As reported by @dkirkby, 60 second exposures often have negative sky models <4400 A, e.g.
This does not seem to be an issue for longer exposures, e.g. the previous spectroscopic exposure on the same night.
20200225/52081 is a case of a 300 sec exposure where the sky model dips negative:
The overall structure looks like an underlying CCD calibration problem (darks, biases) so start by inspecting the raw and pre-processed images. This is related to #978 (update master bias, dark).
As reported by @dkirkby, 60 second exposures often have negative sky models <4400 A, e.g. This does not seem to be an issue for longer exposures, e.g. the previous spectroscopic exposure on the same night.
20200225/52081 is a case of a 300 sec exposure where the sky model dips negative:
The overall structure looks like an underlying CCD calibration problem (darks, biases) so start by inspecting the raw and pre-processed images. This is related to #978 (update master bias, dark).