Step 5. Spectral form, filters, and a first bound on \(S_\Phi\)

Next we turn the general visibility kernel from Step 4 into a frequency-domain formula that’s easy to use with real data. If the lapse fluctuations are stationary (statistics depend only on time difference), their two-time correlator \(C_\Phi(t-t')\) has a one-sided power spectral density \(S_\Phi(\Omega)\). Then the visibility becomes a weighted integral of that spectrum by a filter \(|F(\Omega)|^2\) determined solely by your pulse/sequence \(f(t)\):

\[ -\ln V \;=\; \frac{\omega^2}{2}\int_{0}^{\infty}\!\frac{d\Omega}{2\pi}\;S_\Phi(\Omega)\,\big|F(\Omega)\big|^2, \qquad F(\Omega)=\int f(t)\,e^{i\Omega t}\,dt. \]

This is Eq. (8) in the paper. You pick \(f(t)\) (Ramsey, Mach–Zehnder, spin-echo…), which fixes \(|F|^2\); the integral then tells you how much dephasing to expect for a given \(S_\Phi\).

Concrete filter example: For a simple rectangular interrogation of duration \(T\) (Ramsey/Mach–Zehnder), \(|F(\Omega)|^2=\!\left(\tfrac{2\sin(\Omega T/2)}{\Omega}\right)^{\!2}\). That sinc² envelope shows which \(\Omega\) contribute to dephasing (peaks near \(\Omega\!\sim\!\pi/T\)).

Markovian limit and a back-of-envelope bound: If low frequencies dominate so that \(S_\Phi(\Omega)\approx S_\Phi(0)\), then the dephasing rate is \(\Gamma_\phi \simeq \tfrac{\omega^2}{2}S_\Phi(0)\) and for an interrogation time \(T\):

\[ -\ln V \;\simeq\; \tfrac{1}{2}\omega^2 S_\Phi(0)\,T \;\Rightarrow\; S_\Phi(0) \;\lesssim\; \frac{2\,|\ln V|}{\omega^2 T}. \]

The paper gives a scale: with \(T=1\,\text{s}\) and an optical line \(\omega/2\pi=10^{15}\,\text{Hz}\), a 1% (0.1%) visibility loss implies \(S_\Phi(0)\lesssim 5\times10^{-34}\,\text{s}\) (\(5\times10^{-35}\,\text{s}\)).

Units & conventions checkpoint: They use one-sided \(S_\Phi(\Omega)\) with units of seconds (s) when integrating with \(d\Omega/2\pi\); equivalently, with ordinary frequency \(f\) (Hz), \(S_\Phi(f)\) carries units Hz\(^{-1}\). \(|F(\Omega)|^2\) has units s², making \(-\ln V\) dimensionless, as it must.


Glossary (for this step)

Symbol Name Meaning (units) Typical value/example Metaphor
\(S_\Phi(\Omega)\) Lapse PSD One-sided spectrum of \(\delta\Phi\) (s) Bound via \(-\ln V\) integral "Noise color of time"
\(F(\Omega)\) Sequence filter FT of \(f(t)\); \(|F|^2\) weights \(S_\Phi\) (s²) Rectangular: \(\left(\frac{2\sin(\Omega T/2)}{\Omega}\right)^2\) "Sieve selecting frequencies"
\(\Gamma_\phi\) Dephasing rate Visibility decay rate (s\(^{-1}\)) \(\Gamma_\phi \!\simeq\! \tfrac{\omega^2}{2}S_\Phi(0)\) "Blur-speed of fringes"
\(S_\Phi(0)\) Zero-freq level Low-\(\Omega\) plateau of \(S_\Phi\) (s) \(\lesssim 5\times10^{-34}\) s (1% loss, 1 s, optical) "DC fog of time"
\(\omega\) Clock frequency Internal line frequency (rad·s\(^{-1}\)) Optical: \(2\pi\times10^{15}\) Hz "Metronome speed"
\(f(t)\) Path indicator +1/−1 sequence defining arms/pulses Ramsey/MZ/echo patterns "Which road when"