Jerry Lee DBT

[okiemax]

To gain a better understanding of double-blind testing methodology, I have been experimenting with the shuffle play option on RealOne Player's free service for my computer to determine if I can identify different but similar recordings which I have saved to a playlist. The shuffle play feature will play all tracks on a playlist in a random way, and although I don't know that it works like a true random number generator, it seemed satisfactory for my purpose. A disadvantage of using RealOne Player for double-blind testing is that each trial must be scored when completed and results must be tabulated manually. But hey, it's free.

I chose recordings of two different sessions of Jerry Lee Lewis singing High School Confidental (Bear Family,1994) for my experiment. Session A plays 2 minutes and 37 seconds, and Session B plays 2 minutes and 40 seconds, but I couldn't hear a difference in the two sessions in preliminary listening. In double-blind listening test #1, I correctly identified the session in 7 out of 16 trials, which of course is no better than random, and statistically a null result. I was beginning to think that if there was a difference it was not audible to me.

I switched to contiuous shuffle play, turned the volume down, and went about attending to e-mail and other matters on the computer, hoping the difference between the two sessions eventually would be evident. It worked! I didn't time how long it took, but suddenly it was obvious that there was something different about three-quarters of the way through the one playing at the time. So, I scored a perfect 16 out of 16 in double-blind listening test #2.

For test #3, I only listened to about the first one-half of the song, and nulled again, correctly idenifying only 6 of 16. The next day, I tried what worked previously, listening casually while doing other things. Again I didn't time it, but it probably took under an hour for me to pick up a difference close to the beginning of the song, which was all I needed to score perfectly on test# 4.

What did I learn from all this? Don't put too much stock in nulls. Relaxed or casual listening is more likely to reveal differences for me than test listening. I won't care if I never hear High School Confidential again.

[JoeW2]

Interesting post, okiemax. My experience agrees with your assesment. I have made the worst audio choices of my life when I tried to make direct comparisons intellectually. An audio presentation is a complicated and nuanced event. There are just to many things to listen for in order to make a managable check list.
I've had much better results using what I call 'casual distraction.' I'll do other things as I listen, and pay as little intellectual attention as possible. In other words, I get as far away from thinking about whatever device I am trying as possible.
If I am distracted by something that sounds really good, it's good. Likewise, if I am distracted by something that annoys me, it's bad. If I go undistracted, that's a result too.

Don't get me wrong. I do enjoy dedicated listening where I do nothing but focus on the music. Nor, do I completely discount this when assesing gear. If anything, dedicated listening usually confirms my 'casual distraction' assesments. OTOH, 'casual distraction' has contradicted my assesments from dedicated listening more than a few times. In and of itself, I've not found dedicated listening to be the most effective way to evaluate equipment. YMMV

[Pat D]

What did I learn from all this? Don't put too much stock in nulls. Relaxed or casual listening is more likely to reveal differences for me than test listening. I won't care if I never hear High School Confidential again.
Training increases the sensitivity of a DBT and that's what you did. However, once you learn the differences, you should be able to tell the appropriate passages apart in a DBT.