A 2019 V-band pair permits a fixed 15% D800 return at 1574 days under an empirical error bound

We test one public V-band measurement pair near a proposed October 2019 return of the D800 dip in Boyajian's star. The test fixes the Kepler dip profile and shifts it by two periods of 1574 days. A 15% peak loss predicts a change of +0.074 mag at the quoted times. The measured change is -0.001 mag. We allow independent time shifts of 0.057 day and a baseline slope up to 0.005 mag/day. We compare two error envelopes from the same observer's later season. A Gaussian envelope gives an upper peak loss of 10.5%. An empirical envelope retains a 0.069-mag control change and gives 18.9%. The latter bound permits a 15% return. Small period shifts also change the limit strongly. The empirical envelope has a finite-rank interpretation only if target and control absolute errors are exchangeable. We do not establish this assumption across seasons. The test is retrospective. Its bounds apply to the fixed profile and stated nuisance assumptions. They do not reject all recurrence models. The package preserves source data, hashes, code, control pairs, period and nuisance tests, and figures for offline reproduction.

Boyajian's star · KIC 8462852 · D800 · D792 · 1574-day recurrence · archival photometry · error models · reproducible limits