Re: dawn and twilight sky flats

[Tom Droege <tdroege@veriomail.com>]

I remind you that at the other end of this is the shutter opening 
time.  Since I failed to spring for Uniblitz shutters at several K each, we 
are stuck with a shutter that takes about 0.15 to 0.2 seconds to open and 
close.  This should be consistent within milliseconds, and so is just one 
more thing to take out with the flat field.  That is, if the flat field is 
taken with the same exposure length as the object field.

I have settled on 100 second exposures as a compromise.  This should keep 
shutter opening time effects down in the noise.  For longer exposures, the 
sky brightness would fill up too much of the well.  I get of order 4000 
counts of sky in the I camera with the 100 second exposure.

Tom Droege

At 02:26 PM 11/7/01 -0500, you wrote:

>   Tom points out the Mark IV takes so long to read out that
>one can't get very many flats while the sky is bright, but not
>too bright.  D'oh!  I had completely forgotten about that.
>It would help if one set the exposure time to, say, 2 or 3 seconds
>instead of 60 seconds, but that fixed 40-second readout period
>is still a problem.  That makes twilight flats less attractive,
>for sure.
>
>   Tom asks if one could combine flatfield images from dawn and
>from dusk.  Yup.  If you point south (as the Mark IV typically
>does), then the sky gradient will go one way at dusk and the
>other way at dawn, so combining them will tend to remove the gradient.
>Sort of ...
>
> > Note that a lot of stars can be seen in a dawn flat ...
>
>   If one changes the operating procedure, so that the exposure
>time is only 2 or 3 seconds instead of the usual 60, there will
>be many fewer stars visible in these exposures.  That's a lot
>easier at dusk (when the operator is awake) than at dawn (when
>he's asleep), of course :-/
>
> >  The levels will thus vary wildly, and one must do something to
> >  normalize the data before combining.
>
>   Absolutely true.  I normalize all the exposures so that the mean
>level is the same in each before I combine them.
>
>
>                                                Michael