Even listening through phones, the needle riding in the groove is always a constant source of vibration dealt with by the turntable. If the table is not supported in such a way that is compliant (to isolate the table from outside inputs) but also well-damped, then the effects of this self-generated vibration will linger longer and cause more subsequent interference at the stylus/groove interface, with audible effect on the reproduced sound.
Rigidly spiking the chassis to a sturdy rack might also work well in this regard as long as you listened through phones, but listening through the speakers on a springy floor with a so-so rack, damped compliance is the way to go.
What will work best is kind of a hit-or-miss affair, the system Q depending as it does upon the mass of the suspended table (and shelf if used) and the compliant and damping properties of the chosen supports (and their distribution and number used), not to mention the properties of the rack and floor. The whole interaction is probably too complex to predict usefully, especially given that the exact properties of the supports in a state of compression are likely an unknown, so trial and error to hear what works well is generally the easiest way to go about things here.
My own experience has been that compliant footers which are broadly shaped (not spherical or conical, though a very shallow hemisphere seems OK) and of relatively low height, made from fairly dense and soft (somewhat squishy, flexible) material that exhibits a 'slow' rebound character (is not bouncy), and used in numbers that keep each support under only medium compression (I use five 2" diameter pucks) do the trick.
The stock 1200 threaded feet are useful for the necessary task of leveling the table. So what I do is keep the feet on, but place whole shebang atop an inert shelf that is isolated from my rack by the compliant feet (my shelf is a Symposium and the feet are FoculPods; my rack is a Salamander Synergy that is also on casters and is not the most rigid thing going). I have thought about experimenting with removing the stock feet and replacing them with brass cones having the appropriate threaded posts to still work for leveling purposes - thereby coupling the table's chassis to the nonresonant shelf, which might function effectively as a vibrational sink, and eliminating a potential source of phase-reinforced interaction between the undoubtedly differently-tuned resonances of the two sets of compliant footers. But I haven't done it yet.
Rigidly spiking the chassis to a sturdy rack might also work well in this regard as long as you listened through phones, but listening through the speakers on a springy floor with a so-so rack, damped compliance is the way to go.
What will work best is kind of a hit-or-miss affair, the system Q depending as it does upon the mass of the suspended table (and shelf if used) and the compliant and damping properties of the chosen supports (and their distribution and number used), not to mention the properties of the rack and floor. The whole interaction is probably too complex to predict usefully, especially given that the exact properties of the supports in a state of compression are likely an unknown, so trial and error to hear what works well is generally the easiest way to go about things here.
My own experience has been that compliant footers which are broadly shaped (not spherical or conical, though a very shallow hemisphere seems OK) and of relatively low height, made from fairly dense and soft (somewhat squishy, flexible) material that exhibits a 'slow' rebound character (is not bouncy), and used in numbers that keep each support under only medium compression (I use five 2" diameter pucks) do the trick.
The stock 1200 threaded feet are useful for the necessary task of leveling the table. So what I do is keep the feet on, but place whole shebang atop an inert shelf that is isolated from my rack by the compliant feet (my shelf is a Symposium and the feet are FoculPods; my rack is a Salamander Synergy that is also on casters and is not the most rigid thing going). I have thought about experimenting with removing the stock feet and replacing them with brass cones having the appropriate threaded posts to still work for leveling purposes - thereby coupling the table's chassis to the nonresonant shelf, which might function effectively as a vibrational sink, and eliminating a potential source of phase-reinforced interaction between the undoubtedly differently-tuned resonances of the two sets of compliant footers. But I haven't done it yet.