The question I'd like to ask the writer of the Stereophile article is why was there no hypothesis given to try explaining what was happening. Isn't that what most scientists do? Make an observation, then theorize why it happened, and then test the theory to see if it can be generalized and apply to other operating conditions and make the theory useful.
Look at the last Figs 27 & 28 (the last 2 figures). When the test speaker was replaced an 8 ohm resistor, the ringing magnitudes were much diminished and I can hardly tell any differences between the "best" and the "worst" cables. Why wasn't any hypothesis or explanations (or even discussions) given?
I constructed a highly simplified and idealized model with OpenModelica. The cable was modeled with a 0.1 ohm resistance, 7 uH inductance, and 1000 pF capacitance. The loudspeaker load was simulated using the model given in this link (but with R1 set to 6 ohm):
http://www.aikenamps.com/index.php/designing-a-reactive-speaker-load-emulator
The 1 kHz results are attached. It showed that the type of behaviors in the Stereophile test can be easily simulated. The differences/deviations are not unexpected since the Stereophile graphs were generated using a real (and non-ideal) amplifier and a real speaker, and I've made no attempt to tweak the simulations parameters to match.
So what's going on in the Stereophile test? It is simply that when the tone bursts stopped, the speaker could not "stop" immediately. The combination of the speaker driver inertia and stored energy in the capacitors and inductors generated a back EMF. And what were shown in the traces were the voltage drop across the speaker cable due to the back EMF (when the back EMF is shunted through the output impedance of the amplifier).
So what effects will this have on the sound radiated by the speaker? The real test is to use either measure the sound pressure with an instrument mic, or use a laser vibrometer to measure the displace of the speaker diaphragm. If you do that, you will find out that the effects of the cables are either not measurable or totally negligible. They will be totally overwhelmed by the mechanical responses of the drivers.
This are all elementary stuff.
View attachment 72060 View attachment 72061 View attachment 72062