Hi Doug, while ALL instruments, including the human voice, require perfect speed stability to be reproduced correctly, I find speed instablities are most clearly heard (a different issue) in the decay/resonance of instruments, which being but the resonance of a note already struck, simply decay unwavering until vanished, and nothing decays/resonates audibly like piano. The piano keys are hit, the strings vibrate, the box resonates, and any speed variation is CLEARLY audible and exposed, as decay/resonance CANNOT waver in pitch. Bells are good too, as are acoustic guitars, echoes of various sorts, and so on. I don't get my musical enjoyment from hearing evidence of perfect speed, but if I hear a perfect decay, then I know my playback system is spot-on. Besides, hearing a wavering resonation/decay of any sort is annoying and worrying. All that said, a 'table which gets the speed spot on, and sustains perfect speed in real-world conditions - i.e. while playing a record - not only captures the PRaT which should exist in every recording regardless of quality, but also does a better job of retrieving detail, imaging information, energy (dynamics) and frequency extension.
As to whether or not the Linn shows evidence of better speed stability than a well-made and massive unsuspended deck, this is a complex issue. The Roksan turntable has a motor laterally mounted to a spring which allows the motor to rotate with the belt to prevent belt-stretching, thus pulling off the trick of having greater snap and speed, traditionally, than a Linn. On the other hand, these high-mass decks, if they use a rubber belt of some sort, have no such compensation. Regardless of the mass of the platter, the belts will stretch, and the greater mass of the platter ends up not overcoming stylus drag totally so much as reducing it in frequency, as the belt simply cannot react as quickly while dragging such a large mass. It will, however, react at some frequency. But the rhythm lies in the lower frequencies, and if the belt-reaction to stylus drag is lowered to this frequency, then PRaT will suffer, and the Linn, with a lower-mass platter, come out sounding like it has superior PRaT. Thread-drives are a different matter, as it is difficult to predict exactly what is going on here: they are not stretching, but are they slipping? And in what way are they slipping? Perhaps, like rubber belts, they slip precisely when stylus drag is at its worst, which is logical, which in the end means they have some of the same problems rubber belts have, in racing at precisely the point there is greatest energy in the music. Finally, the Linns are suspended, and the subchassis does react to stylus drag and belt reaction in a way similar to the Roksan's motor's lateral freedom: the Linn's entire subchassis twists in reaction to the stylus drag, thus mitigating belt-reaction, and reducing damage to the PRaT. But, while the Linn saves the Prat, it sacrifices the detail higher-mass decks, which have more stable speed overall (above the low frequencies), traditionally excel at. Of course, all this is mere theorizing, in the wake of my experiences comparing the sound of various high-end and classic decks to such high-torque designs as the big idler-wheel decks (like, of course, the Lencos) and the Technics SP10 MKII (which is discussed at length recently on VA).
In the end, it's what makes you happy that makes you happy - providing you are not deluding yourself as so many in fact do (I have a detail-freak friend up here who always goes for detail, and upgrades at least once a month at great cost, indicating deep-rooted dissatisfaction). So good to see the vinyl industry doing so well and generating so much interest! I hope the music-fest goes well for all, Vive la Vinyl!
