Scripts for Centering and Focusing OWL-based Direct Images

John Thorstensen, Dartmouth College

Last revision 2014 June

The point: It is sometimes desirable to center a direct image accurately. The MDM telescopes point reasonably well, but not to arcsecond precision, so accurate centering requires lots of overhead -- taking a test shot, displaying it, consulting your finding chart, figuring out how far to move the telescope (and it always seems like you have less than a 50-50 chance of getting the sign right), and so on. This tedious process wastes telescope time.

Also, you want your images to be IN FOCUS, unless you're deliberately de-focusing them for some reason. The focus of the 2.4m changes with temperature -- the numbers tend to decrease as the night goes on -- so it's very useful to have a reasonably efficient way of focusing. Also, one should be aware that, while sets of filters are generally constructed to be parfocal (equal optical thickness), mixing and matching filters from different sets leads to focus shifts between the filters, which you'll have to measure.

What this is: This document describes two scripts to help with these issues:

These scripts are tuned to work on facility CCD images written by Owl. It may be possible to adapt them to direct images taken by the 4k imagers or TIFKAM. Much of the functionality of centermdmtask is already included in osctrtask; the focustest task would take some adapting, because it looks for the "focus" keyword in the header and assumes it's the last value in the sequence. In addition, as of this writing there is no way to take a multiple-exposure focus frame at the 1.3m. This may be possible, but it is unimplemented.

Getting Started.

You shouldn't need to modify centermdmtask.py, but it will be useful to have it here anyway.

Using centermdmtask

Because there are a lot of knobs you can adjust, the script runs as a pyraf task, so you can use the pyraf parameter editor. taken an image of your field, and you want to establish celestial coordinates in the image and get accurate directions as to how to center your target.

Setting the script up. Do this before you start observing.

Running the script. Once again, epar centermdm (you can abbreviate the task name), and provide

and then press the Execute button, and away you go.

What it does: The script runs sextractor on the image to find objects, and then passes the list through a filter that should find stellar images. It also reads the image header, and from this finds the RA and dec written by the telescope. Then it goes to the PPMXL stars catalog (compiled from a number of high-accuracy astrometric catalogs). gets a list of stars in the vicinity of the telescope coordinates, and tries to establish the match between stars detected on the image and stars from the PPMXL. (Note: the algorithm uses the KNOWN image scale to evaluate putative matches, so you must have secpix set to better than 1 percent!) . If it finds more than three matches, it tries to do a coordinate solution, and presents it to you. if it thinks it's succeeded, you'll get a prompt like 

Accept this fit?
-- above it will be the field rotation (which should be close to a multiple of 90 degrees in most cases), the RMS of the fit (generally below 1 arcsec for a successful fit), and so on, so you can see if it's sane.

If you get a good fit, answer 'y', and the program will go on to tell you how far off the telescope is pointing, where the target is on the chip, and what coordinates to set the telescope to in order to place your object in the desired location on the chip. Just move the telescope with the hand paddle (turn off the guiding if you started it) and you'll be very close to centered. Or, you may decide that you're already pretty close -- if you've used the dial-a-guide-star technique, described in the guiding and acquisition manual and shown live in the video manual -- you should be very close to centered.

Error conditions There are various ways for this to screw up -- fortunately, it's purely advisory and doesn't actually move anything. It generally tells you about various problems in a fairly self-explanatory way -- just read the output. Some common issues:

Other ... The program produces various junk files (imagename.cat, imagename.radec.match.wcs, and so on) that you can chuck. Run in a subdirectory to avoid cluttering up your main workspace. The program is adapted from starfinding code I wrote years ago, which you might find useful but which is a bit tempremental. If you want it, I can set you up with it.

Getting the best focus with findfoc.py

This script is designed for fine-tuning the focus. If you're way out of focus, take test shots first to walk in on the rough focus before using this.

The Owl control system for the facility CCDs has a focus-frame script that runs on the 2.4m only at this point; it creates a multiple exposure by repeatedly opening the shutter for 10 seconds, closing it, then increasing the focus by 10 units and offsetting the telescope by 10 arcseconds north. The last image is offset by 20 arcsec, so you can tell at a glance which image is the final one. The last focus value is written into the header. The image is always written to /home/data/scratchimage.fit on the mdmarc computer, i.e. in /data/mdmarc1/scratchimage.fit when viewed from mdm24ws1. If you've selected your first focus value correctly, the star images should go in and out of focus, and you can interpolate the best focus value. Here's an example (it's ds9, but I've put the toolbar on the side to save vertical space on the laptop.):

The focustest script takes the guesswork out of the interpolation. To use it, simply set up the path to scratchimage as noted earlier, be sure a ds9 window is open, and then 

			./findfoc.py
It assumes the file it's going for is "scratchimage.fit"; if it's not you can specify another image on the command line.

Here's what it does:

Remarks. If the seeing is fluctuating quickly, you will of course get bad results, but cest la vie. Also, the Moffatt fit seems to be rather delicate.