OK, a simple attempt at a four body Kepler

I take four masses, pair then up :

Mass pairs (1,2,3,4) , taken two at a time
2  3  4  6  8 12

My signal, counting radii up by some small increment and computing gravitational force (G=1), totalled across the six radii. So I compute a signal of 1/r^2 dividing each of those mass pairs. Assume a total energy. Play around with variable until I get a smooth spread of quant ratios.

Compute SNR for a total energy E if the noisecounting radii by d in N steps

My signal:
Sig <- function(d,E,N) {
     foo.sig = NULL
     for(i in 1:N){
         foo.sig[i] = sum(m/(i*d)^2)/E +1}
     foo.sig
 }

Here is the SNR +1. These are the bauds, some eight of them that count out six radii in steps. I have really done a spectral decomposition. The result tells me that those signals divided by that number E, is best counted with these bauds. Nothing more.

So I selected a possible function. There were six radii so I suspect they counter out powers of 2 modulo six. So I count out the function, using these quants in an eight digit sequence.  and generate the signal of six radii counting.  My counter was my suspected function is counter squared modulo 6. I was sure to do this:
2^(B[j]*k)/(2^k), divide by the bit power, to normalize the digits to an eight bit twos; I divided by the bit power.

I got this plot.

Six radii counting out a quadratic form.  Is it useful? No, not without some better knowledge of Kepler.  If I set two quant rates, and counted out a polar R, and a theta, then picked the ellipse as a possible function, then yes, useful.  As for this thing on the right, it could just as easily been someone juggling six balls in the air.

But here is the point.  If we know the generating force and the total energy, then we can find the best quants before applying some law of physics in the Schrodinger equation. Do the Schrodinger last.  Do group separation first.