hi Rick,
> Well, at least one problem is easy to deal with. For some
> reason 3dAFNItoNIFTI is not dealing with the obliqueness
> correctly. However 3dcopy works just fine.
That is *not* what I see. 3dcopy yields the same (wrong) alignment as 3dAFNItoNIFTI. At some locations in the brain the alignment may seem somewhat OK but if you look at, for example, the location I described in the readme file (RAI (x,y,z)=(11,25,-17)) it is clearly off. Can you replicate this?
To be sure, I just downloaded the latest afni version, AFNI_2008_07_18_1710
[[Precompiled binary macosx_10.5_Intel_64: Jun 17 2009]]
Strangely enough, using 3dinfo the afni and nifti version show the same affine transformation matrix (IJK_TO_DICOM_REAL) and are in the same orientation (ASL), but the R-to-L extent values differ while A-to-P and I-to-S extent are the same.
Any ideas?
best,
Nick