SilverSmith Audio Cables Interview

Jeffrey Smith from SilverSmith Audio was nice enough to engage in an informative interview about his cable philosophies and self proclaimed sciences behind them. While his ideas about cables and signal propagation are unconventional, they do sell a very interesting story.

What transpires below is an informal transcript between Jeffrey Smith, Gene DellaSala and John Escallier. All content is provided here with permission. Read on to learn more...

Audioholics: Dear Silversmith Audio. I noticed some interesting claims on your products but also noticed no published specifications. Do you have published specs of these cables (R,L,C)? If so, how did you measure them? What distance did you separate the (+) and (-) conductors when doing this? Thanks.

Silversmith Audio
Silversmith Audio does not publish the RLC specs of the cables because the relationship between the conductors can change. Consequently, the LC parameters are not constant. Also, RLC parameters, within the realm of most cable designs, have very little or nothing to do with why/how cables affect sound.

Audioholics: That's interesting. So what you are telling me is you have no control over RLC of your cable and it's a function of position by the end user. What cable metrics are responsible for the sonic signature of the cable? How do you measure this to improve the designs? How do you know when you reach nirvana?

Silversmith Audio
You sound surprised! That's interesting considering you and others on your site have so diligently shown that RLC parameters can not be expected to influence cable sound for all but the most extreme designs. Those who argue that cables can not possibly affect sound most frequently use the same arguments you have presented.

Audioholics: We never argued cables cannot sound different. But we will argue that only badly designed cables DO sound different. Such is the case with many exotic cable vendor designs. Perhaps yours are an exception, but the fact that you don't rely on measurements to govern your cable designs is frightening.

Silversmith Audio
Unfortunately, they do not continue deeper in to the true physics of electromagnetics and/or properly apply the concepts to discover what really happens in electrical flow. They use the simplified model of electricity, which is the water in a pipe analogy. In this model, electrons, like water, flows up the positive leg, and back down the negative. Engineers use this model because its gross assumptions and generalizations of electricity make low frequency, typically 60hz, circuit design and analysis very easy.

Audioholics: Can we stick to a discussion on cables, I'm an Engineer not a Plumber :-) Hmm so you seem to be doing some sophisticated analysis. What programs/tools are you using to govern your designs? I bet these tools (ie. TDR, Network Analyzers, Magnetics Analzyers,etc) were designed by the same engineers that you claim simplify analysis too much.

Silversmith Audio
At low frequencies, those assumptions will not affect the ultimate purpose of the circuit. In the water in a pipe model, RLC parameters are the basis of evaluation, and it can be easily shown, as you have done, that cables do not have enough effect on over all frequency response or impedance to be sonically discernable. Using the low frequency model and that same reasoning, it is pretty simple to also show the since most amps, pre-amps, CD players, etc all have approximately the same specs and should all sound the same too. High frequency design is a different matter. In this realm, the wave guide model, in which the electrical energy in transmitted between the wires instead of inside them, must be used.

Audioholics: So what do you consider "High Frequency Design"? 20kHz ??? Most RF Engineers consider anything less than 1GHz DC. If you would like to discuss transmission line theory, I will be happy to entertain you. However, if you are trying to apply relevancy at audio frequencies, then you may wish to change topics as it would be like trying to make a fish ride a bicycle.

Silversmith Audio
By high frequency design, I do indeed refer to work in the GHz range. The concepts very clearly show why a cable can have an audio impact. My cables are designed using those equations. If the equations are valid (as far as I know there have been no major changes recently in electromagnet theory), the cable design is valid. If you wish to keep the belief that transmission line theory has no bearing on audio, then you will never discover why cables make a difference. The concepts are there. They just have to be properly applied.

This is a much more accurate model of what really happens in electrical flow and describes, when properly applied, exactly why wires alter the signal enough to change your stereo's sound. It also predicts (and explains), interestingly, other phenomena:

1. A digital cable and a power cord, will affect the sound in exactly the same manner and for the same reason that the analog signal cables do.

Audioholics: And how is that?

Silversmith Audio
Properly applied transmission line theory explaints it very nicely.

Silversmith Audio

2. Lifting a cable off of the floor, depending on the construction of the floor, will audibly change signal transfer.

Audioholics: How so? So if I live in an apartment and are on the 7th floor for example, will the sonic signature of the cable be different? How does it change the signal transfer? Is it measurable? How do you know when you reach the right elevation for optimal performance? Do you use any sort of objective test tool to measure this? Is there any provable relationship in electromagnetics that validates this? What happens if someone lives in an underground dwelling? Theoretically if they placed their cables on the ground above them, they would be elevated, so would their sound be improved?

Silversmith Audio
Properly applied transmission line theory explains it very nicely.

Silversmith Audio]

3. The changes caused by cryo treatment of a conductor will audibly change signal transfer.

Audioholics: Really, how does that translate into provable measurement? Does the cable have to remain in its cryo state to maintain the alleged improvement in sound? Why not build a cryo freezing device to keep the cables frozen all the time? Better yet, why not use a superconductor? I believe you can purchase some from a National Research Lab for a few billion dollars. Who cares about price, this is high end audio :-)

Silversmith Audio
There are several changes that can be measured in a cryo treated metal. Plug one of them in to the proper transmission line equation, and, well, it becomes obvious. In fact, that parameter is also one of the reasons why my palladium alloy outperforms silver, copper, aluminum, gold, etc.

John Escallier
What does cryogenic treatment do to either the dielectric, or the conductors, which provides any change in characteristics that are measureable? When a pure metal is subjected to a cryogenic environment, the electron mean free path increases as a result of a drop in lattice collisions. Pure copper, in fact, will see it's mean free path extend to about 10 cm, and it's conductivity will increase three orders of magnitude. But, it is, to the best of my knowledge (and that of my co-workers) a process that follows a reversible path. We are unable to determine, by any tests that we are aware of, that a copper specimen has been subjected to a cryogenic environment. This does provide any parametric shifts which are capable of explaining a cryo change of any type. So I am unable to understand what parameters you would measure and use in a transmission line model. Only for alloys such as steel, where the lattice changes from FCC to BCC below the martensite start temperature, can we find evidence of a cryogenic induced transformation. Granted, I only work with liquid nitrogen, at 77K, liquid helium at 4.5K, and superfluid helium at 1.88K, and have no experience with the really cold temperatures.

SilverSmith Audio Cables Interview - page 2

Audioholics: Correct me if I'm mistaken, but isn't cryogenic freezing a form of stress relief for some metals? I'm a Mechanical Engineer so you're speaking my language on this subject, as I've been through many metallurgy materials classes. I believe cryo freezing is sometimes applied to brass instruments which rely on resonating for their tonal qualities. When they are formed into shape during manufacturing, internal stress build up at the bends, which may change the sonic resonance of the instruments. Note, I said change, not degrade. Some people or manufacturers then stress relieve the instrument after forming either by heat below melting, or via cryogenic freezing. Since heat costs money, and energy along with associated risks, most who do this stress relief use cryo freezing. With this, it only works and is valid on certain metals as a form of stress relief, and it has been known to alter (improve to some) the sonic resonance of brass resonating instruments.

Could it be that some companies actually believe this form of stress relief changes electrical properties of AC signals too? I wonder if this is where the confusion is and somehow, cable companies have latched
on to this valid process in an invalid way.

John Escallier Response
I gave an example of the FCC to BCC lattice change as an example, because I had to use that a while back. You are correct in the stress relief apps, I only gave an example which provided an atomic level lattice change, because I was discussing conductivity at the atomic level. Nobody here (at work) gives a hoot about wire changes through cryo, as they have found nothing, from kiloamp to nanovolt levels..only the fact that kapton is really the only polymer that is flexible at 4.5K.

I think you are right that some have latched onto a valid thing in an invalid way.

I will admit, I see a high percentage of engineers who really don't understand e/m theory. At work, though, the percentage is a whole lot lower than outside the lab.

Audioholics: There are even a higher percentage of exotic cable vendors that don't understand basic E/M theory. I suppose its because most of them actually don't have degreed engineers on staff. They simply repackage cable products from China.

Silversmith Audio
4. Connectors will have an audible effect on signal transfer.

Audioholics: I suppose that's true for bad connectors. Since I assume you utilize only good connectors, this should be a problem right?

Silversmith Audio
5. The true cause of skin effect is also has the single greatest audible affect on signal transfer (and it has nothing to do with impedance).

Audioholics: OK you got me there. But you may wish to run that one by a few PHD's in the field of electromagnetics and signal propagation. BTW, here is such a PHD that graciously wrote an article for us. Since you sound like a scientific guy, I am sure you know Dr. Howard Johnson. Check out his article about Skin Effect here . While your at it, you may wish to peruse some of the peer reviewed analytical analysis (based on proven and well documented relationships of skin effect), and measurements (utilizing $40,000 test gear specifically designed to measure it).

Skin Effect Relevance in Speaker Cables

Calculation Cable Inductance

Speaker Cable Face Off I

(What you will find in this article are measurements that illustrate Skin Effect causes an increase in AC resistance, while at the same time a reduction in internal inductance) at the frequency extremes. Seems to me that skin effect has everything to do with impedance changes in a cable. Of course it's obvious that at audio frequencies this plays little role in measurement changes, let along audible ones.

Silversmith Audio
I have read the work of Dr Johnson. His description of the skin effect as it relates to impedance is correct. However, he uses the "water in pipe" analogy of electrical flow and does not seem to realize how true nature of the skin effect affects audio cable design. It has nothing to do with impedance. Understanding skin effect in the transmission line world is the key here. And yes, it has the blessing of the PHD types. Again, you accurately discuss the skin effect as it applies to impedance. However, this is a very narrow application. Understanding skin effect in the transmission line world and how it really applies to cable design is the key. It has nothing to do with impedance.

Silversmith Audio
6. Dielectric effects, like "break in" will be audible.

Audioholics: Great, if they are audible, certainly they are measureable, else how would you know when you have reached dielectric nirvana?

Silversmith Audio
7. There is one, and only one, wave guide design that will eliminate, or reduce to the greatest possible levels, ALL signal altering mechanisms.

Audioholics: What Star Trek Episode was this from? I seem to remember this line :-)

Silversmith Audio
A common misunderstanding is that only the super high frequencies travel between the conductors. This is simply not true. ALL frequencies do. It's only the high frequency engineers who have to deal with those affects on their circuit.

However, while it does not affect the circuit at relatively low frequencies of human hearing, the consequences are easily audible. A simple way to confirm this is to look at the equation for the speed of electricity flow down a wire.

Audioholics: OK so if cable A is propagating at .7C and Cable B is propagating at .8C, where C is the speed of light, how does this affect audio frequencies? The wavelength of audio frequencies (3*10^8m/s /20*10^3kHz = 49,200 ft or 9 miles!). Surely you would be more concerned with resistive losses with cable lengths that long? I know, don't call you Surely ;)

Silversmith Audio
Actually you are incorrect. I am very concerned about propagation speed. It has a far greater impact on cable sound quality than resistive losses. However, your example of propagation has nothing to do with cable sound quality and is a good example of the mis-application of the concept. Again, properly applied transmission line theory is the key.

Silversmith Audio
It's parameter's are primarily dielectric related, not conductor related. Dr. Malcolm Hawksford, an Acoustical/Electrical Engineer at the University of Essex in England, also has several papers on audio topics that have been presented at AES and other venues that would also be helpful for you.

Audioholics: It's funny how cable soothsayers (aka Black Knights, Forum Cult Hobbyists) cling to the alleged research of Dr. Hawksford. I appreciate the reference. Perhaps you should check your reference in these discussions:

http://db.audioasylum.com/cgi/m.mpl?forum=cables & n=63636 & highlight=malcolm+hawksford

http://db.audioasylum.com/cgi/m.mpl?forum=cables & n=63801 & highlight=hawksford

http://db.audioasylum.com/cgi/m.mpl?forum=prophead & n=1373 & highlight=hawksford

[John Escallier] Regarding Dr. Hawsford

I note that you mentioned Hawksford. The equations he derived are absolutely valid equations, well manipulated to provide some interesting relationships. That is not where the problem with his essex echo paper lies.

The Hawksford analysis, as printed in the Essex Echo, neglects to include the storage of energy within the conductor...the 15 nHenry per foot number with copper. This is a result of the treatment of the wires as conductors whose voltage and current arise as a consequence of external fields. This is not the case for current carrying conductors. In addition, Hawksford neglected to test various guages of copper wire conductors, instead, substituted a steel conductor with a mu of approximately 100. Since the internal inductance is proportional to mu, the actual inductance he did not accout for was 1.5 microhenries per foot per wire, or 3 microhenries for the pair. On the assumption he used a meter of wire, that is about 10 microhenries unaccounted for in his simulation, and hence, the inductive overshoot in his test. Clearly, had he modelled this inductance, with the loop resistance of his wire, he would have found that the wire matches the formula for inductance provided us by Termen in 1947.

[John Escallier] Regarding Transmission Line Relevancy and Cryo Treatments

I have worked with transmission lines, transmission line theory, and application..and I am unable to understand how high frequency transmission line theory relates directly to analog audio applications. Although I note you have referred to transmission line theory several times in your responses to Gene, you have not elaborated in the least..Please do. I deal with superconductors. For me, skin effect is a way of life. Please elaborate on how properly applied transmission line theory deals with skin effect, and how it pertains to audio.

Audioholics Response: Jeff, do you have anything else to add to this discussion?

Silversmith Audio
I can't think of any additional comments but if I do in the next few days, I'll drop you a line. Thank you.

Please feel free to use my name in association with this correspondence. For you, and your readers, I am in no way trying to say that snake oil salesmen do not exist in the audio world. Unfortunately they exist in too many areas of our lives, audio or otherwise. In the world of audio, all cables, indeed all components, degrade the sound of an audio system. The key is to use properly applied engineering principals to design cables and other components to cause the least degradation possible.

Properly applied engineering and physics principals describe exactly why these components behave the way they do and even allow us to predict how a component will sound compared to one of a differing design, but same specs. These principals have never been disproved. Regardless of the math, however, there is a personal and subjective side to the music. Do not buy on the basis of a white paper, technical description, impressive sales pitch, or the ears of a reviewer. They can be useful tools, but LISTEN for yourselves. If you find yourself tapping your foot, moving and grooving to the music, there is probably something right about the sound, and it's not the frequency response. If you have the ears to tell the difference between live and recorded music, you have the tools to make the right decision.

Best Regards,

Jeffrey Smith
Silversmith Audio

Audioholics Epilogue

We would like to thank Jeff Smith of Silversmith Audio as well as John Escallier for participating in this informative discussion.

I find it interesting to note that Silversmith stresses the importance of applying proper transmission line theory to audio cables to understand the nature of their sonic attributes, yet furnishes no concrete proof, analysis, or technical contribution in this regard. While transmission line theory has relevancy at much higher frequencies (starting in the MHz range, depending on cable length), these frequencies are decades past the audio range of human hearing and the audio equipment in question. When designing high performance audio gear, it is critical to limit the analog bandwidth of such gear in order to attain optimal Signal to Noise (SNR) ratio and stability of the system. As example, Redbook CD incorporates a brickwall filter at 20kHz to bandwidth limit the payback system. While DVD-A and SACD have slightly higher analog bandwidths (about 100kHz), these frequencies are still far too low to be concerned with transmission line issues (reflection, return loss, power loss, etc) for any reasonable cable lengths. For more details, Current Trends article. We must also consider the fact that no current loudspeaker system (as example see our article on Skin Effect ) offers any reasonably controlled dispersion characteristic at frequencies much higher than 20kHz (the limit of human hearing). Let us also remember that most musical content is harmonic in nature at frequencies beyond 8kHz and based on the Fletcher & Munson curves, human hearing sensitivity is dramatically impaired at our 20kHz range limit. The argument of transmission line relevancy, or lack thereof at audio frequencies can be easily analyzed without delving into classical S parameter or Smith chart analysis. If we use classic and proven transmission line theory to determine the importance of cable impedance matching at audio frequencies, we derive the following:

wavelength (in meters) = v / (f*sqrt(er))

where f is the frequency of the signal in Hz, v is the velocity of the signal = 3x10^8 meters/second (vacuum) and er = velocity of propagation factor

If we consider the audio bandwidth limitation of 100kHz (based on maximum bandwidth of high resolution audio formats, considering sharp rolloff of power amplifiers to preserve SNR and stability), using a solid dielectric, with approximately the same dielectric constant for our cable as PE, we're running at around 66% velocity of propagation. This cuts the physical wavelengths where impedance matching matters to 1,980 meters or 6,496ft. A quarter wavelength, which is often used as the benchmark for where the characteristic impedance of a transmission line becomes critical, is then 495 meters or 1,624 feet. Many think a quarter wave is a bit too long to use, and prefer to go with 1/10 wavelength or so. If we are considering ultra high performance, and use the 1/10 th wavelength more conservative estimate to determine the point at which the cable's electrical length becomes long (with respect to a wavelength) to consider transmission line effects for audio cables, we derive 198 meters or 649ft. If you consider using common low resistance 10AWG Zip cord (2 ohm/1000ft loop resistance) as speaker cable at this length, then you would effectively be placing 1.3 ohms of resistance in series with your amplifier and speaker. With a nominal 4ohm impedance speaker, the resistive losses at these cable lengths would dominate and result in over 2.4dB of signal loss alone, not to mention destroying the damping factor of the system. Add in the reactance losses (inductance and capacitance) and we see a whopping 11dB of loss (not factoring in any potential amplifier stability issues from the high reactance of the cable).

My point here is the lumped impedance losses alone would have a far more profound affect than transmission line and/or distributed losses could ever dream of having.

The bottom line here is to consider the validity of sciences some exotic cable vendors enjoy promoting about their products and wonder why they often go through such great lengths of conjuring up such complicated reasoning as to why cables sound different, or for that matter allegedly better despite the lack of proof behind their convictions and claims. Much of the concepts of transmission line relevancy and cryogenically freezing cables contained herein can easily be disproven by classical analytical methods all credible engineers utilize on a daily basis to understand and design electrical systems. There is no rocket science to cable design, and to simply abandon proven critical metrics such as impedance (R,L,C) and geometry/shielding in favor of "just listening" or elaborate pseudo-science reasoning is something all consumers should be cognizant of if their goal is to obtain accurate, quality-controlled products in any facet of electronics and/or cabling.