Comment
N.B. (R.Campbell 14 Apr 2021): this problem was a specific to the correlation of the first part of the quasi-real-time ftp fringe test, and not intrinsic to those scans themselves in the observation and should not apply to the actual full correlation of the NME. The problem arose by using the clock offset passed during the chat without taking into consideration the "parity" whether it stemmed from gps-[fmout|maser] or [fmout|maser]-gps. The program that prepares the correlation vex-files knows the conventions in use across the various stations, so has this factor built it. By importing a value "off-line", we essentially overwrote the correct sign with the wrong one, in this case a net offset of ~22us (~1400 lags), outside the delay window being used in those ftp fringe-test correlations.