RE: Using flat field image

["Albertson, Chris" <CAlbertson@primeadvantage.com>]

You could scale the dark subtracted flat so that it's
mean pixel value equals 1.0

Or you could simply divide the program image by the unscaled
flat and then multiply the result by the mean of the flat.

Watch out for 1) the possibility of a zero pixel value in the
flat that can happen when you subtract the dark frame.  2) rounding
errors in the computations. 

Note that after processing an image that has a range of
pixel values that fill 16 bits after dark/flat processing
the range is larger.  I'd convert everything to 32 bit
floating point pixels before doing any processing.

> -----Original Message-----
> From: Creager, Robert S [mailto:CreagRS@LOUISVILLE.STORTEK.COM]
> Sent: Monday, December 18, 2000 3:35 PM
> To: 'Tass Mailing List'
> Subject: Using flat field image
> 
> 
> 
> Quick (hopefully) question:
> 
> How does one divide a flat field image into the dark-subtracted target
> image?  I.E. what scaling is done before/after the division 
> to either/both
> images?
> 
> Related question:  Should the flat field image (since it's 
> derived from sky
> pictures) be filtered in any manner (box, ???)?
> 
> Thanks,
> Rob
>