The best way to obtain imaging flats with NOTCam is differential twilight flats. The differential mode cancels both the thermal component and the dark. In principle, one could use differential dome flats with the lamps on and off, but at the NOT we have no flat field screen and the inside of the dome is highly structured (and daylight leaks through the hatches). The flat field stability has been checked on a few master flats taken with the WF camera and the JHKs filters. There seems to be a larger difference between consecutive nights due to pointing on the sky than if the same blank field is repeated. This points to scattered light and different light distributions, and we decided to re-install the WF-camera imaging mask, a square mask in the aperture wheel to limit stray light in the system. Because the aperture wheel has only 4 large slots (+ 8 small) we decided to remove the 64 micron slit in order to find space for this mask. The mask was tested once in 2003, but not for twilight flats.
The photometric response over the FOV may not be flat even after having divided by the nightly flat field. For many IR arrays the so-called photometric surface correction (or illumination correction) is needed, believed to originate in scattered light in the system. A model made by observers for the H-band has been described and made available for download to the public through the NOTCam calibrations web page:
http://www.not.iac.es/instruments/notcam/calibration.html#illuFull data sets for the J and Ks bands are still pending, while the possible effect of the new WF imaging mask (see above) should be investigated for the H-band model first
Also, a new KG2 filter was installed to reduce the Z-band red leak. This needs to be tested.
Thomas Augusteijn 2013-12-02