We have noticed that some images have a JD time that differs from the JD implied by the time information in the FITS file header. By checking (almost) all files that have a JD in the filename against the time in the header the following list of suspect images are revealed:

2455633.8734508
2455633.8734679
2455633.8734848
2455633.8735015
2455641.6260333
2455643.4725818
2455643.4800437
2455643.4817785
2455643.4846651
2455643.4866171
2455643.4891739
2455643.4911491
2455643.5007650

This list is the JD from the filename – the value is 1 hour larger than the JD implied by the DATE field in the FITS header. None of these appear on the list of good images as determined by Chris. and his ‘tunelling’ code which inspects magnitudes vs lunar phase.