I used bbregister to coregister diffusion images with anatomical.

I would like to apply the bbregistration transformation to the white and pial surfaces

What I've done is:

I calculated new transformation  matrix using fs_read_dat.m and then apply it to the surfaces with mri_surf2surf.

When overlaying the new surface on the diffusion data- it seems that the transformation is inaccurate.(screenshot is attached)

Any idea why?.

Thanks, Shani