The serial photometry that I perform requires precision pointing and rapid cadence. The target star needs to remain on the same pixels for every shot. The precision of my GEM mount is the limiting factor but generally its pointing errors are under 2 arc seconds when using plate solving. Donuts seems to be the ideal solution to meet my needs but unfortunately it does not currently work with GEM mounts. I have recently been in contact with the author of the Donuts, and he has kindly offered to modify the code in a future release to work in Voyager with GEMs.
In the meantime, I have tried to use the “Realign To Target” feature under the Guide/Dithering tab in Sequence Configuration. I found the realign process goes through the following steps when the target star has drifted a few arc seconds from the CCD’s center.
GOTO move using the target star’s coordinates. Depending on the pointing model of the mount this can move the scope well away from the target. For my setup, the pointing error was only a few arc seconds before the GOTO but increased to about 1.8 arc minutes after the GOTO completed.
First pointing image acquired, the image was plate solved and then the scope moved to recenter the star.
Second pointing image acquired, and plate solved. If the target was not within the “Precision Pointing Max Allowed Error”, the scope is moved again to center the star. This step was generally required for my setup.
A third pointing image acquired, and plate solved. If the scope was pointing to within the max allowed error, then science imaging was restarted. At this point, the target was usually dead center on the CCD.
As you can see, the “Realign To Target” feature is time intensive. A GOTO is performed, 3 pointing images are taken, and 3 plate solutions performed just to realign the telescope a few arc seconds. This really slows the cadence when doing serial photometry! I tried doing “Realign to Target” less frequently (every 8 minutes) but the results were unsatisfactory.
So here is my suggestion for a new feature. Why not just plate solve each science image. If the plate solution indicates the pointing is within the max allowed error, then take the next science image. If outside the max allowed error, move the scope then take the next science image. If the mount is not accurate enough to recenter in one move, then an option to realign using one or two pointing images should solve that issue.
Thanks for considering this feature!