[Aavso-photometry] Photometry of stars in emission nebulae

[Wolfgang Renz]

Hi

> > Wolfgang Renz wrote:
> > I didn't mean that this will have "any benefit in finding better comp stars".
> > Without eliminating the emission lines you might just have to skip stars
> > with a strong nebula gradient in its surrounding (within the aperture) or
> > accept lower accuracy and zero point shifts for them.

> Arne Henden wrote:
> I'm not quite sure this is correct.  As long as the nebula contribution
> stays fixed, and as long as you use the same centroid for your aperture,
> the nebula is just extra background and only dilutes the stellar light.
> It is not much different than having a close companion that is included
> in your aperture. The net effect is to reduce the signal/noise and
> to modify the amplitude of variation.  You can still look for periods,
> flares, etc.  So in many cases, modification of the bandpass is not
> essential - it just helps.

Arne, with photometry packages using a good algorithm for centroid
determination and identical aperture sizes comparable seeing
conditions (FWHM) for all measurement you are right.
When I wrote the "MIGHT have to skip stars" comment I thought about
the preview tools and old software packages that are out there and still
used for photometry but do NOT use subpixel centering of the stars to
measure, use square apertures (that might be sensitive to field rotation
under the described conditions) or use a very poor algorithm for centroid
determination (thats especially difficult for low SNR stars under the
described conditions). Using such software will make it very difficult to
reach a precession of much better than 0.03 mag if the star aperture is
choosen pretty narrow (e.g. as a precaution of measuring stars in a
nebula).
A couple of days before I took a look at the Axiom Research Mira
Technote 6 ("http://www.axres.com/ -> Technotes -> 6 The Effect of
Centroiding Noise on Precision Astronomical Photometry"). It states
that varing the centroid position by just +-0.2 pixel in x and y might result
in a significant +-0.01 mag difference. Rounding the position to the next
pixel might result in much higher values than extrapolated linearly. Actually
this is a nice example for the effects of selecting a small aperture mask
(~ 2.6 FWHM diameter). Using a slightly larger aperture mask (3.0 FWHM
diameter), the "Centroiding Noise" would be (at least in this example) no
issue anymore (on the cost of a 1/3 larger "sky background inclusion" that
causes also some noise). 

So my guess was that larger gradients in bright nebulae like M42 within
the star and sky apertures using such software packages might also
have their effects on photometry of stars with low SNR.

Cear skies
  Wolfgang

-- 
Wolfgang Renz, Karlsruhe, Germany
Rz.BAV = WRe.vsnet = RWG.AAVSO