With local quadratic baselines, VV Ori eclipse profiles are narrower in TESS sector 98 than in sectors 6 and 32

We test whether the later TESS eclipse profiles of VV Ori differ after each profile has a free depth scale. We use public 120 second SPOC light curves from sectors 6, 32, and 98. These visits span December 2018 to January 2019, November to December 2020, and November 2025 to January 2026. With local quadratic flux baselines and even empirical splines, both eclipse types have smaller widths in sector 98. From sector 6 to sector 98, the primary width at half depth falls by 3.73 minutes. Its conditional 95% interval is -4.28 to -3.20 minutes. The secondary width falls by 3.98 minutes. Its interval is -5.24 to -2.62 minutes. These intervals come from resampled event blocks. The widths also fall from sector 32 to sector 98. All 14 tested reduction choices retain the decrease. The primary half-depth change from sector 6 ranges from -3.89 to -3.71 minutes across these choices; the secondary range is -5.00 to -3.71 minutes. Binned profiles and separate visit halves support the change. Constant scaling of the measured deficit cannot explain these fractional-width changes. The inference depends on the local flux baseline and does not exclude flux-dependent aperture losses, coherent pulsations, or more general added-light models. We do not infer an inclination rate or a companion. The package includes the three source FITS files, checksums, source records, code, bootstrap results, figures, and reproduction commands.

VV Ori · TIC 50897998 · TESS · eclipsing binary · eclipse width · stellar pulsations · public astronomical data · reproducible research