[Aavso-photometry] [Norton AntiSpam] Re: Light box question]

[arne]

Ben Davies wrote:
> Guys,
> 
> Thanks for the replies.  It was the S/N ratio that I couldn't see how to 
> resolve.  I assume that in this situation you are not going for the 1 
> million electrons/pixel that Arne was mentioning in a previous post, and 
> are trading signal for accurate frequency matching.  What is reasonable?
> 
> And Michael, you mentioned a "chip flat".  Could you explain what that is?
> 
> I too, have never had great success at completely removing the ghosts of 
> stars in combined images.  I should see Dr Kao soon.  I will ask him 
> what software he uses and report back.
> 
For a master sky flat, where you are medianing a large number of
science frames together, you will have star wings left over because
they are too close to sky and are not rejected.  The first step is
to *not* combine any frame with a bright star or extended object.
The next step is to combine lots of frames, reducing the error of
one by a lot.  Finally, the eye is a marvelous tool in seeing
patterns - you may find that the star ghost image is actually a very
small fraction of one percent and the flat is good, even though it
looks a little weird.

Arne