User issue.
Occasionally the beginning of an observation can include a slew period, which causes the reference sky pixels to be out of range.
This was found for 30 Dor C, which I think is in the CVZ. Usually NuSTAR arrives in earthshadow so it doesn't matter if the slew timing in the observation (which is inserted by hand) and actual timing of the slew is off by a little bit. However, since we arrived while the source was "on", it look like the reference image puts the field of view outside of the 1000x1000 (about 40x40 arcminutes) allowed sky image.
This is a pretty rare occurrence, but should be probably be caught anyways when the projected image is outside of the nominal 1000x1000 grid.
Work around is to remove the slew period at the start of the observation by hand using a usrgti file.
User issue.
Occasionally the beginning of an observation can include a slew period, which causes the reference sky pixels to be out of range.
This was found for 30 Dor C, which I think is in the CVZ. Usually NuSTAR arrives in earthshadow so it doesn't matter if the slew timing in the observation (which is inserted by hand) and actual timing of the slew is off by a little bit. However, since we arrived while the source was "on", it look like the reference image puts the field of view outside of the 1000x1000 (about 40x40 arcminutes) allowed sky image.
This is a pretty rare occurrence, but should be probably be caught anyways when the projected image is outside of the nominal 1000x1000 grid.
Work around is to remove the slew period at the start of the observation by hand using a usrgti file.