.fits and .sdf files are complex data storage structures. When pamela is used, it looks in specific places for several pieces of header information, so these bits of information need to be put in the correct places. The original pieces of header information go under slightly different names at different places, so there is a command which works for several specific file sources.
The pamela command fixhead must be run to sort out various pieces of header information. fixhead requires an integer to tell it which night of the observating run this data is from, and an integer to tell it precisely what to do. Running fixhead without these will give help as to what you should put:
fixhead $ No night number supplied. fixhead 1 $ No format number supplied. $ Formats available are: $ $ 1) WHT July 1998 $ 2) AAT December 1998 $ 3) WHT January 2000/01 $ 4) INT March 2001 $ 5) VLT UVES pipeline data $ 6) SAAO March 2002 $ 7) INT April 2003 $ 8) VLT FORS1 PMOS circular, Oct 2003 $ 9) VLT FORS2, Mar 2004
where the listed formats were those available as of December 2005. If none of these are appropriate then you'll need to add one yourself into the perl script. Now run fixhead on all of your .sdf files:
fixhead [night] [format] *sdf