You were talking about doing it in the 3dDeconvolve
command, which is fine. It is a separate consideration.

The question about whether to do in with 3dDeconvolve
is just whether you want a significance statistic with
the contrast. If you want a single subject significance
statistic, to view, report or us with 3dMEMA, it should
be part of the 3dDeconvolve command. Actually, if you
care about the statistic for anything besides viewing,
consider using 3dREMLfit.

The contrast between 3dDeconvolve and 3dcalc would be
the same, it is just a question of the accompanying
statistic. Note that 3dREMLfit would produce a slightly
different contrast.

- rick