Directional cables - what does that really mean?

Some (most) cables do sound differently depending on which end is connected to which component. It is asserted that the conductor grain orientation is determining the preferential current flow. That might well be, but in most (all) cases the audio signal is AC (electrons going back and forth in the cable), without a DC component to justify a directional flow. Wouldn't that mean that in the 1st order, a phase change should give the same effect as a cable flip?

I'm curious whether there is a different view on this that I have not considered yet.
Post removed 
Post removed 
@Jea48 - when talking about electrons in motion physics calls it current and not wave. The EM wave terminology is usually reserved for radiation. However, we agree - is not "the" electron at the source that runs down the wire to pass on the information.

I'm more curious about the effect of extrusion (that sounds like pulling the wire from the melt, or further purifying it through local melting and re-crystallization to push out insoluble impurities and minimize grain boundaries). How would THAT influence directionality? (I understand that preferential shielding can impact overall noise - that would not affect signal propagation but rather the added noise from external sources).

09-28-15: Jea48
You guys that are using a digital coax cable with a solid core center conductor that connects your CD transport to your DAC, you should be able to hear a difference in SQ when reversing the direction of the digital cable.
Try it!
I suspect the main reason that may occur in many systems is that the supposedly symmetrically designed digital cable is not truly symmetrical at the very high RF frequencies which comprise the risetimes and falltimes of the digital signal.

Minor physical differences between the two ends of the cable in how the connectors and wires are mated and/or soldered together will result, at the tens of megaHertz and higher frequency components which are present in those signals, in differences in VSWR-related reflection effects. Which in turn will depend on the impedance characteristics of whichever of the two components each end of the cable is connected to. Which in turn will result in differences in waveform distortion, or lack thereof, on transitions between the higher voltage and lower voltage states of the signal as received by the DAC. In turn resulting in differences in jitter on the clock the DAC will extract from that signal.

Whether or not there will be audible consequences from all of that will depend on many system-dependent variables, including the exact output impedance of the transport within whatever +/- tolerance it is designed to, the exact input impedance of the DAC within its +/- tolerance, the jitter rejection capabilities of the DAC, the characteristics of the cable (including its length and its propagation velocity, which will affect the timing of how reflections and re-reflections from both ends of the cable are seen by the DAC), the susceptibility of both components to ground loop-related noise issues, etc.

Best regards,
-- Al
Cbozdog - Mitch2 indicated above that the effect of grain boundaries phenomenon is perhaps "minimal"

But something that has far more impact on SQ than grain boundaries is the cable architecture itself.

Here's a summary of what I have tried and observed over the last couple of years.

I started with Van den Hul d102 MK III using the floating shield design - that proved to be quieter and more detailed than the other cables I had been using.

The problem with more "conventional designs" that utilize two conductors in parallel is that they suffer "noise" that is induced from the signal conductor into the neutral conductor. This flows back into the connected components and impacts their performance. Sure, manufacturers put a slight twist in the conductors which is supposed to minimize induce noise, but it is not really enough.

I then moved to Stager Silver Solids - which has a very tight twisted pair (no shield) architecture - which were much more detailed and the tight twisting combats RFI/EMI and induction from the signal to the neutral, resulting in a blacker background and quieter operation.

Then I tried a braided architecture, where the two conductors were braided with thick single strand nylon. This separated the two conductors and increased the angle at which they crossed, reducing RFI/EMI and induced noise. These were significantly better than the Stagers across the board and I actually used the conductors from the stagers to construct them - so the materials used were identical.

I then started playing with my own Spiral designs where the neutral is wound around the signal conductor
- Winding the neutral around the signal effectively places the neutral conductor at close to right angles to the signal conductor which minimizes induced noise.

I have since tried several variants of this design, with the latest working extremely well.
- Basically, the signal is a single strand of solid silver wire
- The neutral is four twisted pairs of CAT6 wound around the signal.
- One wire of each twisted pair is cut short and remains unconnected at one end and basically acts like a floating shield and interferes with induced EMI/RFI.

The signal and neutral conductors are different lengths and different materials - but it does not matter, because when you take a look at the roles each conductor actually performs from the perspective of an actual circuit diagram you will see that the signal conductor is the only one that carries the "music" - the neutral conductor actually only maintains a connection between the neutral sides (i.e. ZERO volts) of the connected components. OK - it does also completes the circuit, but it is very important that the neutral side is kept as close as possible to zero volts for each component to perform to it's optimum.

Friends have also tried the spiral design and reported significant improvements in SQ.

Take a look at these links for a more complete explanation...

A couple of companies that I know offer cables designed along similar lines are:
- Anticables (for IC's)
- KLE Innovations (for IC's and speaker cables)

I have also applied the same spiral design to power cables and the effect was outstanding with a vast improvement in dynamic performance, clarity, image size and performer placement.

I personally agree with Mitch2, so I have stopped concerning myself with crystal boundaries issues - I can't do anything about them anyway - except keep both cables connected in the same direction :-)

For me - selecting the right cable architecture has provided huge improvements in SQ and has elevated my systems performance way beyond its modest price point.