[Freesurfer] preproc inorm question

Hi all, If I were to perform a surface-based analysis, resampling to spherical space, and I did not want to smooth in the volume, would I need to preprocess my data again without smoothing or could I just specify the motion corrected volume in in a new analysis. I think the answer is: I could just specify the motion corrected volume, however, since the global mean for intensity normalization is calculated at the end of preprocessing after smoothing and exists in the fmcsm.meanval file, will there be a problem with the inorm in later stages, will it be performed etc. if I just specify the fmc volume in a new analysis? Thanks, Cameron