Okay - makes sense. I came to the realization that THD_copy_dset_subs was not going to work in this case ... I did the "manual" creation of an empty data set, got the data from the appropriate brik of the real time data set, allocated the memory and copied it in. That worked, but it added a little more code than I wanted to give back to you
After reading your comment - I re-employed THD_copy_dset_subs to create a temporary data set (as you implied it was for), and used PLUTO_copy_dset to copy that temporary dataset to the actual data set I will write to disk, and that seems to be working ...
I'll do a little more extensive testing over the next couple days, and will hopefully have something that you all might integrate back into the main source tree, AFNI gods willing