[Aavso-photometry] Fwd: Light box question]

[Tomas L. Gomez]

If I understood it correctly, one way of doing a flat from the night
sky (specially if you have a lot of light pollution, as I do at home)
is the following:
combine all the (dark-corrected) frames you did during the night.
To combine the frames, use a "median" or "sigma reject", after normalizing
the frames, in orther to erase the stars. To use this method you need
frames from many fields, or at least you have to dither between frames,
so that, there is not a star always in the same pixel.

Note that this is essentially the same as a twilight flat, but instead
of twilight,
you are using light pollution. Therefore you have to use frames taken where
there is not much gradient. Thin clouds will reflect light pollution, and can
spoil this method.

I just did a quick experiment to check this method: I selected a night in
which I had good weather and was able to take twilight flats both in the
evening and in the morning. Then I took several frames made during that
night, each one from a different field (all taken with the same V
filter), already dark and flat corrected.
Then I did a "sigma reject" combine (rejecting above 3-sigma) with Maxim DL.
At this moment, what I get is basically a comparison of the twilight flat
(which I had used to calibrate the frames used in the experiment) and
the "night sky flat". In other words, if this method works, I should get a flat
picture, and this is what I got, within a standard deviation of 0.5%, and
no visible gradient. Then I did a quick estimate of the deviation I should
have gotten (assuming only Poisson errors), and I got about 0.4%.

The problem with this method is that it is very difficult to remove completely
the stars. Even after the sigma reject combine, I can see the shadows of the
bright stars. To avoid this from spoling the measurement, when making the flat
for a frame, you should use only the other frames. Another source of problems is
if there are nebulae or large galaxies in the frames.

By the way, the system I used is specially easy to flat correct, because it is
a newtonian to which I put an oversized secondary in order to have
full illumination
of the chip. In particular, this system has no vignetting.

I hope this helps,

-Tomas L. Gomez (GOT)

On 6/15/07, BailyHill at aol.com <BailyHill at aol.com> wrote:
> Hello Ben
>
> Mike did an excellent job of explaining the dither technique.
>
> Like everything in photometry, it has its limitations.  It is very  hard to
> collect enough images to get real high signal to noise ratio in the  flat.  As
> Mike said, you can spend more time doing flats than  observing.
>
> The other trick, is to remove the stars from these flats.  For me, I  have
> NEVER been able to make them really go away.  I can always see  residual stars
> at a low level.  Perhaps there is a trick that I do not know  able.
>
> I guess the bottom line is that you must match the flat technique to the
> problem you are working to get the optimum results.  What this probably  means is
> to try several techniques, and see which answer you like.   This technique
> will not always be the same.  I can tell you that I spend  some time almost
> every day working on flat fielding techniques.  Its that  important.
>
>
> Clear Skies
> Gary
>
>
>
>
>
> ************************************** See what's free at http://www.aol.com.
> _______________________________________________
>
> Aavso-photometry mailing list
> Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry
>