THanks RIck,
Actually, for each attempted sub-brik merge, the first error appeared as
*** -doall dataset nvals mismatch at file ../sh072303/zregression+tlrc[43,45,47,49,51,53,55,57,59,61,
along with the same error with the other two subjects whose referenced briks started at 43. I just cut and pasted jd as an example.
The uniform fussing about the three subjects' datasets made me wonder if afni does not like merging across differently-numbered briks anymore.
In the meantime, is it possible to adwarp-out only the subbriks of interest instead of all 74 subbriks in each statistical dataset to see if that saves some memory?
Jim