Arny Krueger
September 26th 07, 02:39 PM
AES paper 6858 (2006) : Audio cable distortion is not a myth!
ABSTRACT
Specialist audio cables are often sold to the consumer on the basis of
eyebrow-raising claims for technical performance, though to date no
repeatable test has shown any effect more surprising than mild
frequency-selective attenuation. However, because the loudspeaker load is
typically nonlinear and causes harmonic currents to flow, finite impedance
in an audio cable does indeed cause harmonic voltages to appear across the
loudspeaker. This distortion term is similar to, or even greater than, that
produced by the amplifier's intrinsic nonlinearity.
1. INTRODUCTION
Of all the various arguments to have raged over the general subject of
audio, none has been both so fierce and so enduring as that about specialist
cables. Simply put, the 'subjectivist' premise is that every cable design
has its own 'sound', influenced by every factor imaginable - cable materials
(both conductor and insulation), geometry, stranding etc. - while the common
'objectivist' riposte is that although the series resistance of a
loudspeaker cable can certainly have an effect on the signal if it is too
high (more than a few percent of thc loudspeaker's own nominal impedance),
all other factors pale into insignificance.
Measured evidence certainly provides much better support for the objectivist
position. Faced with this, subjectivists typically argue either that the
measurements were too simplistic and overlooked some as-yet-unidentified
factor of major significance, or that the measurements are invalid because
they have almost invariably been made with sinusoidal test stimuli, and
'music isn't made up of sinusoids'. These two arguments boil down to very
nearly the same thing, essentially that there may be some effects which do
not show up on conventional tests. It is actually very hard to envisage any
form of distortion, in the broadest sense, which would not show up at all on
any given test but it is certainly true that some are far more visible than
others, depending on how a test is carried out.
In addition to this, there is a very large grey area in the whole subject of
audibility of distortion. It is also always possible that more careful
analysis of existing types of measurement may show some distortion that had
previously been overlooked. For all these reasons - but above all driven by
a desire to accumulate some evidence which would be sufficient to refute the
arguments of one camp or the other - I decided to conduct an investigation,
by theory and measurement, of audio cables which would be as thorough as
possible, not excluding investigation of any apparently bizarre claims made
by subjectivists and cable vendors.
7. CONCLUSION
The audio cable market certainly owes something to the appeal of 'audio
jewellery' but there is also, with little doubt, a significant element of
sell-delusion on the part of consumers and probably also reviewers,
retailers and manufacturers. However, this paper has suggested a plausible
mechanism, not invoking any outlandish concept, by which differences between
loudspeaker cables may on occasion become audible. The cure - low resistance
and low inductance - is simple and relatively inexpensive to apply and it is
to be hoped that more manufacturers will attempt to do so.
ABSTRACT
Specialist audio cables are often sold to the consumer on the basis of
eyebrow-raising claims for technical performance, though to date no
repeatable test has shown any effect more surprising than mild
frequency-selective attenuation. However, because the loudspeaker load is
typically nonlinear and causes harmonic currents to flow, finite impedance
in an audio cable does indeed cause harmonic voltages to appear across the
loudspeaker. This distortion term is similar to, or even greater than, that
produced by the amplifier's intrinsic nonlinearity.
1. INTRODUCTION
Of all the various arguments to have raged over the general subject of
audio, none has been both so fierce and so enduring as that about specialist
cables. Simply put, the 'subjectivist' premise is that every cable design
has its own 'sound', influenced by every factor imaginable - cable materials
(both conductor and insulation), geometry, stranding etc. - while the common
'objectivist' riposte is that although the series resistance of a
loudspeaker cable can certainly have an effect on the signal if it is too
high (more than a few percent of thc loudspeaker's own nominal impedance),
all other factors pale into insignificance.
Measured evidence certainly provides much better support for the objectivist
position. Faced with this, subjectivists typically argue either that the
measurements were too simplistic and overlooked some as-yet-unidentified
factor of major significance, or that the measurements are invalid because
they have almost invariably been made with sinusoidal test stimuli, and
'music isn't made up of sinusoids'. These two arguments boil down to very
nearly the same thing, essentially that there may be some effects which do
not show up on conventional tests. It is actually very hard to envisage any
form of distortion, in the broadest sense, which would not show up at all on
any given test but it is certainly true that some are far more visible than
others, depending on how a test is carried out.
In addition to this, there is a very large grey area in the whole subject of
audibility of distortion. It is also always possible that more careful
analysis of existing types of measurement may show some distortion that had
previously been overlooked. For all these reasons - but above all driven by
a desire to accumulate some evidence which would be sufficient to refute the
arguments of one camp or the other - I decided to conduct an investigation,
by theory and measurement, of audio cables which would be as thorough as
possible, not excluding investigation of any apparently bizarre claims made
by subjectivists and cable vendors.
7. CONCLUSION
The audio cable market certainly owes something to the appeal of 'audio
jewellery' but there is also, with little doubt, a significant element of
sell-delusion on the part of consumers and probably also reviewers,
retailers and manufacturers. However, this paper has suggested a plausible
mechanism, not invoking any outlandish concept, by which differences between
loudspeaker cables may on occasion become audible. The cure - low resistance
and low inductance - is simple and relatively inexpensive to apply and it is
to be hoped that more manufacturers will attempt to do so.