Cable geometry, dielectric, and material can perhaps make measureable differences in various ways (e.g. various attributes, and then sometimes in measuring certain types of distortion with certain devices), but that's a much different sort of claim from claims that this is
audible. A pretty large amount of measureable phenomena aren't audible. It seems like people overestimate the sensitivity of human ears when trying to develop theories on what sorts of changes make for audible results, to sometimes surprising degree. I've tried a variety of cables including some rather pricey stuff, and even had the fun of trying to measure performance characteristics on a few, and I have ear training from testing various psy models for lossy codecs since the late 90s. About the only situation in which I obtained a particularly audible change was going from a very short (0.15m) to a longer (2m) cable, and I suspect this may have been due to the longer cable inducing oscillation to the point of audibility in the nearly totally unstable output device (a much more serious problem anyhow
). I'm open to the possibility that my test methodology was in some manner flawed, though I would be surprised to find that I managed to overlook audible differences when I can detect rather minor distortions in psy model testing without significant modifications to test methodology.
Understandably, I think, I decided I'd avoid a big investment in cables for my own use. I use basic cable that gets the job done. I've tried pretty diligently to find audible differences, and at that point I'm looking at differences so subtle (if at all existing) that my money is better spent elsewhere. Even the differences I've seen others describe don't strike me as getting near the price/performance ratio of various other tweaks I could bother with, so it doesn't seem like a wise investment for me. However, nice cables do look snappier and, let's face it, aesthetics is a major part of the hobby for many people as well. There are certain situations where design/construction is very important (such as impedance matching for an S/PDIF connection) and so I take that seriously. I may make some snappy cables for the heck of it if I get bored enough and have the time and resources, and then test them of course.
I'm certainly happy to entertain another cable test if someone wants to put up with me
. I don't have much investment in it either way so I'd be happy to verify your findings if I should encounter such results