% doomed to a null result in ABX blind. testing. This takes lots of training to isolate one specific sound element from the sample that is different (lots of trial & error) & when you have achieved this then doing the blind test is just an assurance. The training is not just to be able to spot this specific sound difference but to be able to maintain focus/concentration on 16 or more trials of an ABX test
@ Scott Wurcer - sighted listening is de facto considered unreliable so how would a protocol improve this? Are you saying some protocol would change your mind about its validity?
Theer are very few people that can
You misunderstand. Blind is the reality; ABX is merely one example of that.mmerrill99 said:
It is known that sighted tests can cause people to hear differences which do not exist. Therefore they are not reliable.
However, as you say, there may be a role for sighted tests to possibly identify things which then need to be confirmed (or not) in blind tests.
why multiple times? when it comes to comparisons between two things are we not looking for things easily obvious/different, as in easily detectable right out of the gate? the more rounds you have to go the more the odds increase for mistakes and attention span/focus being lost?
Last edited:
Yes & we know that ABX is a less sensitive test than other blind protocols. We don't know anything much about forum run ABX tests - if you do have any research on it please cite it. But please donlt try to confuse the matter - my reply to Pano was specifically in answer to his participation in this forum run ABX test - you are just muddying the waters with your post. Would you car eto stick to the topic at hand, please? Do you view forum run ABX tests valid & of what value?
As a generalization, yes but this isn't a rule - CAN is the important word in what you say i.e it's not set in stone - see your next point
And as I said "unless you can reliably spot the difference multiple times in sighted listening by isolating a specific factor in the sample (like the bass attack) then you are almost 100% doomed to a null result in ABX blind. testing. " And wen it is confirmed in valid blind testing then your conclusion must be that the sighted listening was not biased. How do you verify that a blind test is valid & that you are not blindly accepting this test as being the true reflection of what we perceive?
If we are only interested in obvious i.e gross differences then I'm not sure what the relevance of blind testing is? But how do you know it wasn't just a guess - by repeating it, you statistically eliminate guess work?
On one hand I completely agree - it's difficult to retain focus & not become bored with multiple tests of the same small excerpt. But, on the other hand, the nature of perception is such that we are not likely to score 100% on many tests that are of a small difference.
You might say that small audible differences are not of importance but that is how advances are made in audio, at this stage & the accumulation of such small changes can lead to much more noticeably audible differences.
I fear my rebuttal/distinction will be misused, but I'll make it anyhow, and it's not in support of Mr. FUDs points. Reality of how we hear is both eyes and ears (and smell, and most definitely when we last ate a meal). In terms of what is audible vs what is not, blinding is definitely needed, but we also have to accept that there is a cost in terms of false negatives/decrease in sensitivity by most of the population.
Similarly, if one is trying to decide what piece to use, then going with whatever makes one happiest (with all the ensemble of biases, sighting, etc included) is the best choice. With the caveat that the rest of us will reserve the right to say any justification past "I like it better" is purely speculative.
All of this is in line with what you've said before and I am just trying to differentiate between what is audible and what is a listening experience. In terms of telling the difference between an original and looped through a tube amp, blinding is the order of the day.
I just wanted that to be said. You can't remove the human element and I can't see any protocol overcoming the integrity and discipline required of the participants to make knowing the choice not bias them in any way.
You seem to miss the point that "de facto considered unreliable" does not necessarily mean it ALWAYS is wrong - you seem to infer that knowing which is which will ALWAYS bias the participant - is that what you mean?
if the differences despite being "small" are "audible" then i would have to say they are important, their "audibility" makes them self evident,no?
it's when intangible's are allude to make a difference that things become absurd,no?
i once hear a marketing tagline to the affect of "it's all the stuff you can't hear that makes a difference!"
Thank you for asking first, I thought I made it clear it requires a lot of integrity and discipline from the participants not impossible but also not easily proven.
OTOH there also is no way to prove that "forum" ABX is ALWAYS wrong, as you say.
The thing is that "audible" is a loaded term in perception - the closer we get to the threshold of audibility the more unsure our perceptual system is & the more the need for doing statistically significant number of tests - perception is such a fickle & fragile thing that something is audible one minute but not so the next - in other words it never is 100& accurate as we approach the threshold of audibility.
But there is another factor - what do we mean by audibility? Do we mean consciously identifiable or would we consider a signal to be audible if it is reliably demonstrated that it constantly shows on fMRI or other tests?
I don't say that forum ABX testing is always wrong - I simply state that we have no way of judging whether a particular forum ABX test has validity.
That's why I entered the thread when the sine wav tests showed that only one member (excluding PMA) showed that they couldn't hear an audible difference for what should be an audible difference - the sine wav
Both Jakob & I have always maintained that one essential element required to ensure a valid blind test was some sort of control which internally verified the test & therefore the results - the sine wav test was an after test kind of validation procedure which I thought exposed something from which a certain amount of learning could be extracted.
Last edited: