Actually, just noticed that this is an option in selxavg3-sess, never mind! 

On Wed, Jun 7, 2017 at 9:07 PM, Taylor, Johnmark <johnmarktaylor@g.harvard.edu> wrote:
Hello,

I had a quick question. Is there a way to make FSFAST output a 4D volume of all the residuals? In other words: is there a way to see, for each voxel and timepoint, how much of the BOLD response was unexplained by the GLM? I know that it outputs a volume containing the residual variances, but I am interested in looking at the residuals in more detail. I see that there is a directory called "res", but for all of the GLMs I've run it appears to be empty...

Thank you very much,

JohnMark