Thanks for the replies, but there is something that's still confusing me about this obliquity issue.

When I reconstruct my anatomical and EPI scans from DICOMs with the following commands

to3d -prefix 8101-T1 8101_DICOMs/*.dcm
to3d -skip_outliers -prefix 8101-EPI-01 -time:zt 27 273 2.0s alt+z EPI01_DICOMs/*.dcm

both are oblique.

If I then try to align the anatomical scan to the first volume in the EPI (after slice-time-correcting the EPI) with the following command

align_epi_anat.py -anat 8101-T1+orig -epi pb01.8101.r01.tshift+orig -epi_base 0 -volreg off -tshift off

the resulting alignment is *terrible* and when I look at it in the viewer, AFNI warns me about the datasets being oblique.

But, if before running align_epi_anat, I do the following to my scans

3drefit -deoblique -xorigin cen -yorigin cen -zorigin cen 8101-T1+orig
3drefit -deoblique -xorigin cen -yorigin cen -zorigin cen 8101-EPI-01+orig

the alignment is very good. This makes no sense to me, because 3drefit claims not to alter the .BRIK at all, but only to change information in the header. Furthermore, the 3drefit documentation says the "-deoblique" option should not be used "unless you really know what you're doing," and I don't.

I issued 3drefit in this way to the EPI before running 3dTshift, and I'm afraid I may be screwing up the slice-time information, but I can't find anything in the documentation that speaks to this.

Perhaps a better option for me is to leave the EPI alone before running align_epi_anat, and warp the anatomical scan so that it matches the obliquity of the EPI scan. Or maybe I should run 3dTshift, and then deoblique (with 3dWarp, as opposed to just changing the header with 3drefit) both the EPI and anatomy. After running 3dWarp, would it then be appropriate to change the header with the same 3drefit calls I have above?

What is my best option here? I want to preserve my EPI data as much as I can (warp it as little as possible).

Thanks,
Dan