[vsnet-chat 1726] CCD photometry

[Stupendous Man <richmond@a188-l009.rit.edu>]

  William Walker asked some questions about CCD photometry.
Let me answer a few of them.  But I'll start by saying that
it would be a very good idea to read a basic text on CCD
image processing: try 

        "Astronomical CCD observing and reduction techniques", edited by
           Steve B. Howell, Astronomical Society of the Pacific (conference
           series), 1992.
        
>   The people at Mt John suggest that CCD photometry of this type has an
> accuracy limit between 1% and 2%. Presumably this is based upon LMC
> photometry with 0.5 an 1 metre telescopes. What do others find? 
  
  My experience: it's easy to perform relative photometry (two stars
relative to each other on the same image) to 2% with a CCD.  It's
hard -- meaning one must take into account many factors -- to perform
all-sky photometry to a level of 2%.  For one thing, one needs a
site at which the sky conditions don't change a lot over a period
of several hours.  I don't have one in Rochester, NY, USA :-(

  One trick to increase the signal-to-noise ratio of relative measures
on a CCD frame is to throw the stars out of focus (!).   That permits
you to take a longer exposure before the chip saturates.  But it
also means that you must use big apertures to measure the light from
the stars.  It makes sense if you are trying to do relative
photometry of stars which are about the same brightness.

> Another puzzle is when a star is measured for different times - say 1 and 2
> seconds. The magnitudes between these should always vary by a similar
> amount but don't.

    a. you are using an aperture which is the wrong size.  For unsaturated
         stars with a nice PSF, an aperture of radius 1 to 2 times the FWHM 
         is pretty good

    b. your camera's shutter isn't accurate to 1% for short exposures.
         This is very common.  No way to fix it -- you'll have to use 
         longer exposures

    c. you moved the chip between exposures, and you didn't flatfield
         properly.

  Your specific questions:

> 2. Is there any suggestion that the exposure ST6 timing system is variable?

  I'd be surprised if the shutter is accurate to 5 milliseconds.
For exposures of less than 5 seconds, that's 1 percent.

> 4. Should we flat field for each filter? This would make the data reduction
> a little more complex but wouldn't be impossible.

  Absolutely necessary for CCD work if you want to get precision 
of 5 percent or greater.  Sorry.


                                      Michael Richmond

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp