We had a lot of trouble because of that recently as well, in our case
fitting T1 and T2 series. We had a whole mess of results and it took a while
to understand where the problems came from. We then had to manually go
through the whole time series, check the NC_proc values and then do it all
again with a set NC_proc value. It is now possible to load an entire time
series as a pseudo 3D, so there is less of a problem. Still, the parameter
is there, why not use it?
|