> Thank you for fixing the bug. If I understand well, using trilinear or sinc
> interpolation should not make big differences in processing time.
Sinc interpolation is much slower than tri-linear, which is why I originally
intended the code to do most of the work with tri-linear and just finish off
the registration with three iterations of sinc interpolation.
Incidentally, the realignment needs both the new "spm_realign.m" and the
"spm_defaults.m" (as there is an extra global variable included in the
for specifying the quality/speed trade-off).
Regards,
-John
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|