A revised frame rule predicts QLP minus TESS-SPOC times within 0.006 seconds in 16 new sector 83 stars

We test public TESS sector 83 times at native 200-second exposure labels. Earlier work found the QLP time change and a fixed frame model, but left four detector states unknown. We acquire one new star on each of 16 detectors. Three unknown states show the published frame term: camera/CCD 2/4, 3/2, and 3/4. Camera 1, CCD 1 has no later QLP data and remains unknown. The inherited boundary fails at one flagged exposure, 1012806, on each of 13 affected detectors. This is one exposure before the earlier first affected common center. We preserve that failure and freeze a revised boundary and state map before acquiring a separate star on each detector. In these 16 validation stars, all 162,789 finite matched pairs pass the fixed 0.050-second limit. The largest absolute error is 0.005964 seconds and the RMS error is 0.001571 seconds. No star offset, angle, or amplitude is fitted. All 325,578 pairs from both samples have revised errors below 0.010 seconds. Seventy-seven original mission image headers agree with the TESS-SPOC raw time reference within 4.283 microseconds in 154 comparisons. Native sampling narrows the observed onset bracket by 400 seconds, but a gap still spans 18,200 seconds. The package preserves code, lossless timing extracts, complete source headers, image prefixes, provenance, figures, and an exact offline numerical repeat. The result applies to the sampled source versions. It does not establish an absolute clock, a production software cause, or a new planet.

TESS · QLP · TESS-SPOC · sector 83 · astronomical timing · coordinate frames · CCD · barycentric correction · native exposures · reproducible research