Two TESS flux products give different fitted widths for one January 2026 HAT-P-70 b transit under a fixed profile

We test whether a January 2026 TESS observation checks a reported change in the transit width of HAT-P-70 b. The visit has one selected transit. We fit the same asymmetric profile to two flux products from the same images. We keep the time samples and relative weights fixed. Corrected flux gives a half-depth width of 183.69 minutes. Aperture flux gives 187.96 minutes. The difference is 4.27 minutes. A paired residual check with 30-minute blocks gives a conditional 95% interval of 2.67 to 6.05 minutes. Checks with 60- and 90-minute blocks also give positive intervals. These checks hold the profile fixed. They do not show which product is closer to the source signal. A comparison of four visits uses exposure averaging and a shared profile from the 2020 data. The corrected-flux width in 2025–26 is 4.66 minutes larger than in 2020. Whole-transit resampling gives a conditional 95% interval of 2.63 to 6.82 minutes. The January product difference is comparable to that earlier change. Thus, this one transit cannot give a clean check of its persistence. The fitted width assumes a time stretch of one profile. It is not a contact duration. Shared errors in a whole visit remain possible. The study does not establish orbital precession. The package includes raw public data, hashes, code, sample records, results, and five figures.

HAT-P-70 b · TESS · transit photometry · half-depth width · flux correction · exposure averaging · public astronomical data · reproducible research