Marble or Granite shelfs in a hifi rack?

Well, then there's always my approach: I invested about 1K in my Salamander Synergy Twin 40 solely for the reasons of convenience in having adjustable shelf heights, enough space for all my components, and the ease of system reconfiguring that comes with having gotten the Saturn caster option. The shelves are all MDF, the rack itself is fairly flexible and resonant by high end standards (though much better than some more makeshift solutions or mass-market stuff, and heavy as a bear), and way too tall for high rigidity anyway, and of course it isn't positively coupled to the floor, though the fact that the casters employ rubber treads does provide a little isolation.

Frankly, in my circa-$15K system, I've never heard enough differences from various support strategies with most components to place a higher emphasis on this subject. I do own some cones and isolation footers, finding the Iso-Bearings and such mildly useful with certain gear, and don't use the cones. I keep my TT on an original Symposium shelf, which in conjunction with sorbothane pucks underneath, does a very nice job of isolation from energy transmitted through the rack. But I'll tell you what, I'm lovin' life these days with all my gear on one rack and total accesss from all sides (no doors, sides, or rear panel installed) - I probably roll the rack out and make changes (most temporary) around back a few times a week on average, now that this is so quickly and painlessly accomplished. I have decided to replace all the regular corner shelf support brackets with the heavy-duty ones, which employ steel struts that runs front to back between the vertical pillars, for the addtional reinforcing effect they will have on the overall rigidity, but aside from that, I've experienced no objectionable problems in use so far, and I think I'm done with this aspect.

Could it sound better? What do I care - it sounds good enough for me to just relax and enjoy the music, and the ease of use I get with being able to raise or lower shelves and move the whole thing around at will. I personally doubt that any incremental sonic improvements I might get from compulsively obsessing over a more audiophile-approved and tweaked-out arrangement would be worth the candle to me in terms of livability lost. (Oh, and I also don't believe in the concept of "mechanical diodes", self-vibration being more of an issue than acoustically-induced vibration, or the existence of "PRAT", but those are probably issues best left for other threads! :-)

BTW, the preceding doesn't mean I'm in love with MDF any more than I implied two posts ago, just that although I have mused about experimenting with replacing a shelf or two with something else and seeing what turns up, so far I haven't felt compelled to go to this length. The rack probably doesn't justify it anyway. (Ignorance is bliss and all that... :-)

Albert, I did not like the Neuance on an aluminium rack either. I have played with aluminium racks with only a little success, finding that if you have a concrete slab floor, either an aluminium rack, or a steel rack filled with fine sand were possibly better, whereas on a wooden floor a welded steel rack, unfilled, seemed always better.

You are quite right to say there are no universal truths. In particular, turntables, CDs and valve preamps differ in the ways that they are constructed to deal internally with vibrations and so can respond differently to different supports. But with solid state preamps, and amps of any persuasion, I have found the steel rack and Neuance, plus EAR footers has been superior for PRAT and neutrality in several systems. What I believe is much less universal is the degree to which we audiophiles seek PRAT, or agree on what it is. I have had no experience of any heavy shelf approaching the PRAT of a lighter shelf.

I have to agree with Red on this. I find it interesting that the Spectral user finds the Neuance shelves to accentuate the high frequencies. I suspect that what's happening is that the absence of smear and lower-frequency resonances results in a "tighter" (better PRaT, more coherent) presentation that shows the leanness many of us believe the Spectral stuff is voiced to have. I would suggest that room placement and choice of speakers be used to balance a Spectral system, rather than using "muddying" supports.
It IS true that introduction of Neuance shelves under my CDP (EMC-1 MkII) and pre (Aleph P) helped to heighten PRaT, and that quickness and light-footedness can sound leaner as other components' grainniness or edginess is brought into focus. The Neuance allowed me to hear the difference between cables, ICs, and PCs more clearly, and is now allowing me to use a neutral and very fast cable (SPM), for example, as ALL these links become clearer, faster, and less phasey. Neuance really did allow me to "neutralize" any effect of an underlying base material, when set on upturned spikes. I find this aspect more appealing both scientifically and musically than trying to "marry" one of a myriad of base materials to my system's perceived flaws, imbalances, or "character". The Neuance has allowed a predictable and satisfying evolution of my system without a misstep. If I want to change spectral tilt I'll play with speaker/room issues, NOT support hardware. I know I sound like a devotee here, but I love elegant, successful, cost-effective technical solutions to complex problems! Cheers.

Subaruguru, I suspect that what you are saying and what I am saying are in agreement. I am familiar with the prat, lightness, and also leaning out effect of Neuance. I believe, without supporting measurements to say so, that Spectral/MIT is ALREADY more time and phase coherent than most other designs. Neuance helps to move other equipment toward the same coherency and resolution by eliminating smear of the signal. The leaness of the signal that you speak of with Neuance, and that you also attribute to Spectral, may be an inherent result of tightened time/phase coherency.
This is not our understanding of live music. Fullness, bloom, continuity are normal in live music but we are discussing recorded cd's, wherein an entire chain of recording and playback electronics have acted on the signal. It is possible that tight coherency in all aspects of playback equipment can be carried too far, to the point that low level distortions, filter artifacts, a/c line noise, transformer artifacts, and digititis in cd's may be resolved and amplified.
Also, I mention Spectral/MIT as an entity. Neuance has the greatest effect on power supplies, preamp, and a/c filters, The first group of components may be more subject to transformer artifacts than the latter.
Changing speaker placement, speaker choice, or room treatment won't eliminate a need in the end to deal with vibration and a/c. I'm speaking about small refinements to a system whose character I already enjoy and prefer.