Reporting bad exposures

Hi John,

reporting problems on this forum sounds like a good place to me!

What BSM systems appear to be having the centering errors - all of them, or just one?

Arne

Hello,
I often see images that have shifted on Hamren. Most of the time the target is still in the image, just not well centered. The shifted images may lose some comparisons, but are probably useable. It is rare that the target is out of the image.
Duane

Hi John,

Sorry about the length of this reply, but I wanted to be complete in my analysis and to recommend to the BSM team some possible future changes.

eps_Aur was observed on 180405 and 180410 from BSM_NM, as you mention.  It hasn't been observed at BSM_NH since 180103, unless I've missed something!  It was cloudy at BSM_NM at the beginning of the night on 180405, though eps Aur shows on the images.  Astrometry.net shows eps Aur being in the field for 180410, in the same place it has been for several months (the uncentered bright star).  The centering was originally offset to attempt to pick up more comp stars.  The BSM_NM Paramount points *very* well, so if you need a different centering, just let us know.

V2048 Oph was observed on 180402 and 180411 (so, the last two visits that you mention).  Both of these nights had significant clouds; if you look at

http://images.aavsonet.aavso.org/bsm_nh/bsm_nh_180411_images/index4.html

for example, you can see that the images for this field and others are blank (and were bad most of the night; very early stuff, such as at the "index1.html" page, look ok).  The same goes for 180402.  I wish we had better weather in NH!  One solution to this is to use ACP's weather app, and suspend operation until the weather improves.  For those sites with cloud monitors (such as BSM_NM, BSM_NH and TMO61), we need to implement this function.  For other sites, we should supply them with cloud monitors.  They are ~$500 each, so not a major funding issue, especially if donors can be identified.  That would remove most of the blank images.

The last two observations of 11_Cam were 180318 (properly centered) and 180326 (wrongly centered).  It has not been observed since 180326, as it is at 5hrs RA and therefore low in the sky at twilight (and BSM_NH has significant tree obscuration to the west).  The 180326 observations missed the field by about 4 degrees.  This was when we were in the process of swapping hand controllers on the mount, and had firmware issues.  I think many of the fields from March 16-March 26 UT had pointing problems while we solved the controller issue.

With the inexpensive mounts at all BSM sites except BSM_NM, the pointing errors can be large when moving to a field distant from the last field.  Usually ACP takes a pointing image, then recenters before the first science image.  It also plate-solves the science image, and will continue to recenter if the pointing error is larger than a few arcmin.  That is why you will sometimes see a field shift between the first image in a set and the remainder (though that movement is usually only a few arcmin).  During cloudy weather, two problems creap in.  First, if it was cloudy during the pointing exposure, it may not properly center the field, and the first exposure may be very wrong.  The solution to this is better mounts (with APASS on Paramounts, for example, we don't even do pointing exposures, and save overhead). This recentering may be the problem that Duane reports for BSM_Hamren.  Second, if it is cloudy, the system may fail in its periodic autofocus update, and images may become out of focus.  The solution to this is to use temperature-compensation for focusing, which is possible with the brand of focusers we have.  We're looking into that.  Then focusing only needs to be done once at the beginning of a night, and the periodic autofocus is avoided (more time savings).

So, from these two fields, there are some takeaways for the BSM team.  First, there are a team of volunteers that inspect the images from each night and give internal quality assessment.  Those assessments both get posted on a maillist (which the hardware and software folks use to look for problems), and get imported into a database.  The original intent was to make that database accessible to researchers, so that they could see how their images were in comparison to the "norm".  I'll work with George Silvis and Bill Toomey and see if we can make that happen.  In the meantime, I highly recommend to researchers that they look at their images in the full nightly context on the images.aavsonet.aavso.org site.  In addition, *most* of the sites have an all-sky camera, with provided software to create a nice all-night summary image, showing when clouds were present.  It would be good to make those images available to researchers as well, as you just can't tell remotely what the sky conditions were without such ancillary information.  Finally, we need a dynamic "maintenance" log that can be viewed by researchers so that they know when hardware was changed or problems were found that resulted in reprocessing, etc.  The problem with such a log is keeping it current, but I think it is valuable to have.  We'll put that in a list of update features for the network.

John, you are a wonderful participant of the AAVSOnet system, as you let us know quickly when your images are not proper.  PLEASE do not hesitate in keeping us informed.  As I mentioned, this forum is for all things AAVSOnet related, and problems (as well as successes) should be voiced.

Arne

Ken, you are talking about transferring images to astrometry.net and then getting the results?  There is a way in linux to use the astrometry.net library locally and plate-solve from within your program.  That is as fast as pinpoint.

Arne