"paul packer" wrote in message

On Wed, 22 Feb 2006 04:58:03 -0500, "Arny Krueger"
wrote:

You've committed yourself to the
usefullness and veracity of ABX and are therefore not
open to criticism of ABX.

Huh? I can criticise ABX better than most because
trhough experience with it I know it and its faults
better than most.

Good. Perhaps you'd like to list them.

Been there, done that, Paul. But since you are intellectually lazy, poorly
informed and hard to inform on the topic of subjective testing, I'll briefly
go over them again.

For openers, read the "10 requirements" on the PCABX web site and see what
the major points address. Those items pretty well cover the major weaknesses
of ABX and stand as weaknesses of ABX because not everybody is addressing
them or has addressed them.

For example if you take the old Stereo Review DBT articles, compare them to
the "10 requirements" the SR DBT articles come up very weak.

Then there's the slight problem that ABX is a test for differences, and does
not formally address preferences. This is one reason why ABC/hr was
developed.

ABX is also more awkward and time-consuming than other blind testing
methodologies for actually finding hearing thresholds. An alternative
approach is alluded to in the Dolby Jitter AES paper.