Effect of differing extraction aperture radii on the initial fits to our Spitzer photometry. We selected an aperture of approximately 2.1 pixels (i.e., C = 0.0) as the optimum, based on it having the highest ﹩\mathrm{ln}(p)﹩ and lowest residual rms in our initial fits. The actual aperture size varied as the square-root of the noise pixel parameter in each image (Equation (1)). As demonstrated by the rightmost panels, the exact choice of aperture size does not significantly effect our results: the variation off all three of these parameters (black line) as a function of aperture size does not exceed the 1σ uncertainties we find in our final fit (blue shaded regions)—except for the value of ﹩{R}_{P}/{R}_{S}﹩ for small apertures.