[Aavso-photometry] Lightbox flat question

Tue Sep 13 03:13:37 EDT 2005

Hi,

When taking lightbox flats, you must consider what your telescope is actually doing:
it converts an image at angular coordinates (sky) into spatial coordinates (CCD
frame). In fact, you never talk about your telescope spatial field of view, but
angular field of view.

It is always said that the lightbox must be evenly illuminated. I think that this is
only partly true, but more important is that the angular light distribution of the
lightbox is the same than angular light distribution of the sky imaged. If your
field of view is relatively small (a couple of degrees or less) then the sky can be
considered evenly illuminated also at any direction you are viewing it. A
lightsource of this kind is called Lambertian source, and it's angular intensity
distribution follows the cosine function. 

It must be considered that any point at your lightbox should act like a Lambertian
source: emitting light at any direction so that it's luminance is constant. If this
point is emitting light at narrower angle then it is seen darker at large viewing
angles. In case of a lightbox the angle where light is emitted as Lambertian source
must be larger than angular field of view of your objective. Otherwise the edges of
your flats become too dark. There is nothing to do with diffraction nor attenuation,
all what matters is that your angular field of view is evenly illuminated by the
Lambertian lightbox.

It may be not easy to make your lightbox as Lambertian source for large angles. You
may visually inspect your lightbox at different angles (within your objective lens
field of view). If the lightbox illumination remains unchanged it should work fine.
But if you can see clear "hot-spot" because of your light source, or the
illumination level changes significantly at different viewing angles then you do not
have a Lambertian source. It is then expected that flat-fielding does not work
properly. This may be significant when working with wide angle camera objectives.
Rick's propsal of diffuser or two in front of objective illuminated by twilight sky
may work well.

Is there an optimal distance of a lightbox? If your lightbox illumination is even
across the lightbox then the distance should not matter at all. If the illumination
is uneven but Lambertian (this may be because of a single light bulb, or dust in the
lightbox screen) the lightbox must be close enough not to be focused at CCD. I think
that focal length or less should do this (to be more precise, closer than the object
focal length). If your lightbox is not Labertian within your field of view then you
can't save anything by getting the lightbox closer.

Even though you do not have evenly illuminated but Lambertian lightbox you should get
properly flat-fielded CCD images. However, the difference should be seen at the dust
doughnuts, which will be improperly corrected. I have not tested this carefully, but
those who has more experience please show me if I am right or wrong!

Best regards,

Pertti

Lainaus Richard Huziak <huziak at sedsystems.ca>:

> Rob,
> 
> You need equal illumination across the front of the light box, and the 
> box screen has to be big enough for your lens angle to 'see' that equal 
> illumination and not go beyond it.  The viewing angle of your 8" scope 
> is about 2 degrees (ie - approx 'straight ahead').  However, your 28mm 
> lens may have a viewing angle of 75 degrees and the 130mm lens about 18 
> degrees (guessing a bit on the exactness of the angles).  So the wider 
> the lens, the closer it has to be to the screen.  With the 28mm, if 
> angle is 75 degrees, and if your screen is 1 foot wide, you'd have to be 
> no further than 14 inches from it, provided your screen is truly equally 
> illiminated at the extremes.  Seems to me that the closer the better 
> applies here for a light box, except I can envision issues with changes 
> in illumination from the front screen resulting from the 'diffraction 
> angle' from the screen into the lens (ie - attenuation at the edges on 
> very close angles).  That then contradicts my initial impression on 'the 
> closer the better'.
> 
> It may be better to do sky flats with a diffuser over the lens instead, 
> but I can see some real problems with very wide angle lenses, since the 
> sky is not equally illuminated at all angles at dusk.
> 
>  From your '0.6 to 2x lens f/l' statement, it seems you know the above 
> issue already.  Others, are, however, doing telephoto photometry - what 
> are you guys doing?
> 
> Rick
> 
> Robert J. Modic wrote:
> 
> >I've taken lightbox flats with my 8" f/5 Newtonian for some time
> >now with mostly good results.  The lightbox is placed on the front end
> >of the tube (about 1x the scope's f.l. from the primary mirror). 
> >Recently, I've tried to use several short f.l. lenses (28 to 130 mm
> >f.l.) for wide-field CCD photometry.  When I try to use a lightbox to
> >take flats, the results are sometimes uneven.  Depending on the
> >distance of the lightbox from the lens, the resulting flats can look a
> >bit mottled and they don't flat field my science frames as well as I
> >would like.  I've experimented with distances ranging from 0.6x to 2x
> >the lens f.l.  The shorter distances seem to give the best results.
> >
> >Is there an optimal distance to have a lightbox from a lens?
> >
> >Bob Modic
> >
> >
> >_______________________________________________
> >
> >Aavso-photometry mailing list
> >Aavso-photometry at mira.aavso.org
> >http://www.aavso.org/mailman/listinfo/aavso-photometry
> >  
> >
> 
> -- 
> 
> * * * * * * * * * * * * *
> Richard Huziak
> Manufacturing Engineering
> SED Systems, Saskatoon
> tel. (306) 933-1676
> <huziak at SEDSystems.ca>
> * * * * * * * * * * * * * 
> 
> 
> 
> _______________________________________________
> 
> Aavso-photometry mailing list
> Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry
> 

-- 
Pertti Pääkkönen, PhD, Laboratory engineer
University of Joensuu, Department of Physics
Tel: +358 13 251 3238
Fax: +358 13 251 3290