Mprime: I can't give you any specifics about interconnects. As i've stated before, "interconnects are a mystery" as far as component to component interphases go. That's because the spec's from component to component vary so widely.
What i will say is that i hope to start doing some testing in this area sometime shortly. I've finally got all the equipment that i need to really look at this subject in-depth. I do have some ideas as to why some designs are better than others, but at this point in time, they are stricly theories. Obviously, the use of high quality conductors, materials with low DA ( Dielectric absorption ), geometries that are lower in capacitance, etc... I could add a few more criteria that i personally think are important, but i'll save it for later : )
Other than that, much of what we hear has to do with the stability of the circuit loading into the cable. The more stable the circuit, the less variance from cable to cable and mating component. One thing though. Common electronics theory states that you don't want the output impedance of a device feeding into another device that is lower in impedance. This "loads down" the circuit due to increased current draw and can play games with bandwidth, tonality, transient response, etc...
With that in mind, a device with a very low output impedance is much more likely to be more "versatile" / open to variations amongst cables than other devices with higher output impedances. After all, most interconnects will have a nominal impedance that is 50 ohms or higher. I picked 50 ohms as that is a common impedance used for standard coaxial designs. Obviously, a source component or preamp ( tubed most likely ) with an output impedance of 200 - 600 ohms can be quite finicky when you introduce a cable with a low nominal impedance as part of the load that it sees. The end result of such a situation is typically high end roll-off ( lack of "air" and "shimmering highs"), congestion in the midrange and muddy, ill-defined low frequencies. This is what happens when there's not enough current to get the job done.
The opposite can be said of low impedances trying to load into very high impedances i.e. a lack of bass and warmth, loss of dynamics, etc.. but this situation is typically not nearly as degrading as the other way around. That's because current flow is choked and you can't load up as much voltage as might be needed.
Maximum power transfer and optimum transient response occur when all the impedances i.e. output from the source, cables carrying the signals and input to the loads are all equal. There's only one manufacturer that i know of that builds a system like this though and that gear probably isn't real compatible with other brands of gear.
Flex: Some manufacturers of "esoteric" wire have told me that you can have cables built to identical spec's from different manufacturers and they will all sound different. This is true even though they might measure identically. This tells us that there is obviously something more going on inside the conductors / dielectric than just impedances, etc... John Curl posted some info about the "non-solid" parts of metal conductors varying with current flow i.e. conductors actually change shape ( and probably conductivity too ) as various levels of energy are fed through them.
As far as "backwards" cabling goes, that would be speaker cables that are highly inductive, interconnects that are highly capacitive, cables that use very low grade conductors, cables using very lossy dielectrics or dielectrics that don't protect the conductors from pitting / oxidation / corrosion, poor geometries that are highly susceptable to RFI / EMI, cables that have nominal impedances that aren't good matches for the component interphase that they are involved with, etc...
I will only comment on one of your listed. That is, mechanically damped cabling. I've had some cabling here that is PHENOMENALLY microphonic i.e. it is VERY susceptible to both air-borne and floor-borne vibrations. Most of this is due to the type of conductors and geometry used, making the cable very rigid. If the vibrations are strong enough and the cable is coupled to the devices quite sturdily, it's possible to resonate the chassis of the components that are connected to this cable. Given that some designs, such as precision devices like turntables, CD players / transports and even tube gear can be affected by such things, i wouldn't want ANY cabling like that in my system. While not as common, even SS line level and amplification gear can be microphonic / highly sensitive to vibrations. Much of this could be related to brittle solder connections and / or defective capacitors, especially on older gear exposed to high heat levels.
Roofus: Great suggestions. I agree whole-heartedly with your recommendations and the order suggested.
Tommywall: While i thought i was going to hate that website, i actually liked it. Their testing showed the Goertz cables as having the
widest bandwidth with the
lowest inductance and
least skin effect ( yellow charts at top of page ). These are all points / design advantages that i've been preaching about for years and i'm glad to see that a "third party" actually verified this. Having said that, i would have said "neutral third party", but i do agree that these guys have an agenda. Then again, i'm no "fan" of the high end cable market as a general rule either : )
If we look at the
Goertz MI-2's without the Zobel attached in figure 5 ( no far end termination ), you can see that the response at the amplifier and at the speaker are perfectly in phase out to appr 1.8 MHz. Obviously, this is WAY beyond our audible range. Above this point, there are phase shifts that take place and at appr 10 MHz, there is a huge resonance / phase shift at the end of the speaker cable. As such, using this cable without the Zobel creates a high frequency oscillation. This is what i've warned people about when using Goertz or other VERY high capacitance cables ( like Chris VH's 27 pair CAT 5 design or Polk "Cobra Cables" ). I'm sure that many of you will recall me stating that these cables SHOULD NOT be used without Zobels. Now you know why and can see it on the graph. It should be noted that this is SOOO high in frequency that most amps would never notice though, due to their limited bandwidth. Some of you will also recall that i'm a fan of "fast" amps that are "wide bandwidth", so that's why i'm overly cautious about this when recommending this cable.
If we look at this same cable with the proper terminations at the speaker end i.e. Zobel network with the right values used as in figure 6, the output of the amp is identical / remains in-phase with the signal at the speaker. As such, you can't get much better than this. In fact, the output of the amp and the signal at the speaker remain in-phase out to FOURTY MEGAHERTZ!!! Whether or not any of this is audible is a moot point. The fact of the matter is, no other cable comes close to the wide-bandwidth & linearity that this cable delivers while retaining low series resistance.
If we look at this same cable using the factory supplied Zobel's or "impedance compensation networks" that Goertz includes with their cables in figure 7, we can see that the signal at the amp and the signal at the speaker remain in-phase with each other out to appr 2 - 2.5 MHz or so. Between appr 2.5 MHz and 15 MHz, the signal at the end of the speaker cable is slightly out of phase with the signal at the amp. At appr 15 MHz, the output of the speaker cable and the signal at the amp begin to track relatively closely again out to appr 40 MHz or so.
Needless to say, most all of their "complaints" about this cable on Audioholics are gibberish / not applicable to response within or anywhere near the audio passband. The only valid comment that is correct and "negative" about this cable is that the high capacitance per foot can cause some amps to get "squirrely". The use of a Zobel pretty much takes care of that though. The point about Goertz nominal rating of 2.5 ohms for this cable and Audioholic's figure of 8 ohms ( which is still WAY, WAY better than most other cables ) is also addressed in the text. That is, they said "Using classic SQRT(L/C) (which isnt exactly correct, but this topic goes beyond the scope of this article) we calculate about 8 ohms while Goertz calculated about 2.51 ohms".
As one can see, Audioholics aka the "cable naysayers" basically confirmed what i've been telling you folks for several years now. That is, Goertz speaker cables are the widest bandwidth / most linear speaker cable on the market. That's why i said that if you didn't like how your system sounded with Goertz installed ( WITH the Zobel's of course ), you better start looking at your system, not the cable itself. Just remember, they tested the "old" Goertz cable using the original dielectric, not the newer design that is using a higher grade dielectric ( Teflon ).
Rooze: It is my opinion that price and cable quality have little correlation. While it is true that higher grade materials ( purer conductors, better dielectrics, higher grade / nicer to work with connectors, etc... ) do cost more, some cables that are less expensive actually use better materials than "sky's the limit" type cabling.
As far as interconnect selection goes, see my response above. Yes, it is pretty much "hit or miss" from component to component and system to system. As i've stated before though, choosing cables that take advantage of good materials and decent electrical designs puts you miles closer than if starting off with cabling using random designs and materials.
As to cable reviews being worth anything more than "entertainment", don't make me laugh.
Tommywall: IAR ( International Audio Review ) aka J. Peter Moncrieff did testing of speaker cables many moons ago. Moncrieff provided some rudimentary charts with subjective listening results. His observations were that cables using a very high capacitance / low inductance design both measured AND sounded the best. Like i've said before, spec's can reveal sonics IF the spec's / test results were properly obtained.
Other than that, Nelson Pass also conducted similar tests and the article that he wrote about the subject can be found in the Pass Labs website under the name of
"Speaker cables: Science or Snake Oil". Nelson commented that the radical speaker loads that send less than stable amps into oscillation can be "tamed" by playing with speaker cables. In this respect, the cables that worked best were actually those that measured worst. That's because their poor electrical performance masked the problems due to high levels of loss / specific types of reactance between the cable and speaker nullifying each other. In effect, Nelson's article verified that speaker cables could be used as "band-aids" for improperly matched equipment almost 25 years ago. Once again, i knew that there was a reason that i love this guy and the products that he designs. That is, he's years ahead of most of his competition : ) Sean
>
