Al Pacino's gift to impressionists was a line in - you know, the one about how every time he breaks away, they dra-a-ag him back. For Michael Corleone, it was the hold of la Cosa Nostra. For me? The humble BBC LS3/5A. Just when I thought that Rogers' new LS3a clonette would be the last I'd ever write about the Greatest Small Two-Way Monitor In History™, BAM!! another LS3/5a wannabee appears. Only this time, the tale doesn't involve wishful thinking: the Stirling LS3/5a V2 is real. It exists. And it's very, very special.

Let's back up a bit, though, to get all of you up to speed so that you may appreciate the importance of this latest chapter in the edda. Most LS3/5a aficionados, especially those who communicate by mouse, know the following facts:

1) The genuine, KEF-driver-equipped original is no longer made in any serious, serial production form, though one or two entrepreneurs will argue to the contrary. Any new LS3/5as that turn up are made from 'new-old-stock' drivers and parts, while cabinets are less of a problem.

2) Genuine LS3/5as of various stripes now command crazy money on eBay.

3) Reviving the speaker, which specialists estimate could still sell up to 500 pairs per annum, has proven to be just about impossible because KEF no longer makes the drivers for OEM users, the company with the tools and dies for the drivers - Harbeth - has no interest in resuscitating the speaker for fear that it will affect sales of its own mini-monitor, the grille material is hard to come by and licensing permission from the BBC is complex.

With all these obstacles and plenty of second-hand speakers around, why, then, would someone like Doug Stirling and a bunch of other LS3/5a speculators (for lack of a better term) even bother? Harbeth, ProAc, Spendor and others have spent years developing more modern substitutes, with much success, and the world isn't short of brilliantly-performing, small, two-way dynamic speakers between £500 and £1000 per pair. But this is the nature of both LS3/5a-lust and audiophilia. Like the Lancia Fulvia coupe, the Rolex 'Paul Newman' Daytona chronograph, the Leica rangefinder camera and the AGA cooker, the LS3/5a exhibits an inimitable, unrepeatable magic.

No matter how hard the manufacturers try, nothing will dislodge said icons from their positions in their respective disciplines. You can make a better-built replacement, faster, sleeker, more efficient, more sensual, more attractive - whatever virtues you care to enhance, but you cannot shake the faithful. Call it nostalgia, mass hypnosis, delusional behaviour - doesn't matter. The cognoscenti want the original and will Accept No Substitutes. Which is good news only for antique dealers. And the LS3/5a, no matter what its detractors argue, has a unique charm that - for certain listeners - cannot be denied, imitated nor dismissed.*

Stirling, though, would not be deterred, in no small part because a goodly chunk of his business is in the professional sector rather than the domestic, and he's a BBC licensee. And however much certain rivals of his would like to believe, even in the cynical world of pro studios there's still a cult following for the LS3/5a. Although Doug has done his best to refurbish old LS3/5as and make new ones from original parts when stocks are available, he wanted a regular source of brand-new speakers to sell to his clientele. He wanted to be able to offer the added benefit of a 5-year warranty and the absolute assurance of the future availability of spare parts.

Enter the LS3/5a V2.

That 'V2' suffix is extremely important. In order to avoid any accusations of deception, and yet still be allowed to use the BBC logo, Doug created the V2 appellation to distinguish it from KEF-based originals. Not that it would ever be a problem, because you only have to look at the drivers to know they're not KEFs. But suffice it to say, in a world of whining wankers, Doug didn't want to court accusations of foul play. So, after the initial trial batch of 100 pairs, every Stirling LS3/5a will come with an invoice and owner's certificate that clearly indicates 'V2' status.

Now what, exactly, is a V2? If every single element or the original must be present for a speaker to be a proper LS3/5a, and the key components - the KEF drivers - are not being used, what gives Stirling the right to call this an LS3/5a? Here's where this speaker becomes the most interesting solution so far to the lack of LS3/5as.

Cabinets? No problem. Stirling even went a stage further and reverted to the screwed-back types used in the very earliest examples, based on his thin-walled 'Reference' cabinets modelled after the famous 001/002 pair featured in the shoot-out a few years ago. He even tunes them by not screwing the backs too tightly, which changes the sound. Grille material? Stirling bought a few rolls of Tigon some years ago, enough for many hundreds of pairs (and he also discovered a cloth substitute that was used on the very earliest Rogers models).

But Stirling is no masochist, so he's emulated not the single-wire 15 ohm model but the later bi-wire 11 ohm replacement, with readily-available drivers closer to his 'solution'. Doug explained that available drive units would do 11 ohms, but insisting on 15 ohms would have meant changes to both the coils and the surrounds - using rubber instead of more modern materials - and the price would be less resistance to aging. (But see below...)

To achieve this without KEF drivers, he enlisted the help of a design maven who must remain nameless, a crossover wizard who thought laterally about the entire problem. He posited this: instead of desperately trying to reverse-engineer unavailable drivers, why not use readily-available, top-quality drivers and alter their characteristics - aided by a carefully-developed crossover - to ape precisely the sound of the LS3/5a?

Now this sounds like a tall order, but it should not surprise anyone who has fine-tuned an amp with changes in capacitors or valves, or sharpened up a speaker's behaviour with spikes. You can do anything you like to the sound of a component if you know what you're doing. And this guy knows what he's doing.

His mission statement was to achieve the same basic balance and timbre, at the same time improving the crossover quality with audio-grade polypropylene caps and inductors. This, and the new drivers, increased the power handling over the originals at low frequencies - a welcomed benefit regardless of your purism and adherence to originality.

He matched this crossover to a special version of the well-respected SEAS 5in bass unit - a driver found in not a few of the LS3/5a's modern rivals, but tweaked specifically for Stirling. It was a no-brainer: the overall dimensions - 'even down to the mounting positions' - are almost identical to the KEF B110. He told that, 'This is important to keep the off-axis response similar, as well as the unit's interaction with the cabinet. A special surface treatment is applied to the cone, again in order to bring the unit characteristics as close as possible to the B110. The unit has a suspension system that progressively controls cone excursion at high low frequency level, thus increasing the bass overload point.' Doug Stirling added that it also 'does the quirky thing at 1kHz that the KEF did.'

Same with the tweeter: the 'mystery' designer chose a ScanSpeak model, again devising a special version for this application. Both units were mounted in exactly the same positions on the baffle as the original LS3/5a, 'important for both dispersion and integration at the crossover point. It was even found necessary to create a cavity behind the baffle the same as that made by the frame of the B110, as this affected the response in the vital area just below the crossover point.'

Before a pair even reached us for consideration, the LS3/5a community was a'buzz with favourable noises. Key figures in the '3/5a counterculture had already paid real money for them, and it was blessed by no less than the Yoda of LS3/5a-dom: Paul Whatton. OK, so I was pre-disposed toward them before they arrived. But I was going to be swayed by emotion. So instead I enlisted two seasoned LS3/5a users to audition them with me, against a pair of 15ohm Rogers originals.

It took all of, oh, three seconds to see the smiles on their faces. While neither was in the market for speakers and both owned mint pairs, they gave it a total thumbs-up...with one expected proviso.

Even off axis, it was clear that the V2 delivered the sound of an LS3/5a, while at the same time exhibiting certain differences. On the plus side, it would certainly go louder, well past the area seasoned users consider to be safe for a normal LS3/5a. Moreover, despite measuring exactly like an LS3/5a, it seems to deliver more bass. Not a bump to mislead you, but a perception of greater extension. And, as we concentrated on a diet of female vocals, it exhibited the requisite warmth and sheer presence that makes the original so truly revered.

But equally, there were some areas to criticise, for which there seems to be an explanation. One thing that struck everyone who was involved with the original shoot-out was that the more use the speaker had, the better it sounded - in direct contrast to Stirling's concern about woofer surround aging. We have no scientific way of knowing what the 'run-in period' is for these speakers, but it was clear that they needed to loosen up a bit more to provide the delicious lower register flow and the sweeter highs of a well-seasoned LS3/5a.

Equally, there's also a case that simple aging - rather than actual hours of usage - can affect the performance, as I found out when I discovered a forgotten pair of 15 ohm Rogers that had been in storage for , and which proceeded to blow away everything else in my collection. Both listeners agreed - one quite emphatically - that the V2 had the disadvantage of being 'too new to assess fairly.' But even given that condition, both also agreed that, were they bereft of their KEF-driver LS3/5as, they'd have no qualms about replacing them with V2s.

Yes, they're that good, if a brand-new LS3/5a is what you crave. But is it worth considering the V2, when you can buy a second-hand pair of originals for £500 (if you hunt around)? I'd have to say 'yes'. Most obvious are two practical concerns: a warranty and spares availability. But, having used these for a while, I have to admit that 98 percent of the LS3/5a sound, plus an impression of deeper bass and audibly greater power handling (and therefore louder playback) are mighty tempting. It's as if the V2 rectifies the only two complaints any sane individual could level against the original. Which, by a certain logic, could mean to some that the V2 is actually than the original!

Now I don't want to see that as a 'pull quote' on web sites. I add emphatically that it's only better than an original if you worry about spares, fragility and maximum SPLs. Or if you like things virginally new. In which case, you should know that the standard walnut V2 sells for £758.00 plus VAT per pair - just under £900 - while crown-cut maple is £798 plus VAT per pair and pairs in birds-eye maple, rosewood and ebony veneer are £828 plus VAT. Note that the luxury veneers will be produced in series of only 98 pairs each, and Doug reckons that over a third are already sold. Which reminds me: he estimates, conservatively, that he will sell 200-250 pairs per annum.

A year ago, I wouldn't have believed that the dilemma of the LS3/5a would be solved so elegantly. How could I? The adamant refusal of KEF and Harbeth to respond to a certain, admittedly limited coterie of consumers pretty much nailed shut the coffin. This is no condemnation of their commercial savvy, although the latter's vehement hatred of this innocent little speaker might suggest to some the presence of deeper psychological issues. Suffice it to say, Doug Stirling and his designer have, through careful deliberation and experimentation, produced a surrogate that even the most hidebound LS3/5a purist could not fail to admire. But I wouldn't bank on it: audiophiles can be very strange, contrary people when you tamper with their idols.