Proof Non-Blind A-B Listening Tests Are Nonsense

[Robert Dunn]

I am so laughing at myself....

For a while now, I've been testing my current project with one new speaker on the left channel and an older speaker on the right. I've been doing A-B tests by turning the balance knob back and forth and listening intently. Left-right-left-right.

Meanwhile, I've been repairing an old Realistic LAB-2100 turntable. So I put on some beloved old vinyl and turn it up. I start fiddling with amp tone controls and, wow, when I turn up the bass boost, there's a clear shift in the mids from right to left. Since this is unexpected, I turn the boost up/down/up/down several more times and there's still a clear shift between the speakers.

So I turn the balance knob all the way to the left and try the bass boost knob again. Oddly, I'm still hearing both speakers. Turn the balance all the way to the right. Still both speakers.

Eventually I realized that I had the "pure direct" switch on (Yamaha amp feature -- disables tone controls, balance controls, etc.). The bass boost had never done anything; I had unconsciously learned to associate turning knobs with hearing something different.

The lesson is clear: An A-B listening test is meaningless if not blind.

[Soundslike]

That's a great story -- thanks for sharing it. Perhaps that should be referred to as the Dunn Phenomenon...

[daryl]

That's a great story -- thanks for sharing it. Perhaps that should be referred to as the Dunn Phenomenon...

Or N-rays, or just the emperors new cloths.

Julian Hirsch knew this long ago.