FourStar – Calibrations
If exposure times are not known in advance, darks can be taken at the end of a the night. A python program is available, called mkdark.py, that will create a marco called mkdark based on the exposure information it finds in a certain raw data directory. To run the program, open a terminal on the Control-Room-Mac and change directories to the Raw Data path:
# cd ~/Data
[fourstar@Control-Room-Mac]# mkdark.py <selected-date> <# of darks>
# mkdark.py 2011_09_30 9
This script will take approximately 112.0 min to run.
The program calculates the necessary time to conclude the sequence, so that a note can be left on the “astronomer’s night report” for the day crew to know that calibrations are running (in case the script finishes after 8:30 am).
Once the script is ready, in the FourStar-Control window select “Dark” from the FilterCombo list, type mkdark in the macro “File” box and click execute.
The script will automatically set the obstype to dark and vary the relevant exposure times, loop numbers, gains, and modes based on the sequences of exposures found in the <selected-date> (2011_09_30 in the example) data path. The number of darks to take is set by the trailing number in <#of darks> (9 in the example).
Twilight flats are recommended to correct pixel-to-pixel variations in sensitivity. Generally one should start evening Twilight flats at the reddest wavelength and progress blue-ward and in the morning start at the blue and work toward the red. A twiflat.objects list can be opened from:
which contains regions of (relatively) low stellar density. We recommend a rot_5 or rot_9 dither pattern.
The observer can “Send” the twiflat coordinates to the telescope operator and just after the sun sets one can start testing the sky in the Ks filter with 1.456s exposure times. Once the counts reach the desired levels it is recommended to use at least 5.8s when taking any data on sky to mitigate well depth and non-linearity effects.
A rough guideline for flat-field order and exposure times is below.
Of course, in reality it is not likely possible to get all these filters in one twilight session.
Rather, an observer should create flats for filters they plan on using that night and scale these times/dither sizes to reach the desired count level; we suggest 10,000 to 25,000 counts with exposure times greater than or equal to 5.8s.
For Dawn flats, reverse the order, starting at J and working towards Ks.
Dome Flat Fields are not generally recommended although they can be taken to construct bad pixel masks and used for flat fielding if twilight flats are not available.
If the Calibration Unit GUI is not running already on the observer desktop, from a terminal on Llama type the command ffs to open it (it is best not to have two calibration unit GUIs running at the same time on the same account).
Click on the screen in button to move the screen over the secondary mirror. The lamps on the secondary are too bright to use with FourStar so the auxiliary variable lamp located in the dome must be used. The power supply is accessible in the dome; ask the Instrument Specialist for assistance locating it. The lamp is not very stable and some structure can actually be seen in the Hl filter from the lamp diffuser.
Some exposure times and lamp settings are shown in the following table.
|Filter||Gain||Mode||Lamp (V)||Exptime (s)||~Flux (DN)|
When you are finished make sure to switch off the variable lamp in the dome and remove the dome flat screen by clicking the screen out button.
For more information on calibrations with FourStar (including bad pixel masks) please refer to the manual (pages 34 – 39).