[Part 1 begins with an overview of amplifiers, and discusses basic requirements for current and voltage output, and transient response. Part 2 discusses nonlinear distortions in amplifiers, as well as amplifier classes and modes of operation, beginning with Class A and AB. Part 3 covers Class D and other amplifier classes, output transistor types (MOSFET or BJT), and choosing an amplifier. Part 4 begins a discussion about loudspeaker cables and their effect on performance. Part 5 explores some provable characteristics of cable performance.]
6.11 Some passing comments
Unfortunately for people who have not had a very great degree of experience of the subject, sorting out what is solid from what is nebulous in terms of so much that has been published in the hi-fi press about loudspeaker cables is not an easy task. Nevertheless, despite some critics claiming that it is all a load of nonsense, it is the opinion of the authors that far too many people of good repute have claimed that sonic differences do exist to be able to dismiss the subject of loudspeaker cables out of hand.
One of the main obstacles to verification has been the cost of double-blind, controlled tests, with sufficiently large groups of subjects, coupled with the extraordinarily large number of possible equipment combinations and test locations, and the fact that a cable which shows benefits in one of those combinations may show no benefit in another. One component in isolation may not exhibit any noticeable difference over another component. However, when they are used in some combinations with other components, the differences may be readily apparent.
Whilst writing this book, a short, written note was received from Martin Colloms, an eminent authority on the subject of high fidelity loudspeakers. There is nothing to suggest, from either his professional conference papers or his well-respected text books, that he is either cloth-eared or a charlatan, and neither can the authors of this book refute his opinions. An extract from his communication will serve as a good summary.
A fair percentage of my reviewing life has been spent listening to and dealing with cables. Most evaluations I still do blind. I think that I have correlated with reasonable precision the key aspects of cable design from some 400 examples.
Metallurgy: the actual metal or alloy and its state/annealed/crystalline composition/extrusion direction and subsequent build orientation.
Dielectrics alter the sound, as do different plastics in film capacitors, the issues including dielectric factors, DF with frequency, piezo effects, self-damping including semiconducting insulation, and dielectric constant.
Geometry: spacing, stranding, Litz or bunched.
Mechanico-acoustic: physical strength, rigidity, self-damping, microphony issues.
RFI: can be very important; if you spectrum analyse (I work up to 1.5 GHz) the dominant RFI (radio frequency interference) is about 1 MHz for many loudspeaker cables. Many amplifiers are no longer anything you would recognise by 100 kHz, never mind the 1 MHz, and the RF gets in the output terminals and intermodulates around the feedback loop. A speaker cable is often a good medium-wave aerial.
Cables vary greatly in how they dump RF interference into the amplifier output port. The Zobel filter has little effect as it is generally buffered by 10 ohms. If the RFI doesn't get in the positive line, it common modes into the ground line.
At 1 MHz or more, the RFI hardly cares what kind of amplifier it is.
Contacts and connectors also matter, as do their tightness and vibration resistance. Acoustic energy ends up being mechanical watts and everything shakes about.
It is amazing how mild that vibration can be and still affect the sound of an electrical component, including a cable. Simple tests on vibration isolation and suppression show this clearly. Acoustic/ mechanical coupling remains one of the most insidious modes of sound quality loss.
In a recent conversation, Paul Frindle, one of the designers of the Sony Oxford R3 digital mixing console and many of the Sony professional plug-ins, recalled an occurrence during his time as a designer at Solid State Logic. There had been complaints from many studios around the world that mixes done on the small faders, which were simple potentiometers, sounded better than mixes done on the large faders, which operated VCAs (voltage controlled amplifiers).
They proceeded to make countless measurements at the factory but could find no apparent cause for the described difference in the sound. Some of Paul's comments were very reminiscent of loudspeaker cable anomalies.
P.F. You could end up in a situation where a single signal sounded fine, but its relationship with everything else was ill-defined. You could set up a mix on the pots (the small faders), with no VCAs, and it would sound great; but going through the main faders, with the VCAs, it would sound oddly wrong. Yet, if you soloed any single VCA channel you could hear no difference compared to the small fader. It was fascinating!
I found that this was caused by very small amounts of signal dependent delay variance through the VCAs, which was a complex kind of distortion that was unfamiliar within the design context. In my opinion this was one reason that led to the perception that the VCAs sounded bad, and many people went in the direction of consoles with motor driven faders instead.
This problem has some similarity to issues resulting from analogue to digital converters. For instance, if you have different clock-jitter, between channels - this is a lovely one! If your system has timing errors, when you use an ADC and a DAC back to back in a monitoring capacity, the clock is common and simultaneous to both the encoding and decoding stages. Since the timing errors are synchronous they partially null out.
However, if you delay the signal - or store it then play it back later - even though the system is the same, the jitter is no longer synchronous, and the resulting sound quality is compromised. In fact, the sound is reminiscent of the mixes done via the VCAs on the old SSL consoles. Of course, this is one of many good illustrations of how the sound can change through a system, even though the recorded bits are unaffected and no numerical errors have actually occurred.
Another similar problem can exist when you dither signals, because of relative correlation. When you dither a number of channels, the dither noise of any one channel should be unrelated and uncorrelated to any other, or you risk the relative correlation of the dither noise starting to become evident. You can end up with statistically partially 'mono' dither-noise, which can close in the stereo effect, especially on fade-outs.
Again, a single channel works fine, a stereo channel apparently sounds fine whilst the music is playing, but as channels build up with the same dither, the stereo begins to close in. We spent a fortune on the R3 finding 256 independent noise sources, because it was a strange problem to solve. It is hard to measure, and the sound problems are something which you often need to be working on day after day before they become clearly apparent.
You feel a sort of unease, that something is wrong, but you can't quite explain what it is. At one time, before we found an easier solution, 30% of the processor of the R3 was dealing with this problem. [And the R3 had about 3000 times the processing power of a current (2005) Pro Tools system. P.N.]
The illusiveness of these types of problems are not dissimilar to the illusiveness of definitive, irrefutable evidence about the differences between loudspeaker cables, but whereas Sony and Solid Stage Logic have been able to throw vast amounts of money at the problems, low volume producers of special loudspeaker cables have not enjoyed such luxury.
Two of Paul Frindle's sentences in the previous paragraph are very relevant to the issue. "It is hard to measure, and the sound problems are something which you often need to be working on day after day before they become clearly apparent. You feel a sort of unease, that something is wrong, but you can't quite explain what it is".
There are the hardnosed objectivists who would say that if you cannot measure it, then it cannot be important, but, once Solid State Logic solved their VCA problem, the complaints stopped from the users around the world. Surely this is evidence that the problem had been real enough. There is no logic in claiming that if Paul and his colleagues had not found the problem, then that would have proved that the problem was imaginary.
Similarly, if people cannot prove the reasons why cables should sound different, the implication is that not enough effort has been put into finding the problem, and not that the problem does not exist. And, of course, a subtle difference in the sound, given rise to by two loudspeaker cables, will almost certainly not be audible if the resolution of the loudspeakers being used to audition them is not, in itself, high enough to show up the difference.
Nevertheless, the evidence is now overwhelming that cables can give rise to sonic differences. The situation is somewhat paralleled by the relationship between stress and the functioning of the human immune system. Believe it or not, despite the fact that most people 'know' that when people are stressed they seem to be more prone to illness and infection, no clinically proven, scientific evidence currently exists to make a definite link between stress and susceptibility to illness.
However, in 2002, Ronald Glaser, et al, at Ohio State University, published in the Journal of Consulting and Clinical Psychology a paper stating that whilst no proof existed, the circumstantial evidence was too overwhelming to be ignored. The conclusion was that stress did lead to more illness and infection, and that the fact that they could not prove how it could do so did not mean that it could not be said that it did do so.