FITS images may contain pixels in one of the following pixel formats: 8-bit integers, 16-bit integers, 32-bit integers, 32-bit floating point format, or 64-bit double precision floating point format. Converting image files from one format to another is not always an easy task, and you most probably would like to apply the same process over large number of files, e.g. to reduce the disk space they use.
iofits converts pixels from one depth to another. Be aware though that most conversions will incur some data loss if you are not careful about the dynamic range of the input images.
iofits man page