Comparison of analytical noise levels (y axis) with systematics ${\sigma }_{\mathrm{C}\mathrm{C}\mathrm{F}}^{\mathrm{F}\mathrm{C},\text{fund}+\text{syst}}$ (top) and without systematics ${\sigma }_{\mathrm{C}\mathrm{C}\mathrm{F}}^{\mathrm{F}\mathrm{C},\text{fund}}$ (bottom), against empirical noise levels in simulated data (x-axis) from spatial standard deviations ${\sigma }_{\mathrm{C}\mathrm{C}\mathrm{F}}^{\text{spatial}}$ (top) and ERR extensions ${\sigma }_{\mathrm{C}\mathrm{C}\mathrm{F}}^{\text{ERR}}$ (bottom). Each point is a given separation, for a given band and data set. The simulated data considered are those used in Mâlin et al. (2023) for HR8799 (with K_* = 5.24 and t_exp = 171 min per band) and β Pictoris (with K_* = 3.48 and t_exp = 93 min per band).