I've tested some cables in the lab at work using high end signal generators and Spectrum analysers. I've also used real world music files and software (
https://www.libinst.com/Audio DiffMaker.htm). The tests were set up so only the cables were changed (I/C's and speaker cables tested independently).
At audio frequencies (20hz to 20Khz and way beyond) there was absolutely no difference whatsoever, either on the spectrum analyser or the software. The only way you could get a difference (I/C cable only) is by moving the cable about during the test, but the differences were less than 0.5dBm.
I used to be on a Hifi forum, but I got bored with people who couldn't see past the price tag of their kit. Amazing what expectation bias does to someone's hearing.