output = (const*file1*file2)
Note that a 32 bit floating point buffer is used for each pixel. However the input and output data are written as 16 bit signed integers.
The resultant image is written into file1.fits unless overridden by the outfile option