Many years ago when all this cable stuff came about, we did testing. Real engineers, one had an earlier career actually designing wire. Objective as much as we could afford, subjective where we could. Scope, Digi-Bridge, AF generator etc.
We used two systems. One was CJ tube P5, MV50 into Quads. Records of course. My system was a C&M amp , Hafler preamp, into Celeston 44's, modified with SEAS tweeters.
Our conclusion that I have held fast until now is there are two kinds of cables. Bad ones and good ones. I still hold that "truth" for digital cables. Almost for RCAs.
For speaker cables, our testing convinced us that 14 or 12 AWG zip cord was darn near perfect. None of the new boutique fire hose stuff was any better. We also tried several fads at the time like ribbon cables, bell wire ( solid core 16 gauge twisted) etc. One was a ribbon made of twisted pairs. We found some could really upset my amp. ( marginal stability it turns out)
For RCA cables, as XLR was almost non-existent for consumer audio, found plain old Belden "Brilliance" stranded 75 Ohm or RG59 to be about perfect. To the credit of some long gone company, you used to be able to get some red colored cable made to length. I remember the display stands on the register counters. It was again, just 75 Ohm stranded coax. We did not find any benefit to exotic shielding as that is only applicable to RF and most noise is 60 Hz which copper and foil shield do nothing for. It can be a bit more difficult for line level as there is possibly a path back into the feedback so one would be "correcting" external noise and not the amplifier, which would be just amplifying the noise. In power amps, the feedback is before the output filter so it is protected. I guess I do not see the preference for RG-6 for interconnects. Seems like the thinner can have the same C with lower L and the R is not important. From my UHF antenna, 100' to my TV, transmitters 47 miles away behind two hills, yea it matters! Over 3 dB advantage. Granted in the late 70's the level of RF pollution was not what it is today.
Since then, I had some 4TC Kimber to solve an illegal CB radio transmitter problem. It worked for that, but I did not hear any other difference. Overt time I tried several sets of expensive RCA's at boutique shop insistence and did hear differences. ALL of them worse. Snake oil plain as that. One was not bad. Tara I think.
So I have stuck by my zip cord and "well made" RCA cables. I don't have any hum problems, ground loops or other real issues to fix. I have watched with amusement how a braided sleeve can triple the price of a cable and how magic properties can add information that does not exist on the source. It is only insulting when I see a salesman selling $300 cables to someone buying $200 speakers. Yes, I have seen it Best Buy! If you are filthy rich and can toss $30,000 into a cable, that's your problem but I would advocate you can find a better use for it in our society.
Forward 30 years. A few, and I do mean few, serious engineers have looked at cables from the real measurable parameters. The gentleman from Belden, now Iconoclast, the fine folks at Blue Jean, and I am sure a few others. Mogami, Cardas etc. Some has been learned.
The more believable papers admit they are looking at the cable, not the system. I do not suggest this is an oversight, but practical as no two amps source or speakers are the same. So we have cable measurements, but not measurements in systems showing the effect if any.
On speakers, it matters a lot on the speaker design/load. I have always built low Q sealed systems as I think they sound better. They also have a lower broader impedance peak, less phase shift and lower group delay. As I have also designed and built amplifiers, I prefer higher resistance drivers for class AB amplifiers. I have not paid too much attention to impedance at high frequencies, but experiences with class D have identified that as a really serious problem. I am revisiting my last build to see is I can do a little more in the phase and tweeter impedance rise.
I still can't get carried away with fancy copper. What does the signal see inside my speaker? A big honking roll of generic copper coil in series with the woofer that may even have an aluminum VC! On the tweeter side, big aluminum foil caps. So don't tell me crystal orientation in 10 feet of cable, though measurable, makes a sonic difference. If you are active crossover, maybe I'll buy it, but a hard sell.
But geometry and dielectric can. I am not going as far as suggesting $1000 as perfect as can be measured cables are audible or not, let alone $30,000 but come in a Gucci case cables, but maybe, now we have electronics 100 times cleaner, source 20 dB higher resolution, and tweeters getting better every day, actual audible differences are possible. Finding the real ones through the snake oil is still almost impossible. We so far only have subjective viewpoints which seem to track the price or back cover advertisement. It could be different cable parameters are desirable depending if transformer, AB or D amplifiers are used. I may go as far as borrowing some Kimber base again to test.
I am not as concerned with my RCAs. One of the "exotic" DIY cables I tired was just a 30 ga ( wire wrap wire, silver plated copper, PTFE insulation) twisted pair. It was short connecting CD to amp. I heard no difference compared to my short 75 Ohm coax. Flexible but fragile. I will say, this was back in the internal DAC CD player days before I bought my first Wolfson. Source was not what it is today. I did use this wire in my tone-arm to advantage. My RIAA amp was in the base of my modified Thorens, so not trying to send MM signals to the preamp over long RCA coax. Only amplified line level.
Reading the few actual engineering papers and videos I can find, It did get me wondering. Why 75 Ohm cable? Just because it is made by the parsec? 75 Ohm at a Mhz? I don't care. What happened to the low C 100 or 110 Ohm cables we used for instrumentation in the lab? R is not a factor as there is almost no current and we are feeding 22 to 47K typically. Do I care if we have .25dB drop over 1000'? Nope. Ever cut open a Scope probe lead? Tiny steel conductor, rippled inside a tube so majority air dielectric, about a 75% braid. I bet it would make a great interconnect. Belden 2221 maybe? Of course scope probes have an LCR adjustable network so one can compensate the cable for the input reactance. A proper systems approach.
The subject of magic power cords comes up. I still contend that a safe ( welded as some Chinese cords were found to only be twisted to the prong and over molded!) plugs on about any old wire is as good as it gets. Again, consider what is on either side of the cable. But, if one is forced to have cables close to electronics and analog cables, some effort to reduce 50/60Hz emission is desirable. As shielding is useless, that leaves geometry. As far as RF pickup, I contend that is the responsibility of the power supply. Don't blame the cable for a bad amp design. I have long twisted ( better than one per inch) 2-conductor zip cord for isolated cords. Works and is free. How about the standard round IEC cables? Twists are as few as 6 inches and the safety ground is in the twist. Well, how about stripping the outer jacket to the IEC plug. Separate out the green and cut off the paper filler. Twist the power leads at 1 or 3/4 inch interval. Pull it into a sheath for protection and add whatever plug you need on the wall end. Now you have a low emission power cord that just by chance, won't pick up much RF either. Might cost a $2 cord and a $5 plug, with $2 braid. I suggest leaving the amp end in place as the screw on ends are like $11.
HDMI is the bigger mess. I only buy BJ cables as they ae tested to spec. Threw away a lot of total crap cables. I had troubles with POD, noise, ARC and CEC. USB testing also sent me to Belkin and Belden cables. Free ones being crap and easy to measure on the scope the spikes they inject to the ground and then audio. Can't say I heard the bad USB cables, but measurable and cheap to do correctly so why not be safe. Drug out my scope and I have not been able to find any RF noise on my Ethernet. Plain old cheap CAT-5.
Now figure me this: All these high end amps and such in aluminum boxes, then we complain about hum. What about steel? Mu Metal? If I was building the highest performance I could, it would be a steel box for LF, copper plated on the inside for RF. The PS would be in it's own internal steel box.
