AutoFocus Tool User's Guide

This is the user's guide for the AFTool

A reoccuring problem: IDL license server dying.
If (when) the IDL license server dies, we end up with the autofocus software dying every five minutes (IDL will come up in demo mode).
The license server runs on duke, and must be restarted by root (as root), using the sys5_idl_lmgrd, which can be run using:

service sys5_idl_lmgrd restart

update: The root of the problems seems to be MIRSI somehow issuing a command to duke to shut down the license server. We thought we'd fixed this, but the problem came back on 7/25, as reported by B. Golisch.

NOTES Regarding IDL:

• When things go bad in an IDL session: I have had IDL sessions and software freeze up on me.  You should be able to select the xterm session from which you started IDL and use the ctrl-C key command sequence to exit most IDL programs (this will dump you back to the idl shell which you can then exit from using the "exit" command).  On occasion, and with the AFTool, I've had to select the xterm, enter the ctrl-C key command sequence, then mouse over the AFTool GUI to get the shell to process the ctrl-C.

Initial Setup

Step 1: Start a VNC server session on stefan as user autofocus using the appropriate instrument

1a.

1b.

From here, everything is done inside the VNC client

Step 2: Start an xterm, and run the command xhost +

Step 3: Start the GuideDog user interface and an instance of DV.

Step 4:

IDL Version 6.2 (linux x86 m32). (c) 2005, Research Systems, Inc.
Installation number: 216038.
Licensed for use by: University of Hawaii

IDL>

Running Autofocus

Step 5: Start the AFTool

At the IDL> prompt, run the command @aftool (yep, include the @ symbol). This should run the AFTool (which compiles the apropriate source files, loads them into the IDL environment, and then runs the autofocus program), and display the following window:

NOTE: if instead you get an error about file aftool not being found, your IDL_PATH variable is not set correctly. Exit the IDL environment (exit), and try running source .cshrc Your IDL_PATH setting should be already be set up in the .cshrc file, but may not be getting loaded properly. The IDL_PATH variable should include /usr/local/idl_irtf, and that is probably not being set correctly. If running source .cshrc fails, contact Miranda or Charles. If it succeeds, let Miranda or Charles know via email that you had the original problem.

Step 6: Setup Instrument using the Autofocus Setup button, customize, then point telescope at focus object

• Click the "Autofocus Setup" button located in the lower left corner of the AFtool UI.  This should set up the instrument to a generic setup suitable for autofocus (this replaces the running of the autofocus_setup macro in SPEX).
• Customize filters and instrument settings as you feel are appropriate based on current conditions.  The generic settings may not be correct for the conditions you are in.
• Point the telescope at the object you are going to focus on.  Take an image with guidedog and verify the correct position using DV.  NOTE: The object you focus on must be a star of 7th or 8th magnitude or brighter.  Also, use single stars, NOT double stars.  Also, the star must be located AWAY from the slit, so that an out of focus halo does not overlap the slit (both ABEAM position and BBEAM position), and away from the edge of the image.  The following is an example of two cases that should not be used:

Step 7: Set up AFTool

7a

Select the apropriate instrument and secondary mirror. If running in demo mode, change the Data Mode from INSTRUMENT to ARCHIVE.

7b

The TO and support astronomers should have a fairly good idea where focus should be based on current dome temperature.  Please really use the correct temperature, check that it is correct.  If you do not use the correct temperature, the AutoFocus tool will not work correctly and time will be wasted.  (At a later date we will be generating a table and plot for correct focal positions based on temperature)  Use the Estimated Focal Position slider control to set this value.

7c

Set the increments value ("At increments of:")  to 0.10 or 0.05 (0.05 is recommended. 0.02 and less can be used, but takes significantly longer) and the number of points to be used ("Number of Points:").  These two numbers are used to indicate how many images to take and how far apart.  A wider spread would be selecting 0.10 for the increment value and 9 points for the number of points.

As an example, if the estimated focus is -0.20, the increments value is 0.05, and the number of points is 5, aftool will sample data at -0.30, -0.25, -0.20, -0.15, -0.10.  

7d

Leave the Working Directory alone, unless you are working in demo mode, in which case you must enter the directory in which your test data resides.

7e

We currently recommend 1 image per set. This has worked fairly well for all cases tested to date.

7f

Leave the Number of Sets alone.

7g

Ensure that "Estimate Seeing By:" has the fwhm mode selected.

Step 8: Start Data Taking

Click the GO button. This will start the image acquisition portion of the process. Quite a bit of debug is printed out to the shell. If you are running the instrument with DV enabled, you should see the acquired images show up.

Once the GO button is clicked, it should change from a green background to a red background, and instead of reading "GO" it should read "Busy".

Once the image acquisition phase is complete, the GO button background color will change again to yellow, and will read "Waiting".  The user will be presented with a display window containing a sum of all the images acquired:

Step 9: Determine box size around star

The focus object should be obvious. At the shell you will be asked to right click on the center of the object. Please do so. After two seconds the shell will ask you to right click on the right side of the object.  Please do so at a distance where the star has transitioned to sky, but no further.  These two points will be used determine the size of the subarray extracted as useful data.

Step 10: Verify curve fit

The images will then be processed, plotted, and a curve fit to plot points. A view of the plot will be shown, and a window will then appear showing you the recommended focal position. If the data plotted and curve fit seem reasonable, and the focal position seems reasonable, the focal position presented is likely accurate and should be used. A reasonably good example of a successful fit is:

NOTE: the From and To values in this case were from -0.8 to -0.2, centered around a suspected focal position of -0.5

However, if the data plotted and the curve fit shown seem overly suspect, or the focal position seems unreasonable, most likely the AFTool failed to succesfully generate reasonable data based on the settings given, and the telescope will need to be focused manually.

Here is an example of a failed attempt to focus from -0.8 to 0.1 :

Step 11: Accept focal position

The final estimated focal position and FWHM value will then be presented in a new window.

• "Use and Restore" sets the focus position to the calculated position and restores the instrument to it's original settings.  If everything goes successfully, this should be the button you use.
• "Use" sets the focus position to the calculated position, but does not restore the instrument.  The instrument can be restored by using the "Autofocus Restore" button on the main AFTool UI.  If you wanted to re-run AFTool using the calculated value as the center point, you would use this button.
• "Origin" sets the focus to the position that was entered by the operator as the estimated focus value.
• "Cancel" closes the dialog immediately and does nothing.

The Info View Window

It provides a representation of the subarrayed sample actually used for focussing.  It also provides the timestamp of when the data was taken, and holds the focus position and estimated best case seeing value.  It is persistent, and will not disappear when an option from the Calculated Focal Position button is selected.

Returning to normal operation

Step 12: restore instrument parameters