Figure 1

(a) Schematic of a QED system. Vacuum Rabi oscillations can occur if the atom-photon coupling strength $g$ overwhelms the loss rate $\kappa$ of cavity photons and the emission rate $\gamma$ into other modes. (b) Normal photon counting: Periodic ultrashort laser pulses excite the atom faster than all other time scales. The single-photon detector records the arrival time $t_n$ of a photon decaying out of the cavity with respect to each pulse. The gray area in (b) means that no photon is detected because of detector inefficiency. (c) Time-adjusted photon counting: The time (shown in green) between a photon-detection event (in red) and the subsequent laser pulse is eliminated, moving the time of each laser pulse to the time of the previous photon count. Any periods (in gray) of no photon detection are eliminated. In this manner, the system has an effective instantaneous feedback: Once a photon is detected, a laser pulse immediately drives the atom to the excited state. (d) The zero-frequency component of the power spectrum $F(0)$ is positive in the presence of coherent VRS, even if the coupling $g$ is very small. In (d), the color brown is maximum, blue is minimum, and the black vertical lines show $F(0)=0$.Reuse & Permissions

Figure 2

A violation of the extended LG inequality [Eq. (29)] for typical parameters in single-photon cavity-QED experiments [2, 3] using time-adjusted photon statistics. The parameters used here are $\kappa /2\pi =2.7$ MHz, $g/2\pi =10$ MHz, and the variation of the detuning $\delta /2\pi$ can be up to 20 MHz. The violation of the inequality is indicated by the gray island inside the black contour line.Reuse & Permissions

Figure 3

(a) The nonadjusted photon-counting statistics also exhibit a violation of the LG inequality [Eq. (29)]. The combined effect of the atom polarization decay and the cavity decay rates ($\gamma /2\pi =3$ MHz and $\kappa /2\pi =2.7$ MHz) prevents a violation being seen in the data in Ref. [1]. However, a slight decrease in the polarization decay rate $\gamma$ or an increase in the coupling strength $g$, combined with the ultrashort laser pulses in Ref. [1], should reveal a violation of the inequality [Eq. (29)]. (b) Using the same model (without time-adjusted counting), one can estimate the ratio between the coupling strength $g$ and the dissipation ($\kappa$,$\gamma$) needed to observe a violation of Eq. (29). This is approximately given by $\mathrm{VRS}=4g/(\kappa +\gamma )>4$. In (b), we have chosen $\kappa =\gamma$.Reuse & Permissions