Hi, Daniel-

The grid of MNI152_2009_template_SSW.nii.gz (is that the file you are using as a base for @SSwarper?) is a 193x229x193 matrix, with 1mm iso voxels; here is part of it's 3dinfo output:

R-to-L extent:   -96.000 [R] -to-    96.000 [L] -step-     1.000 mm [193 voxels]
A-to-P extent:   -96.000 [A] -to-   132.000 [P] -step-     1.000 mm [229 voxels]
I-to-S extent:   -78.000  -to-   114.000 [S] -step-     1.000 mm [193 voxels]

Fitting 2x2x2 mm^3 voxels in that should lead to a 96x114x96 matrix, I believe, which is what you are observing.

What data set is your "mask" or seed defined in? You mentioned "MNI space", but there are lots of files that could be. The MNI152_2009_template_SSW.nii.gz looks like MNI152_T1_2009c+tlrc.BRIK.gz or mni_icbm152_t1_tal_nlin_sym_09c.nii from here (just with different brightness scaling):
[www.bic.mni.mcgill.ca]
That seems to overlap precisely (same grid dimensions, unlike for example the "mni_icbm152_t1_tal_nlin_sym_09a.nii" volume, which seems to have fewer axial slices).

--pt