Re: Focus indication

[Chris Albertson <chrisalbertson90278@yahoo.com>]

Focus is not something that you change in real time.  The typical
way to focus is first to have a camera with repeatable focus
settings.  Tom has this.  He can tell it to go 157 steps out and
it does.  Next you make a series of exposures at various focus
settings.  You give the set of frames along with the focus setting
used for each to a program that computes a focus setting vs. sharpness
function then finds the maxima of that function. 

"Sharpness" can't be determined by looking at just one star.  The
CCD is likely not at 90 degrees to the optical axis and the lens is
likely not 100% symmetric either.  The focal "plane" is actually a
kind of sin wave and not a plane at all.  You have to look at the whole
frame or samples of it and do the best _avaerage_ focus.

It would be best if you did not look at cosmic rays, ice crystals,
hot pixels or other defects.  Also "blended" double stars would
not be good to use either.

This program outputs just one number.  You don't need graphics for that.
I'd suggest writing in pure ANSI C so the program will be usable by
more people.

IMO focus belongs later in the pipeline. After you have detected
stars and tossed out the clutter and noise.  It still could be
done in real time, if you are runing a real time pipeline. But
remember it takes 20 or 40 minutes to shoot the focus sequence.
Don't make an interactive program.  Most TASS observers will
run with the CRT switched off.

But wouldn't you know it, We are not the first group to need to 
focus a camera.  THere is already plenty of software that does
all of the above.

Here is a short note with links
 http://iraf.noao.edu/iraf/web/irafnews/dec94/dec94-S-31.html

Here is what I would use:
http://iraf.noao.edu/scripts/irafhelp?kpnofocus



"Creager, Robert S" wrote:
> 
> This is an C/C++ Windows program I'm writing, not Qbasic.
> 
> Could you dumb down your explanation a bit :-)  I believe you're assuming
> I've found a star in my window.  While I'm sure I can find a star, does it
> make sense to find a 'good' star, or will just any old one do?  Marginal
> sums?  Half power I get.  I guess I should of brought my AIP book today...
> 
> > -----Original Message-----
> > From: aah@nofs.navy.mil [mailto:aah@nofs.navy.mil]
> > Sent: Wednesday, June 13, 2001 11:29 AM
> > To: tass@listserv.wwa.com
> > Subject: Re: Focus indication
> >
> >
> > Are you talking about additions to the Qbasic code to
> > do this?
> >   Calculation of the psf is not necessary.  The simple
> > method is to find the marginal sums in x and y, find
> > the center, and then march down until you reach the
> > half-power points on both sides.  That gives you the
> > fwhm to less than a pixel accuracy.
> > Arne
> >

-- 

--
   Chris Albertson             
   chrisalbertson90278@yahoo.com
   Redondo Beach, California
   home: 310-376-1029
   cell: 310-990-7550