Hi Rick,
You are correct, we are not using afni_proc.py. It also occurs to me that we have multiband data, with a TR of .43 seconds, and improved spatial resolution. Not sure if that plays a role, but it is the first time I deal with this type of data and the first time I get such a low blur estimate.
The command we are trying to use is 3dFWHMx -combine -input errts+tlrc -out blur, for each subject. What mask would be appropriate to use if not the automask? If we ran 3dDeconvolve on data that we masked out per run and then using a combined mask tranformed to percent signal change, would that same combined mask be appropriate for masking in 3dFWHM?
I appreciate your help!
Ana