Gorsnak wrote:
This has nothing to do with acoustics. This is strictly a question of electromagnetics. Acoustics only happens after the speaker driver starts creating compression waves in the air. Everything before that stage is strictly a question of electron flow. Voltage, resistance, inductance, and capacitance. There is nothing else. If a scope can't see the difference between two electrical currents, then neither can your amp. They're both just collections of semi-conductors, after all.
...
But that's not what you seem to be saying. You seem to be saying that unmeasurable differences in electrical signals can nonetheless result in electronic circuits behaving differently. How? I just don't understand what you think is going on with the underlying physics here.
This may not have anything to do with acoustics (i.e. the study of vibrations in a medium), but it has everything to do with sound reproduction (i.e. electronically reproducing a recording).
Disclaimer: I am not a physicist. I don't (offhand) know much about electromagnetic forces beyond a high school level. However, I do know a considerable amount about the scientific method and theories of knowledge.
This is an argument about methodology and the meaning of measurement, not about physics.
First off,
unmeasurable does not mean
no difference, or even
no perceivable difference. It means that the difference is smaller than the degree of precision afforded by the tool of measurement, in this case an oscilloscope. You pointed this out in your post, and then promptly ignored it. At the level of quantum physics,
every individual wire conducts current slightly differently.
For
most practical purposes, the differences that cannot be picked up by an oscilloscope are small enough that there is no difference. This is because few electrical applications are as sensitive to minute fluctuations in current as analogue audio. Digital audio is reliable precisely
because it doesn't rely on subtle (and variable) differences in current to transfer information. It's not perfectly reliable — as BobDog pointed out, transferring digital information in real time can lead to variation in timing across a given wire.
However, just because the difference between different cables is below the maximum sensitivity of an oscilloscope (i.e. it is unmeasurable) does not mean that the difference is not there, or that the cables do not affect the signal in subtle and equally unmeasurable ways. And, for the sake of argument, I will grant that these minute differences might be
perceivable (not measurable) when they are converted into sound waves. BobDog is right, it's possible that our ears are more sensitive to these differences than an oscilloscope. I have no idea whether they
actually are; perhaps a properly done ABX test would
help resolve this problem.
Of course ABX testing has its own limitations, including the statistical limitations mentioned by BobDog. Also, unlike an oscilloscope, ABX testing tests an audio
system, not one individual component. ABXing the system after ones component is changes (i.e. a cable) will tell you whether that component affects that particular system perceptibly, but the result cannot necessarily be generalized to say that the component has no effect in
any system. It adds some evidence to the pile, but it is not conclusive. I can think of no way of doing an ABX test that can easily be generalized across all systems. By the way, I consider an audio system to include all system components, the room in which it is tested, and the position of the listener. A system with the listener 5' from the speakers is different from a system with the listener 10' away. It's not impossible that a cable swap might have no perceptible effect on the 5' system but still affect the 10' system.
As BobDog has mentioned, an ABX test that shows a particular cable has no perceptible effect on a particular system provides reasonably strong
evidence that it's not doing anything, but it does not provide conclusive
proof that it's not. The best proof we can have is a large body of evidence that is collated from many different tests of many different systems. Statistically, a large body of evidence is reasonably (though not absolutely) trustworthy. As single ABX test, especially one with few samples, is not enough to conclusively evaluate the performance of a component in all possible systems.