Killingbeans
Major Contributor
4 different situations: 1Without ricable cable ; 2 Only with power cable ; Power cable and signal ; Power cable, signal and power supply.
Did he find that the power cable made a difference? If so it would defy all logic and common sense. It can only happen if the error of margin is large enough to completely drown the actual data, and it should be enough to disregard the test altogether.
And as Blumlein 88 says, using an SPL meter on a speaker for testing interconnects is not very reliable to begin with.