Pixel fits place the TOI-2300.03 lost light within 0.13 pixel of the target image in TESS Sector 40

A nearby eclipsing star can make a false planet signal in a TESS aperture. We measure the source of the single long event known as TOI-2300.03 on TIC 280031353. We use public Sector 40 pixels at 120-second cadence. A time model separates the local stellar image from the lost-light image. We fit the official pixel response model to both images. We test 22 time models in three image regions. All 66 fitted offsets are less than 0.125 pixel from the local stellar-image position. This is about 2.6 arcseconds at the nominal pixel scale. The primary offset is 0.0111 pixel. The largest fitted distance from the Gaia target position is 0.1671 pixel. Changes in the time model have a larger effect than the primary residual-noise estimate. We preserve block residual tests, full nonlinear noise fits, and numerical implementation checks. Each block length has 500 full nonlinear noise fits. Their 95th-percentile radii range from 0.0393 to 0.0462 pixel. This work adds numerical source offsets and model tests to earlier visual checks. The model range is a sensitivity result, not a full confidence limit. The fitted positions support the bright target image over the resolved nearby stars under the tested models. An unresolved source, an eclipse on the target, or a target-centered instrument effect can remain. The result does not confirm a planet. The package contains native input subsets, catalog snapshots, archive addresses and hashes, and response models. It also has code, figures, and offline repeat instructions.

TOI-2300.03 · TIC 280031353 · TESS · Sector 40 · difference image · source position · single transit · planet candidate · Gaia DR3 · pixel response function · block residual bootstrap