Recommended calibration data for MagE
(R. Simcoe, J. Hennawi, C. Williams, 4/6/2008).
Updated 10/2008, R. Simcoe

These data have worked for us but are not guaranteed to be appropriate for all programs. Use at your own risk.

The combination of broad wavelength coverage and lamp selection requires [THREE] TWO or THREE setups with MagE to get adequate counts in all orders for flat fields.

UPDATE:As of 10/2008, the issue with the 5 arcsecond slit seems to be resolved. During this run we took TWO setups: Dome flats for orders 6-13, and Sky flats with the 5" slit for Orders 14-20. This combination produced adequate pixel flats. [We recommend that you DO NOT use the 5 arcsecond slit for flat fielding. It appears to have an irregularity near the slit center that causes an undesirable line down the center of each order.]

It is necessary to defocus some flats to broaden features in the Xe lamp and/or sky. If doing this, be aware that defocusing moves the echelle footprint on the detector. So, you should take out-of-focus flats on both sides of nominal focus to cover the full spatial width of each order.
With a working 5" slit, we recommend against defocus because of the translation mentioned above. Instead, use the 5" slit to take a series of IN-FOCUS twilight spectral flats for orders 14-20.

Xe flash lamps and sky flats have significnat structure in the spatial direction, so you will need to use a high-order b-spline to fit this out when creating your final flat field images for calibration.

1. Trace flats:
1.1: Orders 6-13: Dome screen + Qf lamp + 0.875" slit. 30 seconds x 15 exposures.
1.2: Orders 14-17: Xe flash / 0.875" slit. 20 seconds x 10 exposures.
1.3: Orders 18-20: Twilight flat with science slit (0.875").

2. Pixel flats
2.1: Orders 6-13: Use same exposures as 1.1
2.2: Orders 14-17: Use 2" slit and defocus 1500 units BELOW nominal focus. Take 15 seconds x 10 exposures with Xe flash. Then defocus to ~1000 units ABOVE nominal focus. Again, take 15 seconds x 10 exposures, Xe flash.
2.3: Orders 18-20: Only sky works. 2" slit in twilight. In-focus may work but would be better to do out-of-focus in the morning (again, on either side of nominal focus).

3. Illumination correction
3.1: Use an in-focus twilight sky flat with science slit. Choose exposure time so that all orders are unsaturated. These may be the same frames used for trace flats.

Calibration Files:

Wavelength Map (FITS file)
Echelle Order Map (FITS file)
Wavelength Guesses (non-essential FITS file)
Pixel Image (FITS file)
Scattered light template (FITS file)
Order Structure Template (FITS file)
"Checksum" test data suite for MASE (gzipped tar file)