Comparison of the cleaned sample of likely WDs, ﹩{N}_{{ \mathcal C }}({M}_{G},(B-R))﹩ (left, with units [mag⁻² ]) with the modeled white dwarf LCF, Φ₀(M_G, (B − R)) (center, with units [mag⁻² pc⁻³]). The (M_G, (B − R)) domain covered by both distributions is of course the same. Yet, the density peaks of ﹩{N}_{{ \mathcal C }}({M}_{G},(B-R))﹩ and of Φ₀(M_G, (B − R)) are in dramatically different places in the (M_G, (B − R)) plane. The reason is that the two distributions differ by the selection function integral (expressed here via V_eff), which is shown in the right panel; in the case at hand the selection function ends up being only a function of M_G, not (M_G, (B − R)); the survey volume in the top-right corner of the (M_G, (B − R)) plane is zero, as such objects were excluded by the initial Gaia parent catalog query.