Harman Int'l -- all speaker listeners hear alike!

[jackinnj]

Well, not quite, but a recent article in Test and Measurement describes the testing methodology used by Harman International -- manufacturers of JBL, Infinity (and I believe they own a couple OEM's) -- Their analysis of speaker reviewers demonstrated that all speaker preferences, whether the reviewer was a trained expert, or a new listener, are basically the same -- the "experts" just rate all speakers lower. The article can be found on T&M's webzine:

http://www.reed-electronics.com/tmwo.../CA475937.html

Some other interesting stuff -- for the very large speakers they rent an airplane hanger to do their testing !, they use MLSSA and discuss the difficulties of getting this DOS program to talk with their windows systems.

Oh, the Consumer's Union methodology of speaker testing -- is negatively correlated with listener's preferences.

[Jim85IROC]

That's a very interresting article, although I find it hard to completely agree with. I've seen time and time again when selling audio, where some people will imediately gravitate toward speakers that have a much different sound characteristic than I personally prefer. Most "untrained" listeners seem to go for the speakers with the reversed bell-curve frequency response.

[tiroth]

Even "reverse bell" preference would come down to methodology I would think...even if the initial preference is good it is probably very fatiguing. There was mention of it requiring "20 hours versus 20 minutes" so I wonder if the tests were literally some hours long...that might give unexperienced testers the time to learn the difference between neutral and colored sounds.

It's an interesting thought, because it really shouldn't be an unexpected result. Even though the average person might think Bose sounds great, in all likelihood it is because they have never heard anything better, and then cognitive dissonance sets in if they are an owner. The Harmon test essentially removes both variables; you would expect such a lister to rate Bose highly but a better speaker even higher. This is where, I think, test duration comes in though...the initial "wow" impact of a midbass peak takes time to become tiresome to an untrained listener, whereas an audiophile will probably realize it is going to honk before the first bass riff.

Incidentally, my first thought on reading this was to write it off given that people such as ourselves are such a negligable part of the population that we are within the acceptable error...but I think it should be given serious thought. H-K often comes out with papers that challenge audiophile beliefs, but you can't fault them for doing really detailed, methodical research. Even if it isn't all true there are probably some important things in there.

[Mr Evil]

That's the sort of research I like to see! Well thought out experiments to bring the scientific method to what is probably the most subjective area of audio reproduction.

[Swedish Chef]

Yep, the guys at JBL Pro do have a clue as to what they are doing.
Good article. Like a fresh breeze of air.

[Jim85IROC]

Originally posted by tiroth
Even "reverse bell" preference would come down to methodology I would think...even if the initial preference is good it is probably very fatiguing. There was mention of it requiring "20 hours versus 20 minutes" so I wonder if the tests were literally some hours long...that might give unexperienced testers the time to learn the difference between neutral and colored sounds.

Very good point. One very flawed part of my arguement is that the typical customers I had listened to speakers for a matter of minutes, not hours. And like you said, my reason for not preferring the "reverse bell" is because it typically was quite fatiguing, but that is not something that becomes apparent upon initial listening. Because most speakers are bought after a very brief audition, most mainstream products do seem to follow the reverse bell.