Comparison of the Spitzer pixel-phase-corrected flux using Eqs. (1) (top) and (2) (middle). Data have been averaged every 32 images (≈1 min) and normalized. Vertical bars stand for the standard deviation of each one-minute bin. Both solutions seem rather alike, even though the BIC comparison favors Eq. (2). Bottom panel: difference of the corrected data shown in the top and middle panels. The first ≈24 h of data do not show a significant modulation, so the amplitude of the periodic signal in Eq. (2) tends to zero, and so Eqs. (2) ≈ (1).