SUMMARY
Outline The single setupfield command will perform all steps of the mask pointing procedure as shown right. This includes obtaining mask/target images with/without masks, and move the telescope/FOCAS according to the results of the MOS offset calculator (see below). The calculator reads sets of FITS images obtained within the setupfield command as well as the output file of the MDP (the mask design program) to know the positions of the alignment stars on the mask.
MOS offset calculator Right image shows the 1st mask and 2nd target images (without mask) after offsets were calculated. Detected alignment stars and holes are shown marked by circles. The right sub-panels show the detecting calculation of the holes/stars (the case of stars is shown here). The detection can be done semi-automatically by using the mask design file (and small offset indicated by the observers). Manual re-detection can be made for robust detection by tuning the detecting parameters. Re-detection can be made at any times. You may remove any star-hole pairs for calculating the offset if the detection fails. Currently this software will be operated by the instrument operator or support scientist. ANA workstation at the summit will be used for this software.

Important note

Default FOCAS setup.

• CCD: 1X1 binning
• Filter: none
• This is to get enough sky and star counts in a shorter exposure time. 
• Exposure: 30 sec.
• This is to get enough sky and star counts for stars of V=17-19 mag.

Pointing outline

• Issue the SETUPFIELD command on the SOSS.
• Get a mask image at offset position. (1) 
• Get a target image with the mask retracted. (2) 
• In MOS offset calculator, compare (1) and (2) images, and calculate the offset in RA, DEC, and the theta (the mask rotation angle) utilizing sets of alignment holes/star positions.
• Move the telescope and the mask, and take another image with the mask inserted. (3) 
• Alignment stars should be seen through the alignment holes. 
• Compare (1) and (3) images with the MOS offset calculator, and calculate the offset again. 
• If the errors are small enough, stop the setupfield command, and takes spectra with the getobject command.
• If the errors are still non negligible, continues the setupfield command, and try another iteration.
• In most cases, at least one iteration will be required until the pointing error is sufficiently small. So, you expect at least three FOCAS images for the pointing.
  

Usage of the MOS offset calculator

• On ANA workstation, start up the mospointing2.pro software.
• Input the mask design file (the output of the MDP named *.sbr).
• Input the FITS file ID of the mask image and object FITS file ID.
• IDs are XXXXX of FCSA000XXXXX (just type XXXXX, skipping "FCSA000")
  
• You just need to input the ID of the chip 1 image.
• Wait until the mask FITS images are created on the ANA workstation. It will take about 30-60 sec. Press "OK" of the file input dialog when the images are ready.
  

• Combined image of the mask will appear on the right half of the window. Positions of the alignment holes will be marked by circles.

• All holes (0-4) will be automatically detected following the information on the mask design file (*.sbr).
• See carefully how well/bad the hole detection was made on the right end panels.
  

• Wait until the target FITS images are created on the ANA workstation. Note that the above step will be done before the target images are ready. Then click the "Let's go to Object Image NOW" button.
• The combined target image will appear on the left side of the panel. The expected star positions will be marked by circles, but usually there is 1~2 arcsec offset due to the telescope pointing error.

• Click on the first star position (labeled as "0") on the image.
• Other stars will be automatically detected according to the mask design file (*.sbr) after considering the offset calculated by using the first star.
• Watch carefully how well/bad the star detection was made. In this case, a star comes at the center of the upper panel (marked by a red circle), and the surface profile (below) looks fine, indicating a good fitting.

• Offset calculation will be automatically made after detecting all stars. The results will be shown on a small pop-up window as well as the errors of the convergence.

• Two fitting vector plots will appear on the right end of the window. Examine the plots carefully to find out how well (or bad) the calculation was made.

• Upper (before) panel shows the offsets for each hole-star pair. The vectors should show almost parallel movement of the offset. The RMS error of this plot will appear on the small pop-up window (57 pix ~ 5.7" in this case).
• Lower (after) panel shows the offset after removing the mean offset shown in the sub pop-up window. The vectors should show almost random pointings whose lengths are very small (~<1 arcsec). The RMS error of this plot (0.36 pix ~ 0.036") will appear on the small pop-up window.
• In both plots, lower left vectors are references (fixed to a 1X1 pix vector). Note that lengths of the error vectors in the lower panel are shorter than that of the reference, indicating the good offset calculation convergence.
• Note also that the RMS residual error shown in the pop-up window (0.36 pixel in this case) is very small.
  

• If the plots show one or two unusually large vector(s) comparing with others, there might be a mis-detection of holes/stars. Watch carefully the main panel to see whether circles (indicating the detected positions) are indeed drawn on star/hole positions. If you find any mis-detections, you need to detect them manually. (see below)

• If the convergence is good, copy and paste the offset values from the sub pop-up window to the user input dialog of the SOSS.

• The telescope/FOCAS will move according to the offsets, and FOCAS will take another target image with the mask inserted.

• While taking the next set of images, click on the "File Read" button on the MOS offset calculator to specify the next FITS ID.
• Wait until the images are ready on the ANA workstation. When they are ready, press "OK" button of the file selection dialog.
• New images will appear on the left side of the window, and the star detection will automatically start.
• The offset calculation result will show up on the sub pop-up window.
• If the convergence is good, but there are non-negligible offsets remaining, copy and paste the offset values from the sub pop-up window to the user input dialog of the SOSS for further iteration.
• If the convergence is very good, stop the setupfield command by pressing the cancel button on the SOSS. Proceed to the spectroscopy.

• In case of the wrong and/or mis-detection, use the lower buttons for optimizing the detection parameters or do manual detection.

• Redraw Star/Hole: Refresh star or hole images. No circles will be overlaid.
• Mark Star/Hole: draw circles around the current detected positions.
• Set Star/Hole Position: Detect each hole or star manually. Select star/hole IDs (0-5), and then click the corresponding positions on either image. The position fitting will be made by using the clicked position as an initial guess.
• Refit Stars/Holes: re-detecting all stars or holes automatically. The current positions are used as an initial guess for the star/hole detection.
• Refit Stars Semiauto: re-detecting all stars. In this case, click on the initial star position (labeled as "0"), and other stars will be detected by using the mask design file and the offset calculated by the first star.
• CALC: Do offset calculation by using the current detected positions. Press this button after re-detecting holes/stars or after removing/reusing sets of hole-star pair(s).
• Show images: Review the fitting results, by looking at fitted image of holes and stars.
  
• Change intensity: Set intensity threshold for the hole/star detection. In case of pixel defects/cosmic ray hits, you may need to adjust the intensity cut-off for robust detection of the holes/stars. This may also be useful when detecting stars over the diffuse underlying objects (such as a big galaxy, etc.).
• Delete Hole: Select a hole ID to remove the star-hole pair from the offset calculation. You may want to remove some hole-star pairs in case of poor detection even after trying manual detection. Press CALC button to calculate the offset without the selected hole-star pair(s).
• Reuse Hole. Select a hole ID to re-use the star-hole pair for the offset calculation.
• Change Fit Reg. Size: Change Fitting region size to exclude nearby objects/cosmic rays etc. for robust detection.
• Copy Hole Pos.: This command copies the current hole positions to the star positions discarding the current detected star positions. After issuing this command, you will need to "Refit Stars Semiauto" to fit stars from the very beginning.
• Refit All: Fit all holes and stars by using the current star/hole positions as an initial guess. This will be useful to check how well (or bad) the stars/holes were detected, or to find the mis-detected holes/stars.
• File Read: To input the next set of images for iterations. After reading out of the images, "Refit Stars" and "CALC" will be performed automatically.
  

Do you need iteration? ---  the practical criteria to quit the pointing.

• The telescope offsets (delta_alpha, delta_dec) are ~<0.1" (or RMS error is ~0.1")

And

• MOS theta offset is < 0.01 deg.

• Note: our CCD pixel scale is 0.1"/pixel.
• Error of 0.01 deg corresponds to only 0.7 pixel shift over the entire FOV.
  

Or

• Residual error vectors look almost random.
• Note: random error will come from the star image shift due to the bad seeing conditions and/or telescope tracking error.

Note: Error tends to get larger when seeing is bad, probably due to bad accuracy of measurement of star positions and image dancing.

Overhead estimation

• Image acquisition time per exposure
• Wiping: 2 sec
• Exposure: 30 sec
• Readout time: 130 sec (1X1 fast mode)
  
• Image offset calculation per comparison
• 60-120 sec (depending on how well the automatic detection will work)

• You need typically three exposures and two comparisons.

• Telescope slew: 1-3 min. (depending on the path length of the telescope slew. AZ rotation (0.55 deg/sec) will typically limit the time.)
• Autoguiding start/fine offsetting: ~1 min.
• FOCAS reconfiguration (slit insert/retract, filter wheel rotation, etc.): ~1 min.
  

• Total: ~20 min.