Hi Jim,

Stick with the -stim_times files that you have.

Use 3dDeconvolve's -stim_times option with TENT(0,17.5,8) as
the basis function, for example. This example assumes your TR
is 2.5 seconds. This would use 8 TENT functions, over 7, 2.5
second intervals, as your basis functions. If your stim_times
files happened to be TR-locked, this would produce the same
results as using binary -stim_file files.

See the deconvolution handout for a visualization of the tent
functions, on slide 63 of:
[afni.nimh.nih.gov]
[afni.nimh.nih.gov]

As an example, consider the 3dDeconvolve command at the end
of the script generated by the command:

afni_proc.py -dsets EPI*.HEAD \
-regress_stim_files stim_*.1D
-regress_basis 'TENT(0,17.5,8)'

Of course the wildcards should match the names of your files... :)

- rick