You don't want to completely unstack the images. What you need is to have individual stacks for each micrograph. In theory, that's what EMAN2's Relion import tool is supposed to accomplish. If you don't have the original per-micrograph stacks and you also do not have the original micrographs with box coords, AND if the e2reliontoeman.py program isn't working on your STAR file, then send me an email and we can try to help you debug what's going wrong with the STAR file. That is by far the easiest course. If you still have the micrographs and box locations in a usable form, then that is generally the best route forward.