 |
 |
Hi, Just to clarify this - there is no reason to believe that sinc is any more or less artefactual that spline. Both create ringing in the output image, and small sinc windows will have less ringing than spline. One big advantage of spline is that it is a lot faster than sinc (and is effectively a close approximation). Related to the post before - I think that zeroing the negative values introduced via ringing is probably a good alternative to using abs. All the best, Mark On 14 Mar 2012, at 14:10, Matt Glasser wrote: > I would also recommend spline over sinc, as sinc can sometimes have > artifacts in it and spline is essentially as good. > > Peace, > > Matt. > > -----Original Message----- > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf > Of Ed Gronenschild > Sent: Wednesday, March 14, 2012 8:55 AM > To: [log in to unmask] > Subject: Re: [FSL] Final interpolation in eddy_correct - repost > > Hi, > > Instead of taken the abs() I have set all voxels with value below zero to > zero. > > Regards, > Ed >