Hi Robin, thanks for the offer of help. I would certainly need conventional spectroscopic flats. Looking at the documentation what it seems to be doing is comparing the values of the pixels between the two flats and their exposure to see which are obviously non linear. So I think either spectroscopic or normal imaging flats will do provided they differ in intensity pixel by pixel by a factor of 10 or so after bias subtraction and are not saturated.
Regards Andrew