Primary reason is three points define a plane. If a fourth point is added and is not in the same plane (very likely unless the surface is perfectly flat) then it either does nothing or more likely makes matters worse by not having a solid contact with the object. Now if you have adjustable spikes/cones then you can compensate for the fourth point being in another plane. Someone else will most likely have additonal information as to the pros/cons of three versus four good contact points.
It's simply to insure that the feet or supports for the component all make contact with the shelf. If there are 4 feet and one is not making good contact, this can cause the chassis to vibrate, resonate. I think it's stupid to make these out of turned brass or other exotic materials. If you have a really cheesy rack, then soft rubber or sorbothane feet might help...
Sonically, I doubt that it makes much difference. I prefer four adjustible corner points mainly because it makes a more stable platform - less likely to tip over (speakers). But three points (adjustible or not) are much easier to level, as Pmotz suggested.
I think it is more important to decide where to put than how many. If the purpose of using cones is to "drain" vibration than I would like to drain them as soon as possible. That means they should go directly under the transformer and CD drive because these parts vibrate most.
A more scientific way is to use a stethoscope. Just put a scope against the bottom of the equipment at various points and listen for noise. Place a cone directly under the noisiest point and place the others evenly to balance out the weight.
Spikes under speaker are a different story. It is much easier to make fine placement adjustments to a speaker with three spikes than with four.
For a number of years it has been thought that a pointed cone provides vibration control by "draining" unwanted stored energy out of a component and restricting the movement of vibration up into the bottom of the component due to the small contact area of the point against the shelf surface. Some people have called a pointed cone a "mechanical diode" allowing energy to transfer in one direction but restricting the energy flow in the opposite direction.
A closer look at the interface of the bottom of the component with the top of the cone, the point of the cone with the shelf surface and the material from which the cone is made and its shape will reveal the true nature of a cone's capabilities and limitations. A component placed on a normal shelf is subject to external vibration trying to enter from underneath due to the transfer of energy from the speaker through the floor and up through the equipment rack, plus the shelf itself contains additional resonance due to its vibrating in sympathy with the air-borne vibration from the speaker.
Placing the component on a cone will change the nature of the stored energy present in the component by altering its resonant frequency. The amount of change will differ from component to component due to the way the bottom plate of the component will be damped by the amount of surface contact with the top of the cone, the pressure of the cone against the bottom plate, the resonant frequency of the cone (which depends on its material construction and shape) and the position of the cone on the bottom of the component. That is why some cones sound better in some situations than others. It is really "hit and miss", especially when you consider that a cone is a rigid device. Being rigid, it is IMPOSSIBLE for it to really stop vibration from entering from underneath a component. In fact, it readily transfers vibration into the bottom of a component, it just does so differently than the original feet did. Additionally, raising the component further away from the shelf allows more acoustic energy to reach the bottom plate directly from the speaker. Using only three cones under a component (a popular practice) which reduces chassis "chatter" by allowing the three points to more easily define a plane so the component will sit evenly, allows two of the component's corners to dangle unsupported. This situation is not desirable since the chassis can now be more easily excited by air-borne vibration.
A cone may sometimes be better than using the original feet on a component but it cannot stop or absorb unwanted vibration or resonance.
In 20+ years of listening, I have always preferred 3 feet instead of 4. In fact, I have yet to listen to any component where 4 feet sounds better than 3.
I also prefer the solid brass cones of Walker Audio over any other material including, aluminum, steel, soft rubber, hard rubber, plastic, felt, inner tubes, etc. And I have tried them all.
For a rectangular component, I also feel that it is possible to balance the overhanging mass on three feet by varying the area and location of the support triangle. I have even written and copyrighted a computer program to calculate the optimum placement.
RE: my previous post - I should identify that I am a manufacturer of vibration control products.
Of course 3 is better IF a plane is maintained, inferring a very rigid chassis. If the damn thing's sagging or trampolining then more is better, but maybe it's hopeless at that point! (sorry)
Yes, cones transmit, not absorb, Their directionality should be selected according to which mass is the preferred recipient of said energy. Here's where Ken Lyon's comments on the Neuance on top of upturned spikes/cones would be really useful. KEEEEENNNNNN!....
My point (!) is that a cone is not exclusively a single direction transmitter of vibration. Being a rigid device it MUST transfer vibration from the shelf INTO the component that is resting upon it. The hard point is directly coupled to the shelf and the hard cone material cannot absorb vibration, it can only transmit it. If, as a cone enthusiast might claim, the point has such a small contact area with the shelf that it restricts vibration from passing up into the point, how can they also claim that the tiny point is able to "drain" the vibration so efficiently out of the component? Those two claims are mutually exclusive.
The much more important situation to consider is that the vast majority of vibration control products on the market (and also nearly all of the home-made variety) do not effectively address ALL three main sources of vibration that affect a component:
1) Vibration that is directly-coupled from the loudspeaker, and transfers through the floor and up through the component stand into the feet of the component.
2) Air-borne vibration sent directly from the loudspeaker drivers through the air towards the chassis of the component.
3) Self-generated vibration that is created within the chassis of the component by spinning motors, humming transformers and cooling fans.
The other significant sources of vibration are heating and air conditioning systems and exterior traffic (trucks, subways, trains, airplanes, cars, etc).
A component resting on the cone is immediately contaminated with vibration once any of the vibration sources are active. Attempting to "drain" the problem out is like closing the gate after the horse is out of the barn. The signal has ALREADY been contaminated and altered by the vibration and no amount of "draining" can undo the damage to the signal that has already been done. The most effective vibration control will be attained by thinking of the problem as a system and creating a method to place the component in an environment that eliminates as much of the vibration as possible BEFORE it can enter the component and affect the signal.
Disclaimer: I am a manufacturer of vibration control products.
Mr. Kohan is correct provided one looks at a cone footers function without regards to the substrate material that the cone's tip interfaces with.It's not actually the cone itself which provides the benefits and so-called "directionality" but rather the cone footer's filtering effects makes damping easier for the substrate by raising their frequency, combined with the high pressures applied to the substrate,which, *when optimised* will promote localised deformations of the substrate material beneath the cone's tip and subsequent energy losses via their conversion to heat.
Barry, Welcome to Audiogon - your post is like a breath of fresh air - it is clear and easily understood (and of course I happen to share your opinions).
Unfortunately, perhaps - size, shape, material of cones, location of component in room (rack vs floor), vibration environment (urban, suburban, country) are all part of the puzzle, IMO.
(yikes! Another isolation/vibration control co.)