Observing with the Integral Field Unit (IFU)

How Can We Help?

Observing with the Integral Field Unit (IFU)

Created On
Last Updated On
bylcoadmin

UPDATE: As of October 2018 IFU acquisition can be done either with F/2 or F/4 cameras.

  • If the IMACS GUI has not already been started by the Instrument Specialist, open an xterm, type "imacs ", and follow the instructions in the "Getting Started" section of the IMACS Manual.
  • Confirm that the Instrument Specialist has loaded the IFU in the slit mask server, as requested in the Instrument Setup Request Form that you submitted at least a month before your IMACS observing run. In the IMACS MechGUI, the SlitMask pulldown list should include IFU in position 6 with empty slots at positions 4 and 5.
  • Take a direct image of the IFU to use for setup of the required data extraction and display tools. To do this, first ask the Instrument Specialist to open the telescope mirror covers. Start the Flat Field Screen gui by typing "ffs " in an xterm. Once the gui appears, click the mouse on the "IN" button to put the screen into the telescope beam. Now, in the IMACS MechGUI, make sure that the Hatch is in the "open" position, the calibration lamps are all turned off, and the CF-Guider is in the "out" position. For acquisition, the Slit Mask must be set to the position of the IFU, and Disperser should be set to the camera of your choice: either "f/4-Imaging" or "f/2-Imaging". Set the Filter to "Spectroscopic", and the Focus Control to "Focus". Focus control should now adjust the detector focus with an insertion/extraction of the IFU in the same way as the detector focus is adjusted with a filter change.
  • Now, in the IMACS CamGUI, set ExpTime to "5", Loops to "1", ExpType to "Object", Binning X to "1", Binning Y to "1", and then click on the "Start" button to take an exposure. (You can modify the exposure time if you get too few or too many counts.) [Note: do not use image subrasters for the IFU images as some of the helpful data analysis software tools do NOT work properly with subrastered images]

 Software Setup:

  • Open an xterm window and cd to your data directory.
  • Next step - corresponding spot file should be copied to your working directory. This file contains first guess spot positions for the fiber bundles as viewed from the camera. Depending on the camera of your choice, type use the following :
    • F/4 camera:   "cp /usr/local/magellan/idllibs/ifu_acquire/spot_f4nods.txt ."
    • F/2 camera:   "cp /usr/local/magellan/idllibs/ifu_acquire/spot_f2norm.txt ."
  • Now type "idl" to start idl in the current working directory. This will startup with the default paths already in place to run all the image extraction and display tools (developed by Adam Bolton: you can look at ifu_acquire for some older instructions about these routines)
  • Set the camera configuration for acquisition in IDL environment (this will set the proper extraction parameters for the remaining image display routines). Depending on the camera, at the "IDL>" prompt type:
    • for F/4 camera: ifu_set, camera='f4nods'
    • for F/2 camera: ifu_set, camera='f2norm'

Now display the extracted section of the latest IFU image that you have taken by typing:

        for F/4 camera: ifu_slit, frame='iff????'

        for F/2 camera: ifu_slit, frame='ift????'

where ???? corresponds to the frame number.

You should see now the following image:

Next step is to overlay the IFU fiber bundle locations on the extracted image. This is done by typing "ifu_spot" which should overlay red + symbols marking the assumed spot positions based on the values in the corresponding lookup table ("spot_f4nods.txt" for the f/4 camera and "spot_f2norm.txt" for the f/2 camera). These spot positions will not overlay exactly (indeed the first image may even require a large shift in X and Y). See example below:

In order to properly reconstruct the IFU image from the slit image, you will need to tweak the displayed red spot positions to match the actual spots in the ifu_slit image. You need to specify, for two well-separated x-positions along the slit image, required x-shifts and y-shifts needed to be applied to the assumed spot positions at those x-values in order to line things up (again the offsets that you measure are from the red cross to the white fiber spot position). The routine that you will use is 'ifu_tweak' with the correct syntax being "ifu_tweak, [x1, dx1, dy1, x2, dx2, dy2]" where x1 should be the measured x location of a fiber spot near the left of the slit image and dx1 and dy1 correspond to the measured offset from the red predicted spot position to the actual spot at x1 ... and likewise of x2 measured at the far right of the slit image. To move around the slit image in the ATV window you must move the small green box in the upper right mini-image. First, move all the way to the left and measure a spot (best to take one that is bright enough to be well defined at the edge of a fiber bundle) - this will give you x1,y1. Next, measure the position of the corresponding red predicted spot for that same fiber and record x1predict,y1predict. Now move to the far right of the slit image and choose and measure a spot at x2,y2 and again find the corresponding predicted red cross at x2predict, y2predict. Following the syntax of the 'ifu_tweak' command, you will need to type "ifu_tweak, [x1, x1-x1predict, y1-y1predict, x2, x2-x2predict, y2-y2predict]" - this will recalculate the predicted spot locations to match those in the slit image.

Now, type "ifu_spot" again to overaly the newly 'tweaked' spot locations. Check that the red crosses now overlay nicely on top of the slit image fiber spots. If they do not prefectly overlap, repeat the 'ifu_tweak" process by re-measuring x1 and x2 positions and offsets and again running 'ifu_teak'. Once the overlay is matched to the data, proceed to extract and display the reconstructed IFU image. This is an example of the successful tweaking:

To show the IFU source image type "ifu_im, /im" and to show the IFU background image type "ifu_im, /bg". Here is an example of such image reconstruction:

F2/camera
F/4 camera

Night Operations:

  • The observer can (and should) prepare his/her Observing Catalog in advance with the desired rotator mode and offsets. Note, in order to obtain a suitable guide star and SH star in the two outside probes, it may be necessary to change the rotator offset angle.
  • Have the Telescope Operator (TO) slew to the first object that you wish to observe. From the IMACS MechGUI, make sure that the IFU is out of the beam by commanding the Mask Server to "F/2 Imaging". Once the IFU is out of the beam, send the Center Field Guider (CFG) to the "Center" position with the lense set to "Image". Note: when inserting or extracting the IFU from the beam, the IMACS GUI will present a modal dialog box to request that the instrument be rotated to a particular rotator angle. This is done to be sure that the IFU is only moved in and out along a horizontal motion. If you want to avoid the rotation of the instrument back to horizontal for the IFU insertion and extraction, it is advised that you adopt the Instrument Rotator Offset Mode in your catalog of GRV (to hold gravity invariant) and use an Instrument Rotator Offset Angle of -46 degrees to hold the mask server horizontal to the platform. RECENT TESTS SHOW THAT IT IS POSSIBLE TO INSERT AND EXTRACT THE IFU AT ANY ROTATION ANGLE - SO IT IS SAFE TO IGNORE THE MODAL DIALOG TO CHANGE THE ROTATOR ANGLE.

Note that it is desirable that the first object to be observed at the beginning of a run be one that will be easily visible in the IFU. If this is not the case, we recommend using a standard or a nearby, bright(V ~ 16) star in verifying the image position in the IFU.

  • As soon as the telescope has slewed to the position of the object, the operator can adjust the total exposure time of the CFG until you can recognize your field and center on the IFU target. Once the target is identified in the field, the Telescope Operator will need to move the telescope so that the target is translated to the center of the IFU imaging bundle [currently pixel 719,717 of the CFG] or the background bundle [currently pixel 50,435].
  • Once the object has been moved to the correct pixel, you should move the CFG from the "Center" to the "Out" position with the IMACS MechGUI. With the CFG out of the beam, you may insert the IFU from the IMACS MechGUI Mask Server. After the IFU is inserted in the beam, the Telescope Operator can now acquire the appropriate Guide Star and SH Star for the two probes.
  • Alternatively, following a slew to a new field you can directly offset the telescope close to the IFU source bundle position with the following 'toslit' parameters in IRAF. Set short camera, xbin=1, ybin=1 xoffset=-87 and yoffset=105.  With an up to date pointing model, the 'toslit' offset alone has proven adequate to place targets near the center of the aperture following telescope slews. To center your target in the source bundle without using the CFG (perhaps for a very faint source) you can take an image in the f/4 camera and move your target to the pixel corresponding to the IFU source bundle with the icobject command in IRAF. You will need to set the parameters of icobject to image=iffc5 (this will refer to the latest f/4 camera image for chip 5) xpos=106, ypos=4070. When you run icobject, the image of chip 5 will be displayed and your target should be very near the upper left corner in the displayed image. You can mark the target and the routine will caclulate the appropriate offsets to apply to move the telescope.  NOTE: The background bundle can be accesses with iffc3 at position [370,3947].  Therefore it should be possible to move from the image bundle to the background bundle using a toslit command set for the long camera with an offset of 564 and a offset of -252.
  • The CFG positions (or the alternate toslit and f/4 camera image offset) should place the target within a few tenths of an arcsecond of the center of the IFU bundle, but if you would like to confirm and further tweak the position you can take an image of the field in the camera. To do this, go to the IMACS MechGUI and make sure that the Disperser Server is set to "f/4-Imaging" (or "f/2-Imaging" if you are using F/2) and the Filter is positioned to "Spectroscopic" or a desired blocking filter. Make sure that the Focus Control is set to "Focus". Now, in the IMACS CamGUI, set ExpTime to an appropriate value so that your object will be easily visible (but not saturated!), Loops to "1", ExpType to "Object", and make sure that Binning X and Binning Y are set to "1". Finally, click on the "Start" button to begin an exposure.
  • In your IDL session, you can now properly extract the slit image and reconstruct the IFU image in order to verify the target position within the IFU bundle and to determine fine offsets for the Telescope Operator to apply as coordinated offsets (if needed).

First, display the extracted section of the latest IFU image that you have taken by typing, depending on the camera:

   for F/4 camera: "ifu_slit, frame='iff????'" , where ???? corresponds to the frame number.

   for F/2 camera: "ifu_slit, frame='ift????'" , where ???? corresponds to the frame number.

Next step is to overlay the IFU fiber bundle locations on the extracted image. This is done by typing "ifu_spot" which should overlay red + symbols marking the assumed spot positions based on the values in the default lookup table that were 'tweaked' during the afternoon setup tests. These spot positions may not overlay exactly due to small shifts of the f/4 mirror when removed and inserted in the Disperser Server.

In order to properly reconstruct the IFU image from the slit image, you will need to tweak the displayed red spot positions to match the actual spots in the ifu_slit image. You need to specify, for two well-separated x-positions along the slit image, required x-shifts and y-shifts needed to be applied to the assumed spot positions at those x-values in order to line things up (again the offsets that you measure are from the red cross to the white fiber spot position). The routine that you will use is 'ifu_tweak' with the correct syntax being "ifu_tweak, [x1, dx1, dy1, x2, dx2, dy2]" where x1 should be the measured x location of a fiber spot near the left of the slit image and dx1 and dy1 correspond to the measured offset from the red predicted spot position to the actual spot at x1 ... and likewise of x2 measured at the far right of the slit image. To move around the slit image in the ATV window you must move the small green box in the upper right mini-image. First, move all the way to the left and measure a spot (best to take one that is bright enough to be well defined at the edge of a fiber bundle) - this will give you x1,y1. Next, measure the position of the corresponding red predicted spot for that same fiber and record x1predict,y1predict. Now move to the far right of the slit image and choose and measure a spot at x2,y2 and again find the corresponding predicted red cross at x2predict, y2predict. Following the syntax of the 'ifu_tweak' command, you will need to type "ifu_tweak, [x1, x1-x1predict, y1-y1predict, x2, x2-x2predict, y2-y2predict]" - this will recalculate the predicted spot locations to match those in the slit image.

Now, type "ifu_spot" again to overaly the newly 'tweaked' spot locations. Check that the red crosses now overlay nicely on top of the slit image fiber spots. If they do not prefectly overlap, repeat the 'ifu_tweak" process by re-measuring x1 and x2 positions and offsets and again running 'ifu_tweak'. Once the overlay is matched to the data, proceed to extract and display the reconstructed IFU image.

To show the IFU source image type "ifu_im, /im" and to show the IFU background image type "ifu_im, /bg".

Finally, you must decide on any fine tuning offset of your target within the IFU bundle. Two arrows indicate North and East in the reconstructed image that is approximately 4"x5". Any desired offsets should be communicated to the Telescope Operator who can then command a coordinated offset of the telescope and guide probes.

  • You should now be ready to initiate spectroscopic observations. In the IMACS MechGUI, set the Disperser to the grating or grism that you will use, and make sure the Filter is still set to "Spectroscopic" or a desired blocking filter in whichever camera you will observe with. In the appropriate IMACS CamGUI, set ExpTime and Loops to the desired values, and then click on the "Start" button to begin the exposure.
  • It will be necessary to also adjust the focus offset depending on the grating or grism that is selected.  The offsets are relative to the instrument's Auto Focus value determined by the system. Optimal focus offsets were measured during engineering in April 2014 as follows: Grating 150 -100 Grating 300_4.3 -350, Gating 1200_26.7 +400, Grism200 +50, Grism300_17.7 +50, Grism300_26.7 -250, but these values should be checked for your particular setup.  These offset values should be set using the Focus Offset menu item on the CamGUI.

Table of Contents