A partial copy of a "discourse" printed in the early 1980s. My comments added in square brackets.
QUOTE: I find that a twisted pair of wires will have consistently better imaging and transparency than a cable made from similar materials assembled into a coaxial configuration. This concept is well accepted by most people experimenting with wire in the San Francisco Bay Area. No one has proposed a rational and documentable explanation for this effect. [perhaps a double blind test is required?] I surmise that the interaction between the electrical and magnetic fields which surround a conductor is different in a twisted pair configuration than in a coaxial configuration. This difference in symmetry affects the perceived imaging differences, indicating that the twisted pair passes phase information more accurately. In my recent experience I have found that because recording equipment uses a balanced configuration, the deleterious effects of even a few feet of coaxial (un-balanced) but otherwise acceptable cable are audible. [Dew tell - a few feet? How many nanoseconds of coax?] Also, most professional equipment uses a different interconnection format than home equipment- Basically, the home equipment standard is a medium output impedance feeding into a high input impedance in an unbalanced configuration (signal "low" is the same as ground). On the other hand, most professional equipment uses balanced lines (signal "low" is not the same as ground, or in other words, both signal "high" and "low" are balanced to ground.) The impedances are lower and in the US generally are the same for the output as well as the input (600ohms). Europe has a more logical standard of a very low output impedance feeding a medium input impedance-Regardless of input and output impedances, professional applications use a balanced configuration. Even in instances where I might be connecting unbalanced equipment, I have followed (mostly for consistency) a balanced configuration. Little did I realize then that this was as significant as it seems today.
1 will evaluate some Belden cables as typical examples of commercially available wiring using these three criteria. This isn?t to suggest that Belden?s versions of these cables are either recommended or not recommended, but they are well known and I have a Belden catalog right here to thumb through. Under "Hi-Fi and Stereo Cables" the catalog lists number 8421 as "Hi-Fi Connecting Cable." Made with 25AWC inner conductor which is built up with three strands tinned copper and four strands tinned copper covered steel, it uses cellular polyethylene insulation, is fabricated in a "coaxial" configuration, and uses a tinned copper spiral wrapped shield. Thus, it meets only one of the three suggested criteria for acceptable audio interconnect cable (insulator). Under the same heading they also list number 9456 as a "Low Capacitance Stereo Connecting Cable " Other than a slightly lower capacitance per foot (between conductors) specification (12pF/foot vs 18pF/foot for 8421) it is similar to 8421. The inner conductor is 30AWC tinned copper covered steel rather than the composite copper/copperweld inner conductor in 8421. It too is acceptable only for its insulator. Still worse, both use a copperweld inner conductor - and you can hear the steel. [yea, sure, and there is this bridge in Brooklyn..] Listed under "Computer Cables" is number.9721 [that is a control cable, 8 unshielded pairs; try 9271] which 1 have used on several occasions for long permanent microphone lines. It has two 25AWG stranded tinned copper conductors in a twisted pair configuration, uses polyethylene insulation surrounded by an overall "Beldfoil" (aluminized mylar) shield. This cable is acceptable both for insulator and configuration. END QUOTE