With the fitted Kepler pulse shape and white noise, 30-minute bins keep over 99% of information in two TESS visits of KIC 12254688

We tested the effect of time binning on two public TESS visits of the known self-lensing binary KIC 12254688. We fit a pulse shape to three Kepler events. We then fit pulse amplitudes and local quadratic baselines at the published TESS pulse times in Sectors 41 and 56. We combined the same samples into longer bins and averaged every model term over those samples. For the best Kepler shape and independent white noise, 30-minute bins keep 99.488% to 99.861% of the amplitude information across 20 bin phases, two variance choices, and three fixed time positions. The measured off-event amplitude scatter changes by less than 0.6%. The original pulse amplitudes are 1679 and 1025 parts per million. The off-event scatter is 662 and 549 parts per million; these are descriptive noise scales, not calibrated confidence intervals. Baseline changes affect the amplitudes more than binning does. A broader fixed-shape grid keeps at least 98.907% of the white-noise information. The result supports combining existing short samples for this amplitude test. It does not measure hardware changes or the completeness of an unknown-pulse search. The package contains all 22 original public FITS files, input hashes, source requests, code, full results, and figures.

TESS · Kepler · self-lensing · KIC 12254688 · time binning · Fisher information · correlated noise · reproducible photometry