Fig. 4: Phase-dependent two-photon interference.

a Cross-correlation traces measured with the setup shown in Fig. 1b for three different excitation intensities: \(\bar{n}=0.0062\) (left), 0.25 (middle) and 0.62 (right panel). b Measured (solid symbols) and theoretical (solid lines) coincidence probabilities at Δt = ±τ delays, normalised to the coincidence baseline (dashed line). We use \(\overline{{{{\mathcal{C}}}}(\pm \tau )}=\frac{1}{2}\left({{{\mathcal{C}}}}(+\tau )+{{{\mathcal{C}}}}(-\tau )\right)\). c Normalised experimental (solid symbols) and theoretical (solid lines) coincidence probabilities at Δt = 0. Experimental data in b and c are extracted from data in (a). The theoretical results are fitted using maximum likelihood estimation. Fitted parameters \(\{{p}_{0},{p}_{1},{p}_{2};{M}^{{\prime} }\}\) corresponding to different excitation fluxes \(\bar{n}=0.0062,0.25,0.62\) are {0.98, 0.023, 8.0 × 10⁻⁶; 0.96}, {0.69, 0.30, 2.2 × 10⁻³; 0.94} and {0.49, 0.50, 8.0 × 10⁻³; 0.92}, respectively. A fixed value of M = 0.89, extracted from the plateaued ∣g⁽¹⁾∣ = 0.946 shown in Fig. 2b through M = ∣g⁽¹⁾∣², is used for photon indistinguishability.