I sort of can't believe I'm writing yet another cable post.
If a cable's task is to deliver a signal from a source to a destination without changing it, then most cables should not, and cannot have an audible impact. However, in the real world, cables all have a set of electrical properties that are all degenerative to the signal being carried. In typical cable applications those degenerative effects are far, far too small to be audible (many occurring way above the audible frequency range), and may be quite difficult to measure. Other electrical properties don't have an impact until they make up a significant part of the total electrical circuit. This can happen with speaker cables, but pretty much never with interconnects.
The speaker cable scenario is that if a speaker presents a frequency variable load (i.e. not 8 ohms at all frequencies, rather peaking at 18 ohms, dipping to 3), AND the cable presents a significant resistance or complex impedance, then the result will be a variance in frequency response of the speaker. This happens if the cable gage is too small, or the cable is too long, or both. The effect can be audible and measurable. Again, this only happens when the cable's properties become a significant factor in the total circuit, and that only happens in the consumer audio world with amps and speakers. A speaker's impedance is low. If a cable's impedance is 10% or greater than the value of that load, you have a frequency-variable voltage divider.
Every other interconnect uses a voltage transmission system, where virtually zero power is being passed through the cable, being driven from a moderately low source impedance, and very lightly loaded at the destination. And in consumer audio cables are generally quite short. Extremely long cable runs would obviously compound any negative electrical properties, and would at some point become audible. For example, a 2 mile telephone wire has very significant high frequency loss, often approaching 12-15dB at 15kHz. So, don't run 2 miles of phone wire in your system.
The question is: If cable electrical properties are always degenerative, how can a cable ever improve anything?
The answer is: If the total system has a particular characteristic, such as a rising high end, that is not favorable, AND a particular cable's properties counteract that characteristic, then the degenerative effect is perceived as positive. These combinations are quite rare, though, and mostly, cables don't change a thing audibly. But since it's impossible for most consumers to do any sort of controlled ABX test on cables of any kind, the "tests" are sighted, and highly biased. As a result, nobody hears a negative effect with more expensive cable, often a positive effect because the cable looks and feels "better", and costs more.
The other fact to keep in mind is that some exotic cables have more than cable built into them. They have impedance-modifying components hidden in the connectors, or in the "mystery blob" in the cable. That's cheating, in so many ways.
If the system requires some form of performance modification, there are far better, more controlled ways to get there. The random cable combination stands little chance of changing anything, and even less of actually being beneficial.
Just to keep it real, I'll say that some really cheap interconnects actually are, well, a bit too cheap. They may fail from minimal use, some connectors are pretty terrible, and some interconnects aren't even made with shielded cable. Yeah, they still work fine most of the time, but if you need shielding, you won't have it.
Might as well hit one are where cable can easily have an impact: the cable from a tone arm to a phono preamp. Moving magnet cartridges "expect" a 47K ohm resistive load, but also expect a certain amount of capacitance, made up of the input capacitance of the preamp AND the capacitance of the cable, which may not be insignificant. Changes in capacitive load can significantly alter frequency response. But that's a whole deep subject, the short answer is, you have to fully test the system anyway to get it right.
It's still not a bad idea to use good cables, but good doesn't have to be expensive. And with very few exceptions, no cable will change the signal it carries. The few that do can't be expected to offer a positive change in any but the most rare and specific cases.
