No, by making a change, you may expect a difference and hear it, even though none occurred. I have fallen victim to this a number of times, hearing a difference, ohny to realize the change was not active! The power of suggestions is incredible here.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Are DBTs the defacto standard for Audio Science research?
- Thread starter AJ Soundfield
- Start date
Where that type of comment fails to convince, is that it assumes that the system does not change its behaviour because of the change of one part of it - that everything outside of the speaker is a "black box", invariant in its behaviour at all times. A trivial example: two speakers have widely differing impedance characteristics, one of which causes the power amplifier to somewhat misbehave.
In this forum we do care. We want to discover the true fidelity differences, not what is attributed to what something looks like. It is the way we can find cause and effect. There is just no other way.
The people who are swayed by the looks will find themselves in situations where that factor wears out and all of a sudden the system does not sound as good.
Okay, then it does fall under the "bias-controlled" umbrella?
OP
AJ Soundfield
Major Contributor
- Thread Starter
- #25
That irony is that the ad hoc audio show "test" jkeny cited was clearly an A/B test at best, not an ABX as he claimed.
Existing scientific knowledge such as the expressions of the conversation mass, momentum and energy and what that leads to is common when considering sound. It has nothing to do with sighted or unsighted tests. If you consider only the tiny fraction of audio research involved with sound perception then again one would work with what has been established over the past century or two and expressed in, for example, audibility thresholds. As I mentioned before blind/sighted tests has no meaning when detached from the hypothesis being tested. The form of perception being measured may, for example, involve the subject knowing what is going on in which case a blind test would make no sense. The failure to define the quantity being measured is yet another indication of the major loss of plot by audiophiles.
Mivera
Major Contributor
When you say "we" I think you might be alone based on what I've seen around here. But I'm on the same page as you, since the system like what I shared in my last photo, has been very similar to how my various system's I've built have looked for several years. However, when I compare them to the shiny boxes I find at the local hifi shop, they sound far superior to my ears. So perhaps the problem is the people, and not the systems. I think you are looking at things from the wrong angle.
OP
AJ Soundfield
Major Contributor
- Thread Starter
- #28
Science is keenly aware of those measured parameters, unlike ignorant audiophiles who can't measure. You have zero evidence there was any misbehavior and it's irrelevant because if there was it would have not changed whether sighted/unsighted.
If the speaker sounds different, how do I know that it's due to the speaker, rather then the amplifier changing its behaviour because of the differing impedance characteristics it sees?
OP
AJ Soundfield
Major Contributor
- Thread Starter
- #30
That's a Red Herring, auditory perceptual testing is based on (mainly) blind listening tests. Most of what we discuss on this site (or at least what I discuss) is audio based and thus based on blind listening tests.
OP
AJ Soundfield
Major Contributor
- Thread Starter
- #31
Read the and comprehend the test. It the same speaker, just rated differently sighted vs blind
There were four speakers in fact, which were to be rated against each other. How does the reader of the report about this test know that the amplifier used for the test was invariant in its behaviour with all four speakers. If I had a suspicious mind, I could say Harman used an amplifier which they "knew" would behave differently with the competitor's speaker ...
OP
AJ Soundfield
Major Contributor
- Thread Starter
- #33
Just your overactive imagination, as usual
Phelonious Ponk
Addicted to Fun and Learning
- Joined
- Feb 26, 2016
- Messages
- 859
- Likes
- 230
You're right. It looks like an unfinished DIY project, does not inspire confidence, and sets expectations low. A very good reason for you to preference test blind and, if it tests better than high-priced DACs, find yourself a nice, shiny box to cut your logo into.
Tim
Phelonious Ponk
Addicted to Fun and Learning
- Joined
- Feb 26, 2016
- Messages
- 859
- Likes
- 230
He is not alone. He is with those who understand that in the scenario above, your DIY projects may have sounded better, but you almost certainly had your expectations set by factors other than the case, or lack thereof, and those biases need to be addressed as well.
Tim
Last edited:
Testing without knowledge is to me good enough. So, blind in the fact that you don't know what specific component is playing, etc. For example, I am sure you all have played around changing a circuit or part and listening for the "difference". I do it plenty enough, and I have to keep switching things around, or ask my wife to attach wires, etc, then cover with a cloth, so I can just switch back and forth to my hearts contentment.
Funny, when I can hear a difference, seldom do I know if it is the better or not, given lack of a true reference, always have to fall back to measurments, I mean there comes a point where the difference is just too small. Of course, differences can add and subtract, so it all goes pear shaped pretty damn quickly. DBT are tough to do right, but done right, they are pretty air tight for the thing they are specifically testing. I have never been to or done a true official DBT test though.
Funny, when I can hear a difference, seldom do I know if it is the better or not, given lack of a true reference, always have to fall back to measurments, I mean there comes a point where the difference is just too small. Of course, differences can add and subtract, so it all goes pear shaped pretty damn quickly. DBT are tough to do right, but done right, they are pretty air tight for the thing they are specifically testing. I have never been to or done a true official DBT test though.
Phelonious Ponk
Addicted to Fun and Learning
- Joined
- Feb 26, 2016
- Messages
- 859
- Likes
- 230
That the preceding chain may behave differently, with different speakers - so, what's causing the audible difference then? If someone, like Harman, wanted to test this out "properly" then they could tap the signal at the point of the speaker terminals under live volume conditions for the test pieces, and run the signal captured there, from two speakers, through a comparator of some form. If there no abnormalities then there should always be a null, under all conditions, throughout the music or whatever track - if there is some difference, how much is acceptable?
Tom, the technique I use, if I think there is a difference, is to stress the system more: high energy, complex, treble infused, higher volume - this teases out which way the difference goes. The same process a car manufacturer uses to test a new design's suspension - build a test track with the worst combination of road surfaces, to confirm competence.
Similar threads
- Locked
