Chu Gai
Senior HTF Member
- Joined
- Jun 29, 2001
- Messages
- 7,270
Recently a poster inaccurately quoted what I have said regarding extensive and carefully-controlled "blind and double-blind" listening tests that we at DAL have conducted over many years to determine if any "truly audible" differences exist between loudspeaker cables representing a wide range of pricing, size and design approaches.
From these comparisons, which encompassed a significant number of competent listeners and a wide range of audiophile amps and loudspeakers, the results we obtained led us to confidently conclude the following:
1) No audible differences existed between any of the cables assessed for lengths under 25 feet - if "stable" power amps and well-designed loudspeakers with reasonable input impedances were used.
2) When audible differences were substantiated they could be traced to:
a) "high-performance" power amps with excessive inverse-feedback and inherent stability problems that caused them to became unstable and oscillate at supersonic frequencies (creating audible distortion) when used with some low loss, high capacitance, low-impedance cables, and/or
b) a loudspeaker cable with a high series inductance and or a high series resistance, which sometimes caused an audible roll-off of high frequencies and/or a "dulling" of transient detail when used with a loudspeaker whose input impedance dropped below about 2 ohms over a reasonable range of frequencies, especially above about 10 kHz.
Beyond these special cases, no audible differences were ever substantiated between the most expensive, exotic-looking, widely-advertised loudspeaker cables and quality #12 AWG ZIP Cord having the same length.
The many listening comparisons we have made over the past 20-odd years between audiophile loudspeaker cables were carefully controlled according to proper scientific method and good engineering practices. Every reasonable effort was made to ensure that listening comparisons did not encompass spurious factors that might bias or skew results. A wide variety of music and test tones (impulses, tone-bursts, etc.) were used, along with a variety of audiophile loudspeakers and power amps. The amplifiers used varied in price from about $200 to over $10,000. The rooms used for critical listening comparisons were always acoustically well-damped, typically about 25 feet wide by 15 feet deep, with the loudspeakers placed along the long wall, about 10 feet from the listener and separated by an included angle of about 90 degrees. Listeners included DAL employees, salespersons of local audio stores, and numerous visiting audiophiles.
Among the approaches used in evaluating whether verifiable audible differences existed between different loudspeaker cables were: 1) pretending to switch cables but not doing so, 2) switching between cables but not letting the listener know which was being heard (blind and double-blind regimens), 3) switching between cables while keeping the listener informed as to which cable was being used.
The results we have obtained consistently correlate very well with those published within professional and trade journals by competent engineers who have performed similar tests and comparisons between cables. And, they have always correlated with those expected from the teachings of well-known transmission-line theory, network theory, etc. and predictions based upon the proper interpretation of a full set of lab-quality measurements.
There really are no relevant unknowns with respect to transmission-line theory and the measurement of meaningful cable performance parameters. The goofy beliefs and theories that need to be questioned are those often loudly annunciated by persons who pretend they possess competent knowledge and understanding of cable theory and measurement but lack the professional-level credentials and underpinnings to do so. The bottom line is very simple: if it can be heard, it can be identified, measured and quantified by well-known means within a well-equipped laboratory manned by personnel possessing appropriate professional credentials. (Those who believe otherwise are doomed to be victims of those who pursue the design and sale of products based upon pseudo science and nonsensical advertising claims.)
Sadly, the allure of expensive, "high-tech appearing" loudspeaker cables can be traced to an industry typically missing qualified electrical engineering personnel but brimming with personnel who excel at composing "great-sounding advertising prose" containing claims for technology and performance that are virtually baseless. (A sad commentary regarding a very large and profitable industry!)
But, the advertisements of some cable manufacturers do contain what are purported to be measured comparisons between different cables, including ZIP Cord, which is portrayed to exhibit only about 3% efficiency at 60 Hz. However, common sense reveals that such a low efficiency would cause a typical "AC extension cord" to turn "white hot" if connected to an ordinary toaster. Hmmm!
Another advertisement compares loudspeaker cables according to their Joule rating - but a Joule is merely a watt-second, used as a unit of energy-storage when comparing batteries or some capacitors. Hmmm! Thus, such graphs portray totally meaningless information that is not only false but also misleading and downright silly from an engineering point-of-view. When asked why they do not publish meaningful measured performance specifications for their cables, such as loss Vs frequency into typical loudspeaker load impedances, series resistance, parallel capacitance, series inductance, frequency dispersiveness, etc., representatives of most large cable companies usually reply that such performance attributes are meaningless. Hmmm!
I have recently asked five very competent Professors of Electrical Engineering at prominent universities their opinion of audiophile loudspeaker cable design and advertising. The language of their replies would probably not be permitted even here on the INTERNET. Needless to say, they share the feelings of all competent and informed electrical engineers that the advertising claims and specifications for audiophile loudspeaker cables are without substance and cannot be verified by theory, measurements nor proven by competent blind listening comparisons. The same conclusions have been stated in a few magazine articles and peer-reviewed audio journal papers by authors possessing credible academic and technical backgrounds.
Thus, the question arises as to why any competent manufacturer would not at least attempt to design loudspeaker cables with measurable electrical properties that represent the teachings of network/transmission-line theory and the fruits of good engineering practice?
For example, at very high audio and low radio frequencies, cables with a relatively long length can best be characterized by applying "transmission-line theory", while at lower frequencies it is easier (and probably more accurate) to design and analyze cables by using "network theory". For example, using the teachings of transmission-line, the "optimum cable" would be one whose "characteristic impedance" was equal to the average impedance of the load. However, while this solution somewhat applies to loudspeaker cables, it results in a cable whose relatively large capacitance and low inductance might cause some "high-performance" (but frequently unstable) power-amps to oscillate - usually at super-sonic frequencies, detectable as audible distortion on transients, etc. (This is the reason that some expensive audiophile loudspeaker cables incorporate an expensive, hi-tech looking box at the loudspeaker end of the cable which houses a simple inexpensive resistor and capacitor, often called a "Zobel Network". Hmmm!
Recognizing what frequently are shortcomings of the "ideal cable" designed according to transmission-line theory, competent engineers apply the teachings of "network theory" to design loudspeaker cables with lengths less than about 25 feet. In this case, an ideal loudspeaker cable becomes one whose series resistance, series inductance and parallel capacitance are all minimal. The combination of these properties insures the lowest loss across the audio spectrum while minimizing the probability of amplifier instability. Such a cable might be one with very large diameter, low resistance wires, separated by a distance that minimizes capacitance without increasing inductance beyond an amount that would alter fhigh-frequency performance.
Achieving either of these design properties and goals can be accomplished without incurring a high engineering and manufacture costs that might lead to high retail prices - such as those currently being charged for some exotic, hi-tech looking cables with questionable performance properties. So, why not design, manufacture and competitively market loudspeaker cables based upon advertising that articulates their meaningful design parameters and competently measured electrical specifications - rather than the flooby-dust, buzzard-salve and gobbledygook specs presently found in too many cable advertisements? Are most cable manufacturers afraid to advertise meaningful performance parameters, such as the resistance, inductance, resistance, propagation-factor, etc, for their cables. (Can most cable manufacturers even measure them?) Hmmm!
So, the best present advise is very simple: CAVEAT EMPTOR (let the buyer beware)! Best of listening,
John Dunlavy
From these comparisons, which encompassed a significant number of competent listeners and a wide range of audiophile amps and loudspeakers, the results we obtained led us to confidently conclude the following:
1) No audible differences existed between any of the cables assessed for lengths under 25 feet - if "stable" power amps and well-designed loudspeakers with reasonable input impedances were used.
2) When audible differences were substantiated they could be traced to:
a) "high-performance" power amps with excessive inverse-feedback and inherent stability problems that caused them to became unstable and oscillate at supersonic frequencies (creating audible distortion) when used with some low loss, high capacitance, low-impedance cables, and/or
b) a loudspeaker cable with a high series inductance and or a high series resistance, which sometimes caused an audible roll-off of high frequencies and/or a "dulling" of transient detail when used with a loudspeaker whose input impedance dropped below about 2 ohms over a reasonable range of frequencies, especially above about 10 kHz.
Beyond these special cases, no audible differences were ever substantiated between the most expensive, exotic-looking, widely-advertised loudspeaker cables and quality #12 AWG ZIP Cord having the same length.
The many listening comparisons we have made over the past 20-odd years between audiophile loudspeaker cables were carefully controlled according to proper scientific method and good engineering practices. Every reasonable effort was made to ensure that listening comparisons did not encompass spurious factors that might bias or skew results. A wide variety of music and test tones (impulses, tone-bursts, etc.) were used, along with a variety of audiophile loudspeakers and power amps. The amplifiers used varied in price from about $200 to over $10,000. The rooms used for critical listening comparisons were always acoustically well-damped, typically about 25 feet wide by 15 feet deep, with the loudspeakers placed along the long wall, about 10 feet from the listener and separated by an included angle of about 90 degrees. Listeners included DAL employees, salespersons of local audio stores, and numerous visiting audiophiles.
Among the approaches used in evaluating whether verifiable audible differences existed between different loudspeaker cables were: 1) pretending to switch cables but not doing so, 2) switching between cables but not letting the listener know which was being heard (blind and double-blind regimens), 3) switching between cables while keeping the listener informed as to which cable was being used.
The results we have obtained consistently correlate very well with those published within professional and trade journals by competent engineers who have performed similar tests and comparisons between cables. And, they have always correlated with those expected from the teachings of well-known transmission-line theory, network theory, etc. and predictions based upon the proper interpretation of a full set of lab-quality measurements.
There really are no relevant unknowns with respect to transmission-line theory and the measurement of meaningful cable performance parameters. The goofy beliefs and theories that need to be questioned are those often loudly annunciated by persons who pretend they possess competent knowledge and understanding of cable theory and measurement but lack the professional-level credentials and underpinnings to do so. The bottom line is very simple: if it can be heard, it can be identified, measured and quantified by well-known means within a well-equipped laboratory manned by personnel possessing appropriate professional credentials. (Those who believe otherwise are doomed to be victims of those who pursue the design and sale of products based upon pseudo science and nonsensical advertising claims.)
Sadly, the allure of expensive, "high-tech appearing" loudspeaker cables can be traced to an industry typically missing qualified electrical engineering personnel but brimming with personnel who excel at composing "great-sounding advertising prose" containing claims for technology and performance that are virtually baseless. (A sad commentary regarding a very large and profitable industry!)
But, the advertisements of some cable manufacturers do contain what are purported to be measured comparisons between different cables, including ZIP Cord, which is portrayed to exhibit only about 3% efficiency at 60 Hz. However, common sense reveals that such a low efficiency would cause a typical "AC extension cord" to turn "white hot" if connected to an ordinary toaster. Hmmm!
Another advertisement compares loudspeaker cables according to their Joule rating - but a Joule is merely a watt-second, used as a unit of energy-storage when comparing batteries or some capacitors. Hmmm! Thus, such graphs portray totally meaningless information that is not only false but also misleading and downright silly from an engineering point-of-view. When asked why they do not publish meaningful measured performance specifications for their cables, such as loss Vs frequency into typical loudspeaker load impedances, series resistance, parallel capacitance, series inductance, frequency dispersiveness, etc., representatives of most large cable companies usually reply that such performance attributes are meaningless. Hmmm!
I have recently asked five very competent Professors of Electrical Engineering at prominent universities their opinion of audiophile loudspeaker cable design and advertising. The language of their replies would probably not be permitted even here on the INTERNET. Needless to say, they share the feelings of all competent and informed electrical engineers that the advertising claims and specifications for audiophile loudspeaker cables are without substance and cannot be verified by theory, measurements nor proven by competent blind listening comparisons. The same conclusions have been stated in a few magazine articles and peer-reviewed audio journal papers by authors possessing credible academic and technical backgrounds.
Thus, the question arises as to why any competent manufacturer would not at least attempt to design loudspeaker cables with measurable electrical properties that represent the teachings of network/transmission-line theory and the fruits of good engineering practice?
For example, at very high audio and low radio frequencies, cables with a relatively long length can best be characterized by applying "transmission-line theory", while at lower frequencies it is easier (and probably more accurate) to design and analyze cables by using "network theory". For example, using the teachings of transmission-line, the "optimum cable" would be one whose "characteristic impedance" was equal to the average impedance of the load. However, while this solution somewhat applies to loudspeaker cables, it results in a cable whose relatively large capacitance and low inductance might cause some "high-performance" (but frequently unstable) power-amps to oscillate - usually at super-sonic frequencies, detectable as audible distortion on transients, etc. (This is the reason that some expensive audiophile loudspeaker cables incorporate an expensive, hi-tech looking box at the loudspeaker end of the cable which houses a simple inexpensive resistor and capacitor, often called a "Zobel Network". Hmmm!
Recognizing what frequently are shortcomings of the "ideal cable" designed according to transmission-line theory, competent engineers apply the teachings of "network theory" to design loudspeaker cables with lengths less than about 25 feet. In this case, an ideal loudspeaker cable becomes one whose series resistance, series inductance and parallel capacitance are all minimal. The combination of these properties insures the lowest loss across the audio spectrum while minimizing the probability of amplifier instability. Such a cable might be one with very large diameter, low resistance wires, separated by a distance that minimizes capacitance without increasing inductance beyond an amount that would alter fhigh-frequency performance.
Achieving either of these design properties and goals can be accomplished without incurring a high engineering and manufacture costs that might lead to high retail prices - such as those currently being charged for some exotic, hi-tech looking cables with questionable performance properties. So, why not design, manufacture and competitively market loudspeaker cables based upon advertising that articulates their meaningful design parameters and competently measured electrical specifications - rather than the flooby-dust, buzzard-salve and gobbledygook specs presently found in too many cable advertisements? Are most cable manufacturers afraid to advertise meaningful performance parameters, such as the resistance, inductance, resistance, propagation-factor, etc, for their cables. (Can most cable manufacturers even measure them?) Hmmm!
So, the best present advise is very simple: CAVEAT EMPTOR (let the buyer beware)! Best of listening,
John Dunlavy
