And it turns out that, for example, in a RIAA corrector, the error of the wire going from the MC head to the transformer will be amplified almost 1000 times!
Does not work that way. If the error is simply frequency response, the relationship between the perfect and imperfect signal never changes. That would be true for most things dependent on the cable with the exception of noise.
And how many different wires are there in the system? Hundreds. And all these errors that occurred in the preliminary cascades will be amplified by hundreds or tens of times and superimposed on the useful signal.
Again, see my last comment.
A microscopic error on the edge of perception multiplied by such a caos will become egregious. But we do not observe such errors.
So there is no polarity, semi conductivity or any other ELECTRICAL assymetry in a wire.
No chaos multiplier. We can observe asymmetry, just not easily at audio frequencies as the variation in the two port transfer function compared to the source/load impedance is not large enough at low frequencies. At high frequencies it is. This is simple fact. No point in ignoring it.
None of which matters to the audibility of cable direction. Unless you purposely went out of your way to create a cable that is directional, you will never hear the change in direction (shielding aside). I could certainly see someone who does not understand how cables (or electronics) work, doing something foolish, like
- convincing themselves they can hear the direction in a single conductor, when what they really did was completely change the interaction of that single conductor with its surroundings
- moving around speaker cables with non-fixed widely spaced conductors (Tellurium).
- You have to be careful how the shield is connected for shielded phono cables, as the connection of the shield can have a big change on capacitive loading.
- With a high capacitance interconnect, the connection of the shield, how the shield is connected in the component, and the source impedance can all come into play. Conductor to conductor capacitance could be 50pF/foot. However, you can have another 50pF/foot from cable to shield, and that shield could be signal ground, capacitively coupled earth ground, etc. The loading difference of a 6ft cable at 20KHz is 13K ohm versus 26K ohm. Some tube equipment has fairly high output impedance. If the connectivity is different in one direction from the other, there could be an audible effect, not just because of noise.
Self proved that there are no audible errors in the signal, even in more complex cases than just a single wire.
Technically I believe he proved he couldn't find any. He also said that wires can't be directional because the signals are AC. He was wrong on AC, so I am not sure how much faith I can put in his other findings.