wjm@whuxj.UUCP (MITCHELL) (01/16/84)
I'm inclined to agree with those who say that ALL audible effects upon audio equipment are caused by parameters that can be measured objectively, GIVEN THE PROPER TEST EQUIPMENT AND CONDITIONS. Therefore, I'm in total agreement with Juilan Hirsch. However, the problem is finding the proper parameters to measure, since Homo Sapiens and his/her audio detection and signal processing system (a.k.a. ears and brain) are notoriously non-linear. In some cases, VERY minute changes in parameter values can make significant changes in the results (on the order of .01 dB in some cases, notably frequency response variations), but in others, gross differences will be ignored by people. Also, there is no simple correlation between the objective result (say a 2.47 dB frequency peak at 5 KHz) and the perceived result - the speaker tends to push the lead singer forward - a favorite technique used by the "West Coast" school of speaker design. It goes without saying that all tests must be double blind, since the bias of the experimenter can affect the results. The problem is that, although one can theoretically quantify all significant audio parameters, this is rather difficult to do in many practical cases. Which parameters are the significant ones?? How sensitive are we to them?? Example, should two (amplifiers, speakers, widgets, etc.) be matched to within 3 dB, 1 dB, .1 dB, or .01 dB ??? to make a valid comparison between them. One place where this has been coming home to haunt us is with CD players. People say CD player A is "better" than player B, yet the specs that have been published in test reports seem quite similar. Does this mean that we have not been measuring the right things? or that our tests are not sufficiently accurate? It would seem that CD players should NOT sound the same since the D/A conversion techniques and the low-pass filters used after the D/A vary between manufacturers. While the digital circuits may be quite similar, the analog portions are not, and there lies the source of the differences. [as always, these are my personal opinions and not those of my employer] Bill Mitchell CSO Whippany, NJ (whuxj!wjm)