Military manoeuvres come no more complicated. Merely considering a 'top secret' review involving a panel of seven or so listeners is to court disaster, as 'secrets' and 'journalists' are mutually incompatible. But we knew, as the only British hi-fi magazine surviving from 1972, that it was down to us to mark a momentous occasion: 25 years of the controversial, notorious Linn Sondek LP12. Linn, we knew, would be celebrating with the gorgeous, highly desirable, limited-edition LP12 bearing suitable anniversary cosmetics gracing our front cover. We, on the other hand, decided to ring in the changes by gauging the LP12's entire evolution in one searing, never-before-performed comparison test.

Arranged, under Linn's auspices, was the comparison of a Mk I, circa-1972, pre-upgrade, pre-killer-power-supplies LP12 versus the very latest incarnation. We would chart its evolution not bit by bit, an upgrade at a time (though that would have made an interesting book-length test), but by comparing the eldest with the youngest to learn what 25-years' worth of upgrades offered over the undiluted original. With the exception of the two decks themselves, everything else would be absolutely identical: arms, cartridges, phono stages, shared amplification and speakers. But we hadn't reckoned on the Seventies Linn disease and its Linn-gering symptoms. Which you could look at another way: at least that particular aspect of Linn ownership appears to have remained constant over a quarter century...

It started out with the inability of a Linnman to understand that this article was about the LP12 - not Ittok, not Kan, not Arkiv, not Ivor. What we required, just as in Olden Times when such things as comparative reviews of turntables existed, were nothing more than two LP12s representing the beginning and the present. Same arm, cartridge, ad nauseum. After much pleading, it was agreed if never quite understood.

To ensure that Linnman would not be able to whine and moan, it was stated categorically that:
(1) Linn would set up the turntables, be present during the listening sessions AND participate, even though the latter was an iffy proposition as they would know which turntable was on the left and which was on the right, through the screen semi-obscuring the decks;
(2) I would only collate the information and write the report rather than join the panel given that (a) the session needed a moderator and (b) I have been, am and always will be a Linn Skeptik;
(3) The panel would have Linn's approval.

We settled on the two Linnmen plus Editor Steve Harris (LP12 Owner), John Bamford of Pioneer (former hi-fi reviewer, editor and LP12 owner), Andy Whittle (speaker designer, Rogers), Paul Miller (reviewer and test equipment designer) and Mark Steadman (civilian audiophile and too young to have been part of the Linn Wars). The venue, after a number of disasters prevented the creation of my dream studio, was Steve Harris' home. The system? The two Linns were fed into a Musical Fidelity A-1000 amplifier driving a pair of Rogers Studio 7s.

Head Linnman either owns a cheap watch or firmly believes that the press is there to be harrassed. The sessions were scheduled for noon. He arrived at 1.30pm. With two LP12s 'still in the crate': no pre-mounted arms nor cartridges, ad nauseum.

The listening began at 4.30pm.

Our bile having settled by that time, the listeners rated eight pieces of music on each turntable, not knowing which deck was in use, with the following criteria on a grade of 1-10, with 10 being the best: Bass, Mid, Treble, Transparency, Dynamics, Stereo Imaging and PRAT (pace, rhythm and timing). In cases where some listeners chose not to assess a track or a category, the averages were adjusted, but none involved less than five listeners except for PRAT, where only four of the seven didn't feel that it was "A load of bollocks", "A load of old cobblers", "Crap" or "Wha?!?".

Unsurprisingly, seven listeners judging two turntables with six selections and seven categories generate a lot of data. Track by track and turntable by turntable, here's how they scored, with the averaged ratings reading left-to-right for bass, mid, treble, transparency, dynamics, stereo imaging and PRAT, followed by the averaged overall scoring.

1) 'Satin Doll', Bill Berry: (M&K RT101)
Turntable A (new LP12): 5.5/6.0/6.6/6.1/6.1/6.5/6.7 (AV = 6.21)
Turntable B (old LP12): 6.0/6.8/6.8/6.3/6.5/5.9/6.6 (AV = 6.41)
2) 'Cry Me A River', Ella Fitzgerald: (Classic
Records V64053)
Turntable A (new LP12): 5.8/6.4/6.2/6.0/5.7/6.5/6.6 (AV = 6.17)
Turntable B (old LP12): 5.5/6.5/6.0/6.0/5.5/6.4/6.5 (AV = 6.05)
3) St Paul Chamber Orchestra
(Sound 80 Direct-To-Disc)
Turntable A (old LP12): 5.5/6.2/6.2/5.8/6.1/5.8/6.0 (AV = 5.94)
Turntable B (new LP12): 6.0/5.9/5.9/5.7/6.0/6.5/6.7 (AV = 6.10)
4) 'Blues Power', Albert King: (Stax SXATS 1002)
Turntable A (new LP12): 6.2/6.8/6.5/6.8/6.5/6.7/7.7 (AV = 6.74)
Turntable B (old LP12): 5.3/6.1/6.0/6.1/5.8/6.5/7.0 (AV = 6.11)
5) 'I Can See Clearly Now', Ray Charles: (Crossover/Atlantic SD19142)
Turntable A (old LP12): 4.9/5.7/5.5/5.1/5.2/5.1/5.4 (AV = 5.27)
Turntable B (new LP12): 6.3/6.1/6.0/6.5/6.5/6.8/7.2 (AV = 6.48)
6) 'Jerusalem', Alpha Blondy & the Wailers: (Sterns Africa STERNS 1019) (Five listeners only because two couldn't bear it. Uh, Linnman chose it.))
Turntable A (old LP12): 5.5/5.6/6.2/5.8/5.0/5.0/5.0 (AV = 5.44)
Turntable B (new LP12): 7.0/7.0/7.4/7.4/6.5/6.6/6.5 (AV = 6.91)

Now let's look at the overall average scores for the two turntables:
Original LP12: 5.87
Latest LP12: 6.43

While there are those who'll question the validity or methodology, I'm dazzled by the consistency. For example, the old LP12 'bettered' the current version only once, and that was with the first track when the listeners were becoming acclimatised. Then there are anomalies, like the only track which actually had panelists withdraw from the listening was the one chosen by Linnman and which presented the most favourable performance from the current LP12. Other nasty little thoughts which occurred were: Did the Linnmen do a better set-up job on the current model than on the original? Since they set up the decks, could they tell which one was being played just by my movements, and did they score accordingly?

But those are churlish musings unworthy of this magazine. And besides: I'm not worried because the results, even given the desire to castrate an inconsiderate Linn employee, reflect what I heard. From my vantage point, and with the knowledge of which one was playing at all times, the latest LP12 was cleaner, clearer, more precise, better controlled and able to deliver deeper bass. The old one, however, wasn't massacred by the 25-years-younger descendent. It appeared a bit smoother, a shade richer, a bit less hi-fi.

But the real revelation is this: so close are the overall scores that Linn's designers and staff can be satisfied in the knowledge that they got it so right so long ago. By all means, make your way through the extensive upgrade path. But if you can't afford to uprate a vintage example, you can sleep just as easily as the owner of the 1997 vintage.

Linn Products Limited, Floors Road, Waterfoot, Glasgow G76 0EP, Scotland.
Tel: 0141 307 7777; Fax: 0141 644 4262.
Customer Services on Freephone 0500-888909
Web site: http://www.linn.co.uk

BOX NO 1: THE LP12 SYSTEM
Circa 1997, the complete top-of-the-range package consists of the LP12, a belt-drive, three-point-suspended sub-chassis turntable using a patented, low-noise, single-point bearing 'to eliminate noise and optimise information retrieval'. The unit features a solid wood plinth, stainless steel chassis, a beautifully machined 3.75kg platter and stable arm platform. It's driven by the Lingo power supply, which uses two low distortion sine waves separated by 90 degrees from two high voltage amplifiers, the sine waves derived from a low noise crystal oscillator. A special circuit monitors the motor and during start-up drives the motor hard enough to set the platter rotating; when speed 'lock' is achieved the power is reduced to a level at which the motor runs almost silently. It was introduced in 1972 as a 33 1/3rpm-only player, but has since been made capable of playing proper singles.

Both decks were fitted with Ekos tonearms, dynamically balanced and featuring temperature-compensated precision-springs ensure constant tracking and bias forces, ultra-low friction bearings and 'clean room' assembly to guarantee performance and longevity. All of the main components are machined from solid, with advanced alloys and adhesives employed to maximise strength and mechanical integrity. The Ekos is fitted with low-loss cable and gold connectors and an easy-to-use hydraulic lift/lower device. Effective mass is 11.5g This arm has been in service since 1988.

At the business end was the Arkiv moving coil cartridge, feeding the brand-new Linto phono pre-amp. The Arkiv was introduced in 1992; we heard its £1200 successor, said to be "a significantly enhanced version". Tracking has been improved and the output is higher than that of the original Arkiv. The new Arkiv features a revised stylus assembly, with a more highly developed line contact stylus shape attached to a super rigid boron cantilever. Constructed around a solid alloy chassis for its extra rigidity, Arkiv contains a damped suspension mechanism supporting hand-wound coils, terminated with gold contacts. Tracking force is 1.8-2.0g, the coil impedance 4 ohms, the recommended load >50ohms and the output 0.4mV, but none of these concerned us as the Arkiv was fitted to the arm for which it was designed, and fed into a phono pre-amp conceived to match it perfectly, thus ensuring that the review system was fed precisely what the Arkiv could deliver.

The Linto Direct-Coupled Precision Phono Preamplifier's input stage takes the cartridge signal directly into the Linto's amplifying transistors. As such, the phono signal sees neither resistors nor cartridge loading networks sitting in the signal path. The Linto is powered by the 'Brilliant Switch Mode silent power supply', proprietary power supply technology which, with extensive internal shielding, ensures sonic integrity. Signals were fed into the Musical Fidelity A-1000 line inputs and level matched by Paul Miller a track at a time.

By the way, an LP12/Lingo/Ekos/Arkiv/Trampolin retails for £4375.00. I think that for an original LP12, you paid just the last two digits.

BOX 2: LP12 HISTORY AND UPGRADE PATH
Below is the factory-authorised list of retrofittable upgrades and manufacturing changes which delineate the differences between the two turntables in our comparison. The latest upgrade, the Linto Phono Preamplifier, is discussed in the main text.

UPGRADE YEAR S/N (approx.)

1974 2,000
Changed from two buttons to a single switch with
mains neon.

1974 2,000
Motor control PCB changed from terminal strip
to small circuit board.

1974 2,600
Strengthened by addition of strap, spot-welded
in place.

1978 23,000
Two holes added for 6 x 0.5 self tappers into
wood block.

1979 27,000
Lid support prop removed and hinges changed to
spring-loaded, self-supporting type.

late '70s n/a
Added strengthening bar, spot-welded in place.
Increases rigidity and strength.

1981 32,800
Improved material specification of many integral
mechanical components. Improved suspension
and stability.

1981 n/a
Manufactured to tighter tolerance; ground top and
bottom. (changed from zinc to black)

1982 38,800
Electronic speed control with a sophisticated crystal
controlled power supply. Isolates the rotation of the
turntable motor from variations in the electrical supply.

1984 53,000
Enlarged corner blocks. Strengthens and increases
rigidity of plinth.

1984 54,100
Attached strengthening bar with epoxy glue.
Superior bond; increases rigidity.

1986 n/a
Further tightening of manufacturing tolerance.

1987 70,000
Improved bearing liner material and thrust pad
specification. Bearing liner machined to lighter
tolerances, thus creating better speed stability.
Changed to black oil.

1987 79,150
Material changed to MDF core, laminated top and
underside. Increases rigidity of armboard, thus
creating improved platform for tonearm.

1988 n/a
Tightened grinding tolerance.

1989 79,700
Motor thrust pad replaced with stainless steel ball
bearing. Reduces motor noise.

1989 81,000
Composition changed to a new harder, denser rubber.
Improves performance of suspension.

1990 N/A
Direct-coupled power supply for the LP12. Offers low-
noise oscillators, precision filtering, separate drive for
both phases and eIectrical isolation from mains supply.

1991 87,047 (Valhalla)
Small cap factory-fitted (glued in position) to reduce 87,206 (Lingo)
noise level.

1991 87,672
Replacing hardboard base.

1991 N/A
Suspended base board available as an upgrade.

LP12 sold as a mechanical assembly only. Three 1991 87,672
power supply options available: Lingo, Valhalla and
Basik. (Basik PSU supplied with 45rpm adaptor).

1992 88,950
Fitted with additional stud which improves the
coupling of the top plate to the plinth and secures
the motor corner.

1993 90,582
LP12 Cirkus bearing and subchassis upgrade
fitted as standard.

BOX 3: MORRIS DANCING

Unable to accept that making seven professionals wait a tedious four-and-a-half hours was entirely his fault, but utterly convinced that his foul-up caused an invalid result due to the obvious loathing of him evinced by the listening panel, Linn's Brian Morris immediately fired off an e-mail as some warped form of damage limitation. It was unnecessary, as the newer Linn outperformed the original quite markedly - which is what he obviously wanted; the only surprise was that the original wasn't 'slaughtered' by the latest edition. But for an insight into the mind of an audio psychotic, one who firmly behaves as if it's 1979 and the Linn disease is still virulent, we thought it would amuse you to read his bleatings, especially as he fails to understand that the perceived extra loudness of a new LP12 over an original is a quantitative rather than a qualitative gain. And we were only interested in sound quality, not playback levels:

"1. Paul Miller has confirmed to me that the volume differences between the 2 amplifier turntable inputs was only 0 .25db which is significantly smaller than the perceived loudness differences we all heard when listening to the current LP12, compared to the earlier version. This greater perceived loudness of the current LP12 is due principally to the Cirkus and Lingo upgrades in addition to other modifications and upgrades incorporated during the LP12's lifetime. This perceived loudness is a characteristic of the current LP12 which benefits from:

1. a lower noise floor
2. greater information retrieval
3. greater dynamic range
4. more music for the listener
5. greater speed accuracy = more tune
6. being a better turntable

"As we were demonstrating the real differences between the 2 turntables, the listeners should have had the opportunity to listen to music via each turntable in the way the respective LP12s retrieve and communicate to the listener and importantly at the same level of volume set by the amplifier.

"This never happened in the review methodology. Instead, the immediate and most obvious difference (perceived loudness) was obscured in the test by equalising the amplifier/system levels and reducing the perceived loudness of the current LP12. Moreover, since a difference of 0.25db between the inputs/turntables is hardly a noticeable difference to the listener, the doctoring of system volume to equalise the two turntables actually had to have been greater than the equivalent of 0.25db.

"Since the review was set up to compare the real differences between the 2 turntables, each turntable should have been played at its respective level since the perceived loudness of the current LP12 is greater than original for reasons outlined above. As Paul Miller has established that there was little difference between the 2 input signals (0.25db), then an important difference between the two turntables was disguised in the test and should have been apparent for all to hear each time a track of music was played.

"Unfortunately, the listening panel was not given the opportunity with each piece of music to accurately compare the most apparent difference between the 2 turntables and appreciate the inherent musical benefits in this way. In this respect the review methodology was flawed, since the review approach disguised this immediate difference of perceived loudness (components 1-6 above).

"One interesting difference I found between the two turntables worth reporting in the review relates to the Ray Charles track. Played on the early LP12, I found that Ray Charles was clearly mistakable for Joe Cocker. When the same track was played on the current LP12 there was no doubt that it was Ray Charles singing.

"I thought your attempt to enquire into who reported what on their score sheets to be very much against the spirit and methodology of the review i.e.- since the ratings were supposed to be anonymised (sic). However, as stated in my last email to you, the fact that proportions of a small sample of people appeared to prefer both the new and old LP12 does not raise the possibility that Linn has been "dicking" its customers for the past 25 years as you put it. The current version of the LP12 is a far more accurate and revealing reproducer of music compared to earlier versions of the product, presenting music on vinyl in a way which the review methodology obscured in the comparison.

"I suggest we revisit this review and demonstrate the musical differences between the 2 LP12s as originally intended, allowing the perceived loudness of the current LP12 to deliver all the associated musical differences and benefits. Had this happened last week, the actual review comparison would have been more accurate and interesting.

"PPS: Moving volume controls on standard potentiometers will never track left to right within 0.25db and that is why we never use them. Had we used a "hi-end" system anyway as you suggested when we first discussed this review last year, we probably would not be dealing with these issues at this stage."

Final thoughts: All of the panelists were dying to see who wrote what after the sessions, not I. And note the dig at Rogers and Musical Fidelity. One can only assume that to satisfy Morris' idea of a "hi-end" system. It would have to be all-Linn. Again, he misses the point...