MDM Trouble Report for 2010 Nov 7 Telescope: 2.4m Observer(s): Haojing Yan (OSU) Instrument: OSMOS Problem(s) Encountered: I encountered numerous problems this night. After talking to Bob, it seems that a few software problems probably would not have occurred or could have been solved by logging out and a complete restart. However, some other problems are genuine problems. I. On the software side, there are two problems. 1. DS9 at Agung The pyraf script osctrtask.py (which we run on Agung during the slit acquisition process to calculate the offset between the target and the slit positions) puts the intermediate products to the directory where this script is first loaded. For example, one could easily do it this way: agung:obs24m% pyraf --> cd /lhome/obs24m/Scripts # go to where all the scripts are --> pyexecute( osctrtask.py ) This means that the intermediate results, most importantly acqimage.fits (the field image without the slit) and slitimage.fits (the field image with the slit in), will all be stored at /lhome/obs24m/Scripts. Now, here are a couple of cases that you'll likely get into trouble: 1) If you have a pre-existing DS9 window at Agung that was brought up in command-line, and if it was not first brought up in the very same directory /lhome/obs24m/Scripts, for example, if you did it in this way: agung:obs24m% cd /home/hiltner/data/osmos agung:obs24m% ds9 & osctrtask will report a miserable error like XPA$ERROR Unable to load fits acqimage.fits (DS9:ds9 8cfc5311:32816) and you will be left with nothing in your pre-existing DS9 window to continue the process. 2) Even worse, if the directory where you brought up your pre-existing DS9 actually happends to have the files of the same names as acqimage.fits and slitimage.fits , you will be in a bigger trouble. The way that osctrtask displays the images will put these two, wrong images of the same names to your poor, pre-existing DS9, and you will be working on the wrong images for your calculations! And, after you applying the offset, taking a set of new images and running osctrtask again, you'll be puzzled because nothing gets improved -- this is because the same, wrong images are displayed over and over again! Note that the above refers to the situation when you have a pre-existing DS9 window fired up in command-line. If this pre-existing DS9 is fired up by clicking on the icon in the toolbar, you'll be fine -- osctrtask will simply ignore it and will pop up a new DS9 window for its display. So, the best way to avoid problems is not to have any pre-existing DS9 windows before running osctrtask (or the other pyraf scripts like osbias ). However, it is ugly to leave any intermediate images in a directory that means to be a dedicated directory for scripts. It is better to load osctrtask.py and the like in this way: agung:obs24m% pyraf --> cd your-working-directory --> pyexecute( /home/agung/Scripts/osctrtask.py ) And then the intermediate images will be stored in your-working-directory. Of course, the best way is to change the way that osctrtask etc. store and display intermediate results. 2. Data directories At Hiltner when setting up Prospero, the default directory to be mounted is /lhome/data/ . However, at Agung this directory has a different name, /home/hiltner/data/ . Agung does have a directory /lhome/data/ , but it is a completely different thing. Overall, the names of the directories at these two computers are confusing and could be improved. II. On the hardware side, there were a number of problems. 1. The telescope once failed to stay at the postion where it just slewed to. This position was fairly close to the previous position (I was pointing away to a nearby bright star to check the telescop pointing), and it happened near the meridian and not far away from the zenith. Unfortunately I didn't get a chance to write down the coordinates and the time because there were some other problems happened then. The telescope actually slewed (I heard the noise from the motor), and reached the desired position (I was staring at the monitor watching the position values changing). However, it only stayed there for less than a second and dropped to a lower declination a few arcmin to the south (again I heard the sound from the motor). I don't recall whether it also changed in RA -- probably not. 2. The telescope position values as displayed on the TCS monitor drifted away while it was staring at the science target field. This happened multiple times at different locations. The telescope definitely didn't move, as the guide star was still in the guide window and didn't move. In most cases the values drifted away by a couple tens of arcsec (or more) in both RA and DEC. When it happend, the DEC value always changed to a smaller value (to South) and the RA value could change to both directions. 3. The telescope once lost its position during a long exposure. The field was at RA=02:17:36.7, DEC=-5:11:42. The telescope actually moved, because the guide star was lost and was moved out of the guide box (but still within the guide window). The happend at close to the meridian. The new position displayed on the TCS monitor, as I recorded on the log sheet, was at RA=2:17:36.9, DEC=-5:12:2, however, there could be a mistake in the DEC value, as (most likely should be DEC=-5:12:21), as I do remember I pulled the guide star back to its original position by using the paddle, and the position value displayed on the TCS monitor indicates that the telescope indeed jumped , but the actually jump was only half of the values as one would have calculated based on the displayed values. 4. The SDSS filters (at least the u'-band filter that I've been using) and the BVRI filters don't seem to be par-focal. The best focus obtained with the V filter always resulted in a non-optimal u'-band image. Finally I decided using u'-band for focusing. After the series of u'-band imaging I switched to BVRI again, and see the same thing. I decided to change it on-the-fly, and adjust from 4378 (u'-band best focus) to 4360, and got much improved BVRI images. From the last u'-band exposure to the R-band exposure before I adjusted the focus, there was only 13 minutes in between, and it is unlikely that the seeing changed by that much during this short interval. ------------------------------ Submitted on 2010 Nov 9 [7:12:51]