Left panel: comparison of the predicted amplitudes for the DoG-HIT ring-only image (red markers), the DoG-HIT image with additional Gaussian component (blue markers), and the self-calibrated visibility amplitudes (green markers) as a function of projected baseline length. Right panel: the difference between the predicted visibilities for the DoG-HIT model and DoG-HIT with the additional Gaussian component as a function of projected baseline length. The plot demonstrates that the additional Gaussian component has no sizeable effect on baselines longer than PV-NOEMA.