My pleasure. Fascinating company as well, formed by musicians with a 'cost no object' approach.
Last edited:
Comparison through blind or ABX test is a wrong way to judge sound.
Listen and judge differences through the memory of a previously perceived feeling is not possible .. is not reliable ..
also .. our information processing system, the brain, is deceived by many mechanisms.
the best method is to listen quietly many times, for many days, in humoral and different physical conditions, preferably when the judgment was consolidated, we came to the right decision.
not choose never through test of skill or as a blind ABX test.
Listen and judge differences through the memory of a previously perceived feeling is not possible .. is not reliable ..
also .. our information processing system, the brain, is deceived by many mechanisms.
the best method is to listen quietly many times, for many days, in humoral and different physical conditions, preferably when the judgment was consolidated, we came to the right decision.
not choose never through test of skill or as a blind ABX test.
Last edited by a moderator:
You know, everybody should decide for themselves what appears to be the
right way to evaluate stuff. Once a friend of mine said to me, after the extensive
and long term listening to a couple of projects I've made for him and myself, that
all of my speakers sounded the same! The loudspeaker units were all different brands
(Visaton, Monacor, Focal and Dynaudio). So if JonBocani says that drivers once
EQed, sounded similar to the listeners, I am not surprised.
Me, I don't plan to buy dsp platforms and a bunch of amps to do the same thing
I can do passively. Not just yet. Maybe in the future when I become a wimp.😀
right way to evaluate stuff. Once a friend of mine said to me, after the extensive
and long term listening to a couple of projects I've made for him and myself, that
all of my speakers sounded the same! The loudspeaker units were all different brands
(Visaton, Monacor, Focal and Dynaudio). So if JonBocani says that drivers once
EQed, sounded similar to the listeners, I am not surprised.
Me, I don't plan to buy dsp platforms and a bunch of amps to do the same thing
I can do passively. Not just yet. Maybe in the future when I become a wimp.😀
Nonsense.
... and this is exactly how the ABX test can be done in your case! Evidently, nothing is wrong with the ABX test methodology.
Actually, with a proper training we can shorten the process of perceiving these quality differences, from many days to minutes...
ABX test is the only certified way to test if one can detect the difference between A and B. If one can not, it means there is no difference "for him". It doen't mean there is no difference. Very small number of people have significantly better ears than the other, and we call them golden ears, but the training can improve one's hearing ability.
I do not disagree with the importance of the long time listening test, but it is not comparable to ABX, they are just different. The long term test can clearly show one's musical preference better than ABX.
The point is, if one can't detect the difference on ABX, his preference on long term test is most probably fooled by or biased by something other than sound by itself, such as equipment appearance, price, brand or his own theory.
I do not disagree with the importance of the long time listening test, but it is not comparable to ABX, they are just different. The long term test can clearly show one's musical preference better than ABX.
The point is, if one can't detect the difference on ABX, his preference on long term test is most probably fooled by or biased by something other than sound by itself, such as equipment appearance, price, brand or his own theory.
I agree, ABX blind test not sensitive at all for listening for differences. It is only a meaningful test if one finds a difference.
We can have a Zika virus test that is 70% sensitive, 100% specific. That means, if you test positive, then you have the virus. If you are negative, that doesn't mean you don't have the virus. This is what sensitivity means.
Last edited:
I'm not sure what you agree about with him. Your statement is somewhat opposite to what he said.
If the ABX result is obvious, trained and experienced ears can instantly tell which is his favorite, I guess. If ABX result is not so obvious, long term test may be meaningful.
I do believe ABX test is very meaningful, because it can solve the common and worst problem in consumer audiophile world, one think A is better than B even if he can't actually tell the difference...
Hi,
Maybe one advantage from DSPs : time alignement in the low area between the medium and bass units ?
There's NO time limit on ABX tests, or any other double blind listening tests. The ONLY requirement is that the test subject doesn't know what X is.
Now if people think that you need to test for long time periods, its up to them to come up with the evidence. Waiting patiently....
But does it make sense that there could be a difference to the sound and yet ABX cannot detect it? There's been ABX tests with rock music and people failed to tell when even 10% harmonic distortion was added. People think everything sounds the same on ABX, which is simply not true. People try to listen for something specific and cannot find it sonically, but the difference in sound can affect them emotionally. You cannot use emotions to reliably distinguish between 2 choices as it is highly variable.
Your father can have alzheimer's dementia and you cannot tell right away, as social graces are intact, even though he has lost half his brain cells. But over time, you notice the poor judgment and forgetfulness. So the difference is there. Doctors can use specialized tests like MMSE that can find the disease 80% of the time, but a negative test can still be wrong. You may still feel your father is somehow "off".
If there IS a difference, it doesn't matter what kind of difference, people can be trained to detect it faster. Of cause its up to the people who think there's a difference to come up with the evidence so others can be trained to detect it to. Waiting patiently....
Hi,
I am not denying advantages to dsp, I was simply saying that basically I don't need
any of it right now because I can handle all the usual issues the passive way. Time
alignment between midwoofer and woofer on a flat baffle is not critical for there
is little Z offset between similarly built units and you can see that is true among
competent designs showing symmetrical filter in that region. If there would be
any need for it, it could be addressed either by physically moving one of the units
or by adding another passive component to introduce an extra phase shift.
This is an interesting observation. Thank you for posting.
I must agree that ABX test is meaningless in some cases. As I stated before, ABX test is basically for experienced people. I believe most of the ordinary people would fail to tell the minor difference, let's say such as DA converter difference.
But, I also believe that all of us here are not the ordinary people, and we must (or should) tell the extremely minor difference on ABX test. 🙂
I am sensing topic drift in this thread. The thread started with a very interesting assertion, that given appropriate eq and level matching it was impossible for 12 people, including people with experience in designing audio equipment, to tell the difference between highly regarded drivers e.g. ATC domes, and lesser drivers, including a woofer. So far the only valid criticism of the test method I have read is that the drive unit off-axis response may vary and allow differentiation. This is a fair criticism and easy to address. ( apologies if it has been, I have not made it all the way through all the posts at this time.)
Much of the criticism to date seems fall into three camps.
1) there must be a flaw in the test method
2) preference has nothing to do with flat response
3) it's using eq, which is intrinsically flawed.
The first criticism has some merit where it is based on previous empirical testing.
The second is a personal preference and has nothing to do with the test ( full disclosure- even with a large ESL and a deqx to switch in an out, I prefer to tailor my system for pleasure rather than accuracy)
The third is frankly absurd. Unless you are listening to live music or a purely analogue recording with no eq applied via a purely analogue system, you are all listening to eq'ed music. It's a staple part of the industry used to enhance the quality of the recording. With a very few exceptions all recorded music is eq'd. Even 1950's and 1960's recordings were eq'd using analogue filters.
What this test has revealed is very interesting and deserves serious consideration. If 12 people can't tell the difference between specialist midrange drivers, an ATc dome and a woofer when equalised and level matched that is telling us something significant. We may not like it, and some may argue it may be flawed to some extent, but it's a valuable result
Much of the criticism to date seems fall into three camps.
1) there must be a flaw in the test method
2) preference has nothing to do with flat response
3) it's using eq, which is intrinsically flawed.
The first criticism has some merit where it is based on previous empirical testing.
The second is a personal preference and has nothing to do with the test ( full disclosure- even with a large ESL and a deqx to switch in an out, I prefer to tailor my system for pleasure rather than accuracy)
The third is frankly absurd. Unless you are listening to live music or a purely analogue recording with no eq applied via a purely analogue system, you are all listening to eq'ed music. It's a staple part of the industry used to enhance the quality of the recording. With a very few exceptions all recorded music is eq'd. Even 1950's and 1960's recordings were eq'd using analogue filters.
What this test has revealed is very interesting and deserves serious consideration. If 12 people can't tell the difference between specialist midrange drivers, an ATc dome and a woofer when equalised and level matched that is telling us something significant. We may not like it, and some may argue it may be flawed to some extent, but it's a valuable result
Last edited:
What this test result showing us is the frequency response is the ONLY important factor to determine the sound quality. This conclusion is soothing new to me, because the past result of the similar scientific study I read have concluded that frequency response is the MOST IMPORTANT factor, not the ONLY factor.
What I really want to know is how trivial the other factors, such as distortion etc.
What I really want to know is how trivial the other factors, such as distortion etc.
You know what Dr.Geddes thinks of non-linear distortion, so much that
he won't trouble his mind much over it when choosing speaker units,
although drivers of his choosing already are of good quality.
he won't trouble his mind much over it when choosing speaker units,
although drivers of his choosing already are of good quality.
People do not think everything sounds the same.
But you must find a threshold when you organize such test. Back in 2010 when i organized similare blind test but with MP3's v.s. AAC v.s. HD files, i had to find the threshold, which would happen to be 64kpbs MP3. Even 128kbps was not enough of a threshold... 192kbps+ nobody was able to spot it, not even remotely.
The great thing about ABX tests is they cannot lie. People DO or DO NOT identify differences. Saying you can spot differences is no proof at all. ABX is that proof.
There is the DAC thing as well.
I'm using a somewhat dubious source/DAC and i plan to try a nanodigi+high-end DAC very soon. Maybe that would help. Maybe not. 😉