Re: [Ida] RE: Cameras with Fisheye Lenses

[Jan Hollan <jhollan@amper....muni.cz>]

Andrej, Herman,

what's only needed, whenever the vignetting, geometry etc. will be known,
is: which darkframe is to be used for the image. One option is to shoot
images and darkframes alternatively, so e.g. odd numbers will be
darkframes, the even ones images. The other one, if you don't have so many
darkframes, is to say the software which darkframe it should use for the
given set of images (say, a whole directory with them) -- this is the only
task I do with my images (it is often complicated if I use various
exposures, e.g. to find luminances and brightnesses of distant lamps).

If you have your images and darkframes on any linux/unix machine, I can
process them there, just make me a user and give me a read access to the
images. I'll try to make a script to process all images in a fully
automated way. Of course, the calibration would be but very preliminary.

Alternatively, put some raws (inc. darkframes, ideally in a separate
directory if they are not identified easily by something as an odd/even
rule) somewhere so that I could download them. I would evolve the script
here at my computer.

Finally, with some rule for darkframes, the photometry task would be just
starting the script from the directory with the images. Hundreds of them
may be processed in a single batch, providing there is enough space on the
disk. (Of course, no flag would be given to RAWs -- these will remain
intact. Just the pngs or jpegs will be generated. Another ones might be
generated later, if better calibration would be available.)

Magnitude loss may not be the most robust characteristics, as the natural
sky luminance varies with time and angular height (and in 3000 m it is
brighter close to horizon than at the sea level). But the excess luminance
of an image compared with an image from a non-polluted site (difficult to
obtain... when nobody makes such an image with the same equipment on such
a remote site) might be a good proxy. If expressed in 0.n mag steps, it
would be magnitude loss in fact. Using medians, the two images may be
simply subtracted, to get the artificial part of the sky luminance.

The only thing I cannot do now is an automated stellar photometry. Another
people might be able to provide this task. As a colleague of mine (Filip
Hroch) works with a similar equipment (not really a full fish eye, no
giant CCD providing a circular field), there is a hope that he will do it,
being an author of "munipack" software. Of course it would be the best way
of calibrating the camera completely...

cheers,
 jenik