Hi Alfonso, 

I'm wondering if an issue with the patch you sent earlier is replicated in the latest release of CONN 19c. I encounter an error with the normalization process when I use the option "Structural Normalization preserving Grey/White/CSF masks". A handful of structural files (T1 and FreeSurfer ROI masks) end up oddly warped and poorly aligned with MNI space. The strange part is that these same scans normalize fine using the default preprocessing pipeline (i.e., using the simultaneous segmentation and normalization). You can refer to the images attached for reference. Even if I use this preprocessing step "Structural Normalization preserving ..." alone (to test whether the order of steps affect results), the resulting images are still problematic. 

Even more puzzling is that using "Structural Normalization preserving Grey/White/CSF masks" might also affect the functional images somehow because I see that my functional images fail to normalize when this step is included in the pipeline. First I tried with the option (1) Functional indirect normalization preserving Grey/White/CSF masks and then I tried preprocessing another time with the option (2) Functional indirect normalization. Again, I have no issues with functional normalization using the same dataset and the default preprocessing pipeline in CONN.

Could this be a bug with these preprocessing steps or am I missing something?

Paul