Hi Rick and Daniel,
> > Regarding the anat_al+orig and SurfVol+orig alignment, it
> seems
> > likely that the 3dAllineate matrix is the inverse of what you
> > want to apply, meaning you should use -ixmat_1D instead of
> just
> > -xmat_1D.
>
> I tried both, and neither works. With -xmat_1D the surface is
> off up to about half a centimeter (and the angle is off, too),
> while for -ixmat_1D it is a couple of centimeters.
I got it to work now
and indeed I had to use -ixmat_1D. Two other 'ingredients' were necessary:
-transform the surface coordinates from LPI to RAI, then apply the allineate matrix, then transform the coordinates back from RAI to LPI
- 'neutralize' the ALLINEATE_MATVEC_{B2S,S2B} matrices (set by the align_epi_anat.py) in the aligned anatomical by setting those to the identity transformation (b.t.w., is there any way to remove those from the file using 3drefit?) .
Thanks again for all your help.
best
Nick