This is actually the related item I mentioned I'm working on.

AFNI is not written to handle oblique slices. As far as I know,
none of the relevant programs (programs where orientation is
important) such as to3d have _any_ code to deal with them.

However, sometimes, they work by a lucky alignment of the
stars. I'm looking at several cases that _don't_ work, such
as the Siemens Trio scanner, and trying A) to interpret scads
of undocumented data members in order to correctly align
such images, and B) to see how much general support for
oblique slices can be easily integrated into AFNI.

I don't think it's been a big issue until recently, but over the last
4 or 5 months, we've suddenly been bombarded by user questions
on this issue. I particularly would like to hear from any internal NIH
users that need more support for oblique slices.

That's the long answer. The short answer is yes, if to3d isn't
currently working with your data, you'll have to use something
else to register it, unless somebody else can suggest something.