How essential is shielding?

No, I don't think any such generalizations can be made, Bryon. Many shielded interconnects having low or very low capacitance are available from Cardas, Nordost, and Blue Jeans, among others. Unshielded interconnects with similarly low capacitance are available from Kimber and Nordost, among others.

That's what I figured, Al. Otherwise life would be too simple.

How to know if a cable is shielded.
Well, usually you can just LOOK at a picture of the cable. Most interconnects that are shielded are coaxial, a smooth round cross section. Most non-shieded are just wires, usually woven or twisted together.

Elizabeth - Although this is often true, like with the Kimber PBJ you mentioned, it's worth noting that some unshielded interconnects still have sleeving, and so don't have the appearance of bare wires. Morrow cables are an example.

All of my power cables are fully shielded with shield grounded. That is absolutely necessary for the amps I am using. All other wires are completely bereft of any insulation save for 2 ml.

Muralman - Having some cables shielded but not others seems to me to be a good compromise, since, for any 2 cables that are close enough to potentially contaminate each other, it seems like only one of them would need to be shielded to prevent it. My system has a similar arrangement, but I arrived at it more or less by accident.

Bryon

Bryon, as you know, the Speltz is a tightly wound small gauge wire with a no-see-um negative wire running through. Being as the two carriers are at an almost right angle to each other, I think they would be more influenced by external electromagnetic forces rather than self contaminating. What do you think?

My SCs are purposely run separate from each other.

...the Speltz is a tightly wound small gauge wire with a no-see-um negative wire running through. Being as the two carriers are at an almost right angle to each other, I think they would be more influenced by external electromagnetic forces rather than self contaminating.

Hi Muralman - I wasn't trying to say anything about the design of the Anticable I/Cs, or their vulnerability to internal/external noise. Sorry if that was unclear.

I was trying to say that, like you, I have some shielded cables (digital interconnects, power cables) and some unshielded cables (analog interconnects, speaker cables), and that this arrangement, at least in my case, results in a system that is more or less dead quiet. So I was trying to suggest that shielding some cables, but not others, might be an effective way to reduce inter-cable contamination (e.g., power cables to interconnects, or vice versa), without having to shield everything.

Bryon

Capacitance of many typical Belden cables run in order of 25pF per foot while the best interconnects go as low as 3.5pF/ft thanks to use of expensive dielectrics (foam Teflon) and special geometries (oversized tubes). Cable with no shield eliminates hot-to-shield capacitance but hot to return wire capacitance exists and most likely is increased because unshielded wires tend to be twisted pairs. Twisting two wires exposes them to EMI evenly, working effectively as a shield (both for radiating and receiving) but unfortunately increases wire to wire capacitance.

Shielding itself is a very complicated business. For instance, shield made of non magnetic material does not protect against EMI (it cannot) but induced high frequency interference travels on the outside of the cable - a shield, because of skin effect. Add multiple shields + shield's inductance + twisting + additional returns and you'll get something nobody can understand. If it sounds good in your system go for it and don't worry about the science. Noise pick-up is system, and not the cable, dependent - all components participate.

It is worth to mention that EMI pickup of lower frequencies such as approx 500kHz generated by many class D amps can be picked-up as direct connection thru capacitance (important to keep wires apart or right angle) and not thru the electromagnetic pickup. It is because of lenght of the wires that would have to be hundreds of feet long to be any antenna for transmitting or receiving. Antenna works quite well at 1/4 wavelength but drops rapidly below that becoming practically ineffective below 1/10 of wavelength.

As for balanced being remedy - it works but problem starts at high frequencies where noise gets thru transformer capacitance (if used) or instrumentation amplifier (if used instead of transformer) rectification effect (uneven slew rates going up and down cause extraction of the modulation of the signal)

I have found it can make a difference in noise levels in some cases and not in others.

You have to experiment to determine.

Orientation and position relative to other wires with current passing through or powered devices with transformers that emit EM fields matters a lot.