View Single Post
Old 10-18-2008   #19
Lotus
Super Member
 
Join Date: Nov 2007
Location: Texas
Posts: 938
Default Re: I have some issues with this test too

Quote:
Originally Posted by JerryDelColliano View Post
Lotus,

Kenny brings up some good points/questions above.

Here are some of mine...

1. Are these tests being done blind? Any speaker evaluations must be done that way so that the look of the speaker doesn't effect the results. I have done tests at Harman and Paradigm and they are ALWAYS blind. A MartinLogan has a visual advantage in my opinion over say a Gallo and or a Vandersteen as the later are some ugly ass speakers in comparison if you are just looking at thier exteriors.
You can't see the speakers. We give the test subjects dials. Then they wear eye coverings. At that point the screen is removed and the test begins.

Quote:
2. How could you demo each track for 3 minutes. That is SO LONG. Its a demo. Normally, people tune out of a demo after the first verse which is rarely past 1 min in the song.
We don't. That would make it entirely too long.

Quote:
3. Who gave you the speakers?
Either bought or donated.

Quote:
4. What order are the speakers played in?
Random. Including speaker designations being switched without the test subjects knowing.

Quote:
5. With over 2000 demos in your test - who is paying for the hearing tests? Who paid for the speakers?
We did through money gathered from the university, government grants, and donations. We have more than a few big donors.

Quote:
I am not trying to bust balls here. It just seems like there are some serious flaws to the science of this test. I know Harman does something similar at their test facilities here in California but a LOT of the variables are taken out of play and they are blind tests.

There are no flaws to the science, trust me it's run better than the corporations I've worked for. The Professor behind these tests made sure the science was done right. It's a controlled test. It started out testing cables. Everything else was added when the donors showed interest. The cable tests should be done by the end of the year. They're just crunching the numbers now on the last 800 or so test subjects. I don't know if the final numbers are going to match the other numbers, but that should be interesting to read the report when finished. The cable testing hasn't been my gig, I'm with the speakers.

There are some "flaws," with the test in that we can't flat say "this speaker is better than all others." Why? We didn't test all speakers. We didn't test all speakers from the manufacturers we tested. We couldn't afford to. Basically we tried to buy the 2nd to top of the line speaker for different lines.

We should be able to however say that certain types and certain cones, crossovers are "preferred." This data can be sold to speaker manufacturers.
Lotus is offline   Reply With Quote