SXT_COMPOSITE combines two or three SXT images with different exposures into
a ``composite'' image. The longest exposure is used as the basis for the
final exposure. SXT_COMPOSITE uses the results of SXT_SATPIX to
identify the locations of saturated pixels and replaces them with
equivalent pixels from the shorter exposure images, scaled for the long
exposure image. This routine uses the algorithm that is used to produce
the ``sfd'' images in the standard SXT movie. Example calls are:
IDL> cimg = sxt_composite(index,data)
IDL> cimg = sxt_composite(index,data,satpix=satpix)
IDL> cimg = sxt_composite(index,data,dc_data,sx=[1,2,3],sd=[0,2,1])
By default, SXT_COMPOSITE will do dark and leak frame subtraction if
data is byte-type. The user may optionally specify the dark-frame images
in dc_data and will default to 
unless the explicit
X-ray and Dark-frame indices are given in 
and 
, respectively.
Note that the replaced pixels are box-car smoothed with a smoothing window of 5 pixels.