Stop 2 of 9 · How much of each axis can be resolved?
Light share and dark share — what each axis resolves
What a pair of observables resolves is its light share: the squared cosine between the two fluctuation vectors. What it cannot resolve is the dark share, the Robertson defect over the product of the uncertainties. They add up to one on every axis. A closed axis of three positions at its maximal current state resolves 4/5 and keeps 1/5 dark; as the ring grows both tend to 1/2.
At the maximal current state, the c of the closed axis, three positions resolve 4/5 and keep 1/5 dark; the open axis with three positions keeps 0. For larger rings the share is not monotone, it goes to 1/2 from both sides (RingSzego). What the window needs is not a value but c > 1: at the ring limit c = √2 it is 0.1444, positive because the infinite limit does not saturate the Gaussian (angle 45°, not 0). The first two plots are computed live from the matrices of the ring; Ωb is the measured datum of each instrument and the spread between fronts is the error of each instrument.
What is exact here
- Lean Light + dark = 1 on every axis, for every state.
- Lean In the band every axis keeps a positive dark share, entangled with the others or not.
- Lean At the maximal current state each axis resolves 4/5 and keeps 1/5 dark; the open cube of three keeps 0.
- Lean As the ring grows, light and dark both tend to 1/2 and the angle to 45°.
- Browser The curves for each size are computed live; the measured Ωb values are the instruments' data.
Theorem trail
| Statement | Name |
|---|---|
| light + dark = 1, every state | TorusDark.lightShare_add_darkShare |
| dark share > 0 in the band | TorusDark.darkShare_pos_of_band |
4/5 and 1/5 at c | TorusDark.lightShare_maxState, TorusDark.darkShare_maxState |
| the limit of the ring: 1/2, 1/2, 45° | RingSzego.tendsto_lightShare, RingSzego.tendsto_darkShare, RingSzego.tendsto_angle |