Sorry when I said Standard L/R/G I meant SE (whatever gets terminated into a TRS connector). Its certainly not the standard (if I were to judge most pro audio equipment, the standard would be xlr connectors -I assume balanced config).
Channels are kept separate in all audio equipment because cross talk is even worse in the internal connection of components. Everything I mentioned in my previous post results in cancelling or reducing induced effects (cross talk) but when you're dealing with signals in the process, nothing cancels out so you have to keep them separate, you're absolutely right about that. But after wards when your transmitting the signal, you can use differential transfer (like telephone lines, data lines) to reduce cross talk and noise, by using twisted cable pairs). And I think this is what we're discussing in this thread. If I'm wrong about what topic I'm discussing then know that all the info I've written down is for headphone cables not for interconnects and or internal construction of amps/dacs etc...
As for balanced cables, I was referring to balanced cables constructed from the quad litz braid (obv no shield, just hot/cold pins with a floating pin left over). In this case, there is no cross talk from using a 4 litz braid because once again all the noise gets canceled out due to the differental mode of operation. The difference between balanced and SE (or whatever you want to refer it to) is that the means of transferring the audio signal in a balanced set up results in the cleanest possible signal, its like your headphones are connected right to the output of your amp and the wire doesn't exist(a bit of an exaggeration but I'm just trying to say balanced is about as clean as you can get while having a long cable). SE on the other hand reduces noise through the twisted cable method (at the very last when dealing with the litz 4 braid). This uses a form of differential transfer but it is not as good as a balanced set up. A balanced set up inherently uses differential means of transferring the signal so at the end so you get just the signal. SE (with twisted pairs) is just a method to try and reduce the noise/crosstalk created from long transfers.
I think that the previous solution somewhere near the middle of this thread (p5?) someone used a twisted pair for the r+g and a second twisted pair for the l+g. All I'm saying is that you get the same result when you do a quad litz braid (see the sample braid in my previous post). The difference is that you're sandwiching another pair of cables in between. But because of the geometry, the magnetic fields cancel each other out or at the very least they are greatly reduced.
Braiding cables for mono cable is fine, my point is a dual channel is fine aswell. In fact I would go so far as to say, that if it were possible to create a geometric design similar to the litz quad braid for multiple channels (greater than 2) then I would go so far as to say that even more channels would be fine. But to be honest, I don't think its possible because other braiding techniques don't utilize a twisted pair geometry in directly opposed axis like the quad litz braid.
No, no braided... too much work, I learned the hard way. You guys ever try flat braiding with jena labs wire it looks gorgeous but...pita I've recabled my headphones for the sake of doing it myself. I've come to learn why so many companies charge high prices for their services. Grant it the cost of material may not be that high, but the amount of time and effort required to recable a pair of headphones adds 2+ hours to any job. I'll be doing a recable on my akg701's and I'm still undecided what to do with them, I think I'll just stick with a quad litz vampire wire braid.
On a side note, if you're interested in making a cable that is left/right channel separate, shielded etc... I think I know of a way to do it. But I think it would more than double the weight of a cable, double its thickness and well... basically turn it into an interconnect rather than a headphone cable.
Final points I want to emphasize:
1) I still question the source of OP's information (particularly that website) If someone can point me to what waxing does (I assumed it was always to prevent oxidation, Dr. Xin used it in one of his tutorials and I figured that was the prupose) and what "low level detail resolution" means then I'll certainly correct/retract any of my comments regarding that site and the information available from the site. Furthermore, the gold issue.
2) As you mention Les, we are making compromises... its the engineering principle. So long as we don't all live near a military base or a radio station, we don't all have to run balanced set ups in shielded bunkers to reduce noise. And as far as I'm concerned, the amount of cross talk you find in a headphone cable is drastically minimized by using twisted pairs and quad braiding. And yes I could go insane using shielded cables for left, right and end up with a garden hose for cables, but its just not feasable. So I make the compromise and accept some noise in my system. Or if you want both a thin cable and absolutely no noise, I would drop an extra 2 grand to get a balanced source, balanced amp, and recable all my headphones to a balanced set up. In terms of RCA connectors though where the signal is still being "processed" by your system, I would err on the side of separate channels, plus I really can't imagine interwoven interconnects, its just...well .. I dunno I find no benefit in having them versus separated lines. As far as I know the cost of building an interwoven 2 channel interconnect might be higher or at least take more time to construct.
3). with the exception of high end dennons, I think every cable I've cut into has some form of arrangement where the signal wires are encased in a metal shield but not from one another.
