Quote:
Originally Posted by Fran Guidry
Just in case someone wants to do objective blind testing, let me point out that there is a technique called double blind ABX which deals with those two issues, and there are free tools to implement double blind ABX.
By double blinding the source expectation bias is removed.
The tools I've used allow looping and selection of starting points, so clips can be compared in detail at different locations, which deals with the focus issue.
http://lacinato.com/cm/software/othersoft/abx
It turns out that doing this carefully (minimum of 16 trials with full attention and concentration, using level matched same source material) is a lot of work but can be quite revealing.
Fran |
Well perhaps both Yes and No
Yes I have known about ABX testing for some time now, and Yes It is extremely useful in many testing circumstances, particularly in placebo type tests and tests that measure group reactions . Where it is unknown to both the participants and person conducting the test as to who is in the control group and who isn't.
Yes the link for the ABX test software is helpful thanks
Given the nature and logistics of audio testing I was thinking it may be somewhat less important over just single blind testing. (provided the person switching is not giving off some kind of visual cues )
And Yes blind testing certainly does serve to eliminate the positive expectation bias (those who believe they can hear a difference) But unfortunately. No it probably has significantly less if any advantage to eliminate negative expectation bias (those who believe there is no difference) because in that circumstance the "objectivity " you speak of, may not actually be coming into play. It can be a matter of the belief in "I don't hear a difference" trumping any advantage or accuracy of blind multiple tests.
Then as far as "deals with the focus issue" depends on what you mean by "deals with" Helps with focus then Yes . While certainly loping back and listening for very short clips (a few seconds ) starting at the same point can definitely help eliminate the inherent inaccuracy of listening to a whole clip of anything say 30 seconds or longer then listen to another clip. It still does not eliminate eliminate the fract that we the listener can still change focus from clip to clip and as James Johnston noted, when we listen and focus on one specific detail ( high frequencies for example ) we tend loose focus on the rest of sound spectrum.
And the other two factors that come into play and double blind really does nothing for, is, 1. the ability or lack thereof to actually listen critically,
to know what to listen for, and what it is you may actually be hearing or not hearing. Critical listening is learned skill set just like mixing, and is by no stretch something inherent to every musician regardless of years of experience.
And 2. the other factor is , each individual's actual hearing range.
If you can't hear above 10k or 12k or 13k then you may have no idea what may or may not be happening to the high end for example . Is it accurate and airy, detailed and present or harsh and sibilant, or dull, or non existent ???.....
So while blind testing is no doubt a very good exercise and I would encourage it , it may not quite be the panacea be all to end all, that some believe it to be.