A fixed CCD and frame model predicts QLP minus TESS-SPOC times within 0.02 seconds in 96 sector 84–85 stars

We test a fixed detector baseline for relative times in public QLP v01 and TESS-SPOC v1 files. Earlier work explained a sector 85 time change with a fixed frame rotation, but it measured a baseline for each star. Known TESS time adjustments predict a baseline of 1.979 minus 0.020 times (CCD number minus one) seconds. We acquire 96 TICA image headers at six epochs in sectors 84 and 85. In 320 matched header and star comparisons, the measured baseline differs from the prediction by less than 48 microseconds. We combine this term with the earlier frame and camera rule. We save the formula, sample rule, and error limits before acquisition of 16 new sector 85 stars. All 171,856 finite matched points in these stars have errors below 0.010 seconds. Their RMS error is 0.001636 seconds and their largest absolute error is 0.005421 seconds. No star offset is fitted. Across all 96 stars, 1,034,336 points have errors below 0.020 seconds. One sector 84 control star fails the stricter 0.010-second limit. A transfer function that uses QLP alone gives the same predictions within 0.086 microseconds. The package preserves native timing extracts, source headers and hashes, the ephemeris, code, tables, figures, and offline instructions. The result tests relative time conventions in these source versions. It does not establish absolute clock accuracy or the production code cause.

TESS · QLP · TESS-SPOC · TICA · time stamps · CCD readout · barycentric correction · coordinate frames · sector 85 · reproducible research