Shear measurement residuals after calibration and binned shear biases as a function of galaxy properties. Top: shear estimation on data described in Fig. 5 after NN calibration. Each point is one “case” with 2000 “realizations” sharing the same axis ratio, size, magnitude, PSF, and shear, but differing in orientation. Middle: multiplicative bias (shown in dark cyan) and additive bias (displayed in blue-violet) are presented as a function of the true galaxy magnitude. The data points from the top panels are categorized into six bins based on the magnitude and fit to a linear function. The y-axis is plotted on a logarithmic scale. The lighter shade corresponds to ±2 × 10⁻³, while the darker shade represents ±2 × 10⁻⁴. Bottom: similar to the middle panel; the m and c are shown as a function of the true galaxy S/N.