Hi Jason -

Sorry for the delay.

At first glance, the command lines look reasonable. Just a few questions about your data:

You are saying,

b0_rev.nii ---- reverse phase-encoded b = 1000
b0_fwd.nii ---- standard phase-encoded b = 1000

but these should be b=0 images. Is that what you meant?

Also, are all the input images you are using the same resolution, matrix size (i.e., number of pixels), and the same slice orientation?

If the answer to each of the above is, yes, then I cannot think of anything obvious that is wrong here. In that case, maybe you could upload your images somewhere from where I could get them to take a look what exactly is happening?

Best,
  Torsten