Local fits with a cubic baseline give times 11.8 hours apart for one V380 Cyg eclipse in corrected TESS sector 82 flux

We test local eclipse times in public Kepler and TESS light curves of V380 Cyg. One B eclipse in TESS sector 82 has a different shape in aperture flux and corrected flux from the same images. We fit a symmetric profile and a cubic baseline to the same 111 time bins. Six starting values give two corrected-flux time solutions 11.797 hours apart. One has a smaller residual sum. The aperture-flux starts give the same time to within 0.001 seconds of numerical variation. This is numerical agreement, not a physical time uncertainty. The reference corrected-flux fit is 255.620 minutes later than the aperture-flux fit and reaches a width limit. Another start gives a time 452.201 minutes earlier. The corrected-flux failure persists with 15- and 60-minute bins. A control eclipse in the same cycle has a product time difference of 0.82 minutes. The reference catalogue has 145 selected events in ten Kepler quarters and seven TESS sectors. In the TESS data, removing the single fit at a parameter bound changes a diagnostic B-minus-A timing slope from +3.31 to -10.96 minutes per year. Baseline choices also affect the slope. This result identifies an event and a local-fit failure that can affect orbit studies. It does not establish a new planet, a physical eclipse shift, or a precise new apsidal rate. The package has all 23 original FITS files, source records, hashes, code, fit trials, figures, and run instructions.

V380 Cyg · TESS · Kepler · eclipse timing · flux correction · apsidal motion · local baseline · public astronomical data · reproducible research