RE: Focus indication

[Tom Droege <tdroege@veriomail.com>]

Rob uses the term "Tom wants".  I am just out front on the problems of what 
we will all need that try to operate this beaste.  I am certainly open to 
better ideas from the experts.  I will soon have three systems to get lined 
up and focused.  Focusing is not much of a problem, Arne's scheme works 
just fine.  Lining up the systems and monitoring the sky is another 
problem.  Lots of things can go wrong.  Clouds smear out the images.  Frost 
does the same.  If the RA drive home position detector is not lined up just 
right, then the interrupter blade puts force on the position detector as it 
moves in and out.  This causes a little wiggle in the RA motion and makes 
the stars double.  There are many things like this.

What is needed is a "good data" detector.  One should then monitor this in 
the program and sound an alarm when the data does not look so good.

Arne says this is hard to do.  If it was easy, I would do it 
myself.  ;^)  Possibly it can't be done.  It does not have to be 
perfect.  I would like to start up a run and have the system give me some 
indication of how good the data looks.  Anything is better than 
nothing.  At present Mondown puts up a section of image.  A quick look will 
tell me that I left the light on when I last went out to do something on 
the porch.  I have done this many times.

The NS and EW components of the fwhm would give a clue as to the RA 
adjustment, the alignment of the telescope, condensation, fog, and other 
problems.

This is not like running a big telescope.  No one will want to sit all 
night monitoring the results.  One starts the run and goes off to watch 
Wrestling on TV or whatever one does to waste time.

Tom Droege

At 12:27 PM 8/2/01 -0600, you wrote:

>Lovely.  Thanks for all the ideas folks.  I believe I'll start simple, with
>Michael R's stepping +- the centroid in x,y, and maybe +-45 to see what
>happens.  If that doesn't provide what Tom wants, I'll go to 1d gaussian
>fitting, as that gets me both Michael's/Arne's marginal fitting and Chris's
>radial profile.  Then we'll move on to Doug's 2d fitting.
>
>Andrew, are you then interpolating with 3rd order splines, since the 4th
>order was too painful?
>
>Anyone have thoughts on how to determine when the "background tail is
>wagging the star-dog" for Michael's method?  Since I'm not going to present
>information on an individual star basis, but rather a mean of a subset of
>behaved data, this might not be an issue.
>
>Robert Creager
>Senior Software Engineer
>Client Server Library
>303.673.2365 V
>303.661.5379 F
>888.912.4458 P
>StorageTek
>INFORMATION made POWERFUL