Use T2 image to refine pial surface

On freesurfer documentation, I only find one simple command that seems to do all the work, but I’m wondering if there is a requirement for this to work? Does my T2 need to have the same FOV, slice numbers, resolution etc. Also, if I recon my T1 first, then did some editing on the mask/surface, after that do a recon on T2? Does my T2 recon take into account of all the editing I did earlier on T1 or it’s a completely new recon again?

FreeSurfer will align the T2w image to your T1-defined subject space. And you can just run with recon-all -autorecon3-T2pial to only redo the steps that change with the T2 added.