fudd@fluke.UUCP (Mark Freeman) (12/08/83)
I see three problems with the proposed speaker wire test, and one piece
of missing information.
First, what length are the wires? You will want the wires long (to
increase the "wire effect"), and equal length (to be fair).
Second, you will want to match signal levels AT THE SPEAKERS. Some
wires will have enough series to make the speaker output levels
significantly different -- invalidating the test.
Third, you should throw in a control: two wires of the same type.
Perhaps this could be a separate test -- listen to three cables,
labeled A, B, and C. Two cables are identical, and the third one
shown to have a different "sound" from the others in the previous
tests. The listeners will try to detect the odd wire. This test is
useful, because people may show a preference for a certain letter of
the alphabet.
This is psychological testing, and you have to be very careful to
account for all the variables.
Mark Freeman
John Fluke Mfg.
{ssc-vax,lbl-csam,uw-beaver,microsoft}!fluke!fudd