Re: Dumb question

[Arne Henden <aah@nofs.navy.mil>]

Richard Miles wrote:
> A possibly original suggestion on my part is that one keeps optical surfaces
> (in particular the filters and CCD window) as clean as possible and then try
> comparing precision and accuracy of the photometry WITH and WITHOUT
> flat-fielding.  One useful test is to image M67 and model the variation in
> response (un-flat-fielded) across the chip as a polar diagram, i.e. a smooth
> function of radius from, and angle around the optical axis.  Don't assume
> the optical axis coincides with the center of the chip however - it usually
> doesn't!  This type of approach, may I suggest, is an alternative to
> flat-fielding in the conventional sense.  It of course largely misses the
> localised 'donuts' and other dust motes but since you are drift-scanning
> then you are tracking across the chip and such donuts (which usually have
> max. absorptions of <5 or 10%) are averaged out.  Of course you mustn't
> disturb the optical setup between calibration runs on M67 or any similar
> field.
   As Michael mentioned, the Mark IV is a traditional stare-mode instrument,
so dust motes are important.  Even more important are systematic gradients,
specifically vignetting and image distortion.  These have to be flatfielded.
You can check your flatfielded image against known standards or with a
grid exposure to see how well the flatfielding performed, but looking
at raw images is a waste of time.

>>>I intend to advise anyone wishing to use the
>>>spreadsheet to reduce CCD images to make sure that they choose a photometric
>>>aperture that includes 98-99 percent of the light flux from nominal point
>>>sources not only for the entire area covered by the CCD image but also
>>from image to image used in the reduction.
Aperture selection is a well-known procedure.  TASS is blessed (so to speak)
with the majority of the psf fwhm coming from the optics, not the sky,
so you can use a single aperture setting under almost all conditions
and airmasses.

>...  Since the catalogs with reasonable number density
> such as Hipparcos are no better than +/-0.01 mag (against the standard
> system) there's no point in striving for 99.99% of the light to be included
> in the aperture.  Indeed the actual figure depends on the absolute
> photometric accuracy of the catalog you are using (as well as the image
> quality).  The principle here is that the uncertainty introduced by omitting
> up to x% of light from the aperture should be around half the uncertainty in
> the catalog itself (my opinion).  So if we put some numbers in as an
> example:  If the catalog uncertainty is +/-0.02 mag then the aperture should
> include at least 99% of the total flux, and on average for reasonably good
> quality optics the integration software will register 99.5 +/- 0.5% or so of
> the actual light, such that the error introduced in this case is only about
> 1/4 of the uncertainty in the photometric catalog itself - that's a good
> ratio especially since atmospheric transparency will also be competing to
> degrade photometric accuracy!.
   It doesn't work that way.  You don't have to include any specific percentage
of a star's flux, just that the same percentage has to be included for each
star measurement.  50percent works just as well as 99percent, since what
you are looking for is consistency and signal/noise.

Arne