rfg@hound.UUCP (R.GRANTGES) (02/27/85)
[] Um, you are certainly right that the wrong way to do an A-B test can bias the results. You would also have been right if you had gone on to assert that few audio salons had the proper equipment to do it really right. However, I have never heard of one that did it the <stupid> way your friendly salon keeper explained was wrong. He was right, it was wrong. Since some speakers are considerably more efficient than others, a really sizeable resistor would have to be inserted in series with the more efficient speaker. Horrors! It gives me goose bumps to think of it. It would probably also have had the effect of making the more efficient speaker seem much louder in the bass than usual. But that does not mean it can't be done. Use two identical amps, for instance, or switch two pads "upstream" from one amp to change the volume level. While I'm on the subject, let me express another pet peeve. You often read that in A-B tests a level mismatch of 0.1 db or 0.01 db or -you name it - has been shown to bias the results. Now I am not one to deny the validity of scientifically demonstrated fact. However, I say, Bull S*it! Explain to me how you balance the levels within that tolerance of two sound sources that are, by definition, of different spectral characteristics. What frequency do you use? Band limited pink noise? What band? How measured? Fooey. Any speakers I have ever compared this way begin to sound quite different when they are reasonably close to the same level. Two speakers sound with different loudness depending on what type of music is being played on them. You set the levels so they sound "equal" and then listen carefully. Maybe you tweak the levels from time to time to see if it makes any difference. Usually it is more clear which speaker has what coloration than which is actually louder. Now, done this way, I doubt it could be shown that they were within 0.xx db of each other on any particular measurement. Let me put that another way: given adjustment as I have described, I <will> measure them as being of equal loudness within any desired tolerance - just let <me> pick the test signal. -- "It's the thought, if any, that counts!" Dick Grantges hound!rfg
newton2@ucbtopaz.CC.Berkeley.ARPA (03/01/85)
Isn't it time for a "net.awards.consistently_sensible" roster? But there probably wouldn't be much point-- ALL the awarads would probably go to Dick Grantges. Deservedly. Doug Maisel @ Land of a Thousand Intemperate Flames