# 8.1.4 Photometry

• The comparison of Gaia EDR3 photometry to external catalogues shows that issues are mainly seen in the bright end, i.e. brighter than 5-6 mag, and in $G_{\rm RP}$ for $G_{\rm RP}$ fainter than 18 mag.

• Comparison with GOG simulations (Section 8.3) shows that at high latitudes the values of $G_{\mathrm{BP}}-G_{\mathrm{RP}}$ forGOG20 and Gaia EDR3 are in good agreement, but not at bright magnitudes where the colour determination seems to be problematic; however, this discrepancy is slightly reduced compared to Gaia DR2.

• Results of comparison with CALSPEC are globally better than for Gaia DR2 except for $G_{\rm RP}$, which does not go through zero and confirms the magnitude dependence seen in the Hipparcos comparison. This is inside the expected uncertainties of the libraries and is not considered problematic.

• The pattern of the ecliptic that was present in Gaia DR2 for sky maps of sources with phot_bp_rp_excess_factor lower than $5$ is not longer visible, meaning that the background due to zodiacal light has been successfully removed.

• The phot_bp_rp_excess_factor values are improved with respect to Gaia DR2 mainly for faint sources

• Comparison between mean $G$ magnitudes provided in the Gaia DR2 and Gaia EDR3 catalogues for a sample of 140,635 sources classified as RR Lyrae stars shows that 2044 sources have a difference in magnitudes more than 0.5 mag, with the Gaia EDR3 magnitudes being fainter (Section 8.6, Figure 8.29). Most of these sources seem to be extended objects.

• Median differences between Gaia DR2 and Gaia EDR3 photometry in open clusters are very small (Section 8.7), i.e. -0.0087 ($G$), 0.0027 ($G_{\rm BP}$), 0.0074 ($G_{\rm RP}$), 0.0074 ($G_{\mathrm{BP}}-G_{\mathrm{RP}}$), with dispersions of the order of 0.01-0.02 in $G$, $G_{\rm BP}$, $G_{\mathrm{BP}}-G_{\mathrm{RP}}$ and lower in $G_{\rm RP}$. However, for a small sample of stars differences can be higher than 1.5-1.7 mag in $G_{\rm BP}$ and $G$, possibly related with the small number of observations ($<30$).

• Photometry in crowded fields (globular clusters, Section 8.7) is still worse than in non crowded regions, with deviations both in colour and in magnitude already present at phot_bp_rp_excess_factor $>2$.