Why do you think Bi-Wiring improves the sound ?

Alan Shaw, the designer of Harbeth speakers, has typically strong views on the topic. These three quotes were taken from three separate posts on the Harbeth forum.

"I really wish this subject of biwiring would just disappear up its own terminals. I don't have many ambitions in life but killing this discussion by deleting the biwire terminals and reverting to just one input pair is going to be at the top of my 2011 New Year's Resolution list for the remaining models that still feature biwire legacy terminals! You've given me a real motivational boost!

"The terminals were fitted for one reason and one reason only: to give the user choice. Have I ever used them at exhibitions? No. Have I ever used them for critical listening? No. Have I ever used them during the design of the speaker? No. When we were offering the biwire terminals, right at the end of the design process (which has all been with single wire) I took a saw to the prototype PCB, cut in in half to isolate the bass and tweeter sections and then made a pretty PCB layout based on that. Did I listen to the biwired crossover before authorising production? No. Do I believe that even 0.00000001% of enhanced performance can be gained? No. 

"Of all the subjects ranged over in the speaker arena, this one is a complete and utter waste of time - in my opinion. But what do I know about it? I only design the speakers ....... !

"Biwiring does do one thing very well though: it introduces the one and only, much appreciated 'fiddle factor' to allow individuals a physical and psychological interaction with their speakers. What else can you do to them other than dust and polish them?"

***

"A biwire link is gold plated brass about, say, 30mm long. The claim is that this particular 30mm long piece of highly conductive metal is somehow, magically, more important than any other 30mm piece of perhaps less highly conductive metal anywhere else in the chain between the loudspeaker drive units and the power station a hundred miles away which is supplying the current that causes the cone to move and a sound to be generated. Does that sound logical? Does that sound an intellectual argument that a professor of engineering at a university could or should set his students studying? Of course not. It's a daft fixation on what is, from a point of electrical conduction, probably the best "link" in the chain from the point that the mains supply enters the house.

"The biwire link has this fascination for one reason and one reason alone - it's accessible by the user. So it lends itself to being fiddled with and to all the associated gratification of adjusting ones hifi.

"This is a non-issue. Pick a genuinely 'weak' part of the signal chain and experiment, but this big, fat brass part with countless billions of surplus electrons isn't the hold grail. Of that I am totally and absolutely certain as I've stated. You'd be better off paying attention to, let's say, 30mm of copper track on the printed circuit board that the binding posts are pressing onto which is vastly less conductive because it is thousands or millions of times thinner than the biwire link. But of course, that would involve opening the speaker and voiding the warranty."

***

"Do the exponents of this biwire mania have any concept at all that a current is a circulating concept? Circulating from the power station, through your amp, cables, crossover, voice coil and back again to the power station? Anyone into biwire connectors grasping that concept please? That concept of how electricity actually works is why there is a live and neutral pin on your wall socket. There has to be a flow. And what impedes the flow is resistance. And resistance is associated with thin parts, like the voice coil (about 6 ohms). So the fact that the biwire link has a resistance of perhaps 0.000001 ohm compared to the voice coil's 6 ohms means that as a component in the circulating loop, what dominates the resistance by a huge factor is the voice coil.

"If the concept of a circulating current is unclear or distrusted then the whole scientific world we live in collapses."

"So Alan Shaw says it doesn’t matter and Richard Vandersteen says it absolutely matters. Two very respected speaker designers with completely different views of the same subject. Do you just cherry pick one because it supports what you believe or fits your experience? What a logical fallacy. I would not add a set of binding posts to a set of speakers just to try it out. If there are already two sets of binding posts, give it a try for yourself and make your own decision."

There are several problems with your post. First, assuming that you were responding to my post, I didn’t cherry-pick anything. I simply reproduced some posts of Shaw’s that explain his position.

Secondly, can Vandersteen actually demonstrate HOW it matters? Shaw explains why, in his view, it doesn’t, and is typically rigorous in his scientific approach to such matters. That doesn’t mean that he cannot be mistaken about something, but it does mean that he can explain, with a scientific foundation, why he holds a particular position.

With a quick search, I found anecdotal claims by Vandersteen that bi-wiring sounds better on his speakers, and this:

Additional experiments with a Hall Effect probe revealed that high-current bass frequencies created a measurable field around the wires that expanded and collapsed with the signal. We believe that this dynamic field modulates the smaller signals, especially the very low level treble frequencies. With the high-current signal (Bass) separated from the low-current signal (Treble) this small signal modulation was eliminated as long as the cables were separated by at least an inch or two. (To keep the treble cable out of the field surrounding the bass cable.)

Note that he says "We believe...". Not exactly hard science, though perhaps he is on to something.

I found a related article (on "qacoustics" UK) with some experiments appearing to support the ides that bi-wiring confers benefits. Don, a regular and knowledgable contributor to the audiosciencereview.com site had this to say in response:

Because a single wire carries woofer and tweeter current; bi-wiring means the woofer wire carries only woofer current, tweeter wire tweeter current, though voltage is the same for both. A plot of voltage would show the same voltage applied to woofer and tweeter (less changes due to wire loss, insignificant in practice).

Single wire:
Amp -> single cable -> woofer + tweeter = single cable carries all current

Bi-wire:
Amp -> woofer cable -> woofer = woofer cable carries only woofer current; crossover reduces tweeter current to woofer

Amp -> tweeter cable -> tweeter = tweeter cable carries only tweeter current; crossover reduces woofer current to tweeter

The crossover makes the woofer look like a higher impedance to the tweeter cable so tweeter current is reduced in the woofer wire, and likewise the tweeter crossover makes a higher impedance to the woofer cable for tweeter current, so there is less interaction in the wire. The net energy the amp delivers, and that the woofer and tweeter each receive, is the same whether you use a single wire or bi-wire. The wire itself contributes negligibly to distortion and so the bi-wire argument is a red herring (is a false argument). The amplifier and speakers dominate (by orders of magnitude) the distortion.

HTH - Don

And another commenter on a different site:

QAcoustics article cited above shows just how SUBTLE the differences can be between Bi-Wire and Single Wire connected Speakers. Change in IMD Levels are only 1-3 dB, with the IMD voltages being 70 dB below the Fundamental Signal Levels...which is 1 part in 10^7 or 0.00001 %.....probably impossible for even Golden Ears to detect.

If someone has a scientific explanation for why bi-wiring may be superior in practice, please do provide it.

This sure sounds like hard science. Lots and lots of "I believe" in there. He couldn't even be bothered to listen to it. It is no more authoritative than Vandersteen's anecdotal evidence. Is there any evidence that Richard Vandersteen does not use science in his designs? I would say that quite the opposite is true. In fact, he would say that designers that choose to not pay attention to phase in their speakers are absolutely wrong. He does back that with measurements. I am not saying either is right or wrong. I am just pointing out that many people here love to pick out the expert that they want to believe. You know, appeal to authority.

How ironic that you would accuse me of cherry-picking, then pluck just one of a number of quotes from Shaw that I provided, while ignoring those that include scientific reasons supporting his skepticism.

The onus of proof is on those, including Vandersteen, who claim that biwiring DOES make a meaningful difference, to support those claims with science. Has he done so? I am still awaiting proof that he has.

This:

Is there any evidence that Richard Vandersteen does not use science in his designs?

is a silly straw man. No one has suggested anything of the sort. And of course the fact that Vandersteen does, broadly speaking, use science in his designs, is irrelevant to the specific question being debated.

Shaw and the many other skeptics have provided technical reasons why biwiring does not make a meaningful, audible difference. If you are able to provide evidence to the contrary, I'll be happy to listen to it.

The above two posts are valuable in that they both touch on what may be useful points. Bi-wiring, for example, with at least some Vandersteen speakers (e.g. 2CE), apparently allows the user to make tonal changes via dials on the back of the speaker. In other words, the speaker is designed to be bi-wired, and to allow the user to adjust the midrange and tweeter levels.

Assuming that the two features are not distinct (i.e. completely separate), then of course the combination will allow the user to change the sound. Considering that to be an advantage is one thing, but to argue that bi-wiring alone improves sound is another.

With regard to @hshifi's point that bi-wiring allows the user to "tune" the wires is similar. That would be very different from the claim that a simple bi-wire setup is somehow superior to a single cable. Simply adding "more" cable is not in any way supportive of bi-wire being superior, and for what should be obvious reasons.