Hi Bruce, 

Here is what I did. For each subject:
a) used the original (non-skull stripped) brain for the 001.mgz file as input - located in the Subnum/mri directory
b) used mri_convert to get the afni skull-stripped brains and renamed them "brainmask.mgz" , located in the Subnum/mri directory. 
c) Then I ran the following command:

recon-all -all -3T -mprage -s processed/Subnum


I get the following error:
 --------------------------------------------------                                                               
#@# Mask BFS Sat Aug  5 01:31:09 EDT 2017
/vega/psych/users/mrv2115/fsdata/processed/SB002/mri

 mri_mask -T 5 brain.mgz brainmask.mgz brain.finalsurfs.mgz

threshold mask volume at 5
ERROR: dimension mismatch between source and mask
-------------------------------------------------

Can you please advise?