[Freesurfer] follow up on question asked during freesurfer course

Hi

To compare data different version of Freesurfer I was advised to bring it into the same version, and the subjects are the same. I used recon-all –make –all –s to convert from Freesurfer 4.30 to Freesurfer 5.1. The results are now more similar but there are some differences for weak correlations. When I use mri_glmfit I get a different output fwhm with the two different dataset using Freesurfer 5.1. Would you recommend me to rerun the converted subject using recon-all –all? I usually get similar result on hemisphere and results that are more similar than before on the other hemisphere. I am enclosing two representative jpg files. I have used smoothing at fwhm 60 mm.

Best regards Knut J Bjuland