MI/IMACS Observing Procedures
Date of last modification: August 11 2010, 3:40 pm Chilean time
1. Accessing Data Acquisition Computer (Llama)
- Username (obs1-obs9) and
password: will be set up for you by the instrument specialist.
- The data will be placed in /Volumes/Data_Llama/IMACS
- Check your working directory: it should be something like
20070903_04 or ut070904 that will be created by the instrument
specialist before you start.
- In Camera GUI, under Options, "DataPath" should point to this directory.
- 2. Start Control Windows in Llama
- The control software is started by typing 'imacs' in a Llama
- This will bring up the IMACS GUI and the Camera GUI.
Normally, those will be started by the instrument specialist during setup.
- From the IMACS GUI you will select the Slit,
Grism or Grating; adjust the tilt if using a grating, select
filters for direct imaging, and switch to Centerfield SlitViewer if needed.
- From the Camera Gui you will command the CCD mosaic:
exposure time, type of exposure (bias, object, flat), binning, etc.
- The filters, gratings/grisms and long slits will be mounted
previous to your run according to the instrument setup request,
and their names will show up in the GUI pull-down menus.
- In the CCD Camera GUI make sure that the "DataPath" found in the
"Options" pull-down menu points to the directory where you want the data
to be stored.
- Readout speed should be set to "Fast" unless otherwise adviced
by the instrument specialist.
- If the Quick-Look Tool has not been started, go to the "Options"
pull-down menu of the CamGUI and click there to start it.
- IMACS will adjust automatically the focus each time you change filter or grating/grism.
- 3. Subraster
The subraster to be used (if any) depends on which camera you are using (long
camera: IMACS/F4 or short camera: IMACS/F2) and also on the orientation, which
may be 'normal' or 'Node&Shuffle'. If working in 'normal' orientation
the spectrum will normally lay on 2 of the 8 chips (usually chips 2 and 5).
In 'Node&Shuffle' orientation, the spectra will run along chips 1 to 4,
so only those (or some portion of those) need to be read.
Clicking on the green array drawing (in CamGui) will allow you to select
which chips should be read.
Clicking next to the box which contains the word 'Full' on CamGUI will allow
you to select a subraster (a menu will pop up).
The subraster can also be written to
a file and its name loaded in that window, and there's also several subraster
files already available in the instrument default directory which can be
browsed from this window.
For example, if you are observing with the short camera (F2) and the
200-15.0 grism, you may want to save only chips 2 and 5 and trimming the
regions where the object's signal will be substantially null.
A useful subraster for this situation could be as follows:
2 1025 3073 2048 2048
5 1025 3073 2048 2048
If you write these 3 lines to a file and load it as a subraster you'll see
the selected region surrounded by a rectangle on the quick look tool. Before
using it, please check if
this seems allright (you may prefer saving a larger sample or even the entire
Ask the instrument specialist if you are in doubt.
- 4. IRAF imacs package
IRAF will be also started by the instrument specialist during setup, but
in case you need to reinitialize:
- Open an IRAF window and go to the working directory
- Load the 'imacs' package (this will automatically set the
stdimage parameter to 'imt8192')
- Open ds9
- 5. Check the slits and calibration lamps
- Bring up the flat field/calibrations GUI (if it is not already there)
by typing "ffs" in an xterminal of Llama or Burro. From there you can put the
screen in and out by clicking in "In" and "Out" and
turn on/off the comparison and quartz lamps.
- In IMACS GUI set Hatch to 'open' and 'filter' to 'spectroscopic'.
Unless you are interested in doing photometry out of your direct
images, you can use the spectroscopic filter for all your observing.
- Take a short (2-5 s) direct image of each of the slits you will be using
during the night. For example, 0.7 or 0.9 for the objects and 1.5 or 1.2 for
standards (the field 'Disperser' in the IMACS GUI must say 'f/2-Imaging' or
'f/4-Imaging', for the short and long camera, respectively).
- on Slit-Mask = Long 0.7/0.9; Disperser = f/2 Imaging; Filter: f/2 =
Spectroscopic2 (if you are using the short camera; otherwise it should
be f/4 instead of f/2).
- Measure on the Quick Look Tool (or using the display task in IRAF)
the position on the slit where your objects will be placed.
The slit has
several gaps: a good position for the objects should be midway
between two of those gaps.
If you have trouble in finding the slit position, try modifying the 'stretch'
parameter (on the right side of the Quick Look Tool) or ask your instrument
Go to this page
and follow the instructions there to
calculate the offsets for the 'toslit' task. Make sure that you pick up the
values corresponding to unbinned images.
The slit position will move slightly
along the night, thus a new slit image should be taken before each observation
and its position entered under xslit and yslit in 'icobject' (see below)
but you generally won't need to update the parameters for 'toslit'.
- The slit position will vary depending on which camera you are using
(F4=long, or F2=short) and the dewar orientation (Normal or N&S).
For example, if you are using F2, Normal orientation,
and 1x1 binning, a good slit position could be x=232 y=4055 on chip 5.
If you are using F4 and N&S orientation, your slit position
can be close to x=1945, y=3560 on chip 2, for 1x1 binning.
- If you are using the CenterField guider, go
and follow the instructions.
- Set the grating/grism on and the 0.7 or 0.9 slit.
With the long camera so far we have used the 300/4.3 grating,
at a grating angle of 6.1 degrees. With the short camera we have
used the 200-15.0 and also the 300-17.5 grisms.
Take a 5-10s exposure of the He-Ne-Ar lamps
(some lines will be saturated but the spectrum will be suitable
for wavelength calibration).
- 6. Afternoon calibrations
- Take a 10 bias sequence: set ExpType= bias;
this will automatically set the exposure time to 0s.
Dome flats for long camera (F4):
Turn on the Qh lamp from ffs GUI, Flat field screen should be 'in'.
- In camera GUI set ExpType=flat, ExpTime=20s, loops = 7
- In IMACS GUI set aperture=0.7", 0.9" slit, select grating/grism
and set 'filter' to 'spectroscopic'.
- Take another sequence with the 1.2" or 1.5" slit, with ExpTime=10s.
- Dome flats for the short camera (F2)
can be taken using the Ql lamp and exposure times of 10, 20 and 30s for the
1.5", 0.9" and 0.7" slits, respectively.
- Dome flats for the Centerfield slit are taken with
'Slit-Mask' in 'SlitView' and CF-Guider in 'SlitView' too.
In all cases take a test exposure before starting a loop and
check that the exposure time is appropriate (quartz lamps get
often replaced) giving maximun counts around 15000-20000.
If you care about good flatfielding
in the red region of the spectrum, it is recommended to obtain at least 3
dome flats right after or before the object observation. You can do it after
taking the comparison lamp. Otherwise fringing will not remove properly.
- 7. Object list
Your telescope operator will need a catalog with the selected objects
for the night (to which you can add new targets at any time).
You should edit it following the instrucions given
In that file, fields #8 (instrument rotator offset angle) and
#9 (instrument rotator offset mode) should be set to 133.85 and HRZ.
This will set the slit along the parallactic angle.
- 8. Observing
- In IMACS GUI set Filter=spectroscopic, Dispersr=F4/Imaging or F2/Imaging,
Slit-Mask: F4/Imaging or F2/Imaging
- Once the telescope operator has slewed the telescope to your
target, from the IRAF window execute 'toslit' (return has to be typed twice).
- In CCD window set ExpType to "object".
- Take a direct image of the field (10-20s) check that the target
is in the same chip you chose for the slit position; otherwise ask the T.O.
to move the telescope a few seconds and try a new exposure of the field
(use the compass next to the quick look tool to guess in which direction to
- Place the slit in and take a short exposure.
- Use IRAF display and imexamine to measure the position in the
slit where your object will be placed.
- Do an epar on 'icobject' and put the name of the
direct image of the field and the values of xslit and
yslit just measured; then execute the task. You will be presented with the
direct image and will be prompted to mark (with the space-bar)
the position of the object.
Answer 'yes' in order to offset the telescope (again, you need to type
- If the offsets are a few seconds you don't need to take
another image of the field. You can take a short exposure to see the target
through the slit, and then place the grating/grism on to start
taking the spectrum.
- In CCD Camera GUI set the appropriate ExpTime,
object to "object name", and click START to begin the exposure
- To take a comparison move screen into place (wait until
the 'wait' sign disappears) click to turn on the He-Ne-Ar lamps,
and take the arc exposure
(don't forget to remind the NA to disable guider
before the screen is moved to the IN position).
- If you are taking dome flats at the object position, turn the
comparison lamps off, turn the quartz lamp on, select ExpType = Flat and
take 3 dome flats.
- Turn OFF the lamps, move the screen back to the OFF position,
and ask the telescope operator to go to the next object.
- Take a minimum of 2 flux standards (3 is best) with the wide slit.
You don't need to take comparison spectra for those, but if you do, switch
to the narrow slit. If you are using the centerfield slitviewer for the supernova observing, then take the flux standards throug it; but if the supernovae are relatively faint and you are observing with the long slit,
do the standards with the wide long slit.
- Observe one telluric standard with the narrow
slit and take a
comparison lamp after it. Make sure you get high s/n in this data
without reaching the saturation level.