Homogeneous refinement - Mask issue

Dear,

I did all particle ab-initio model followed by homogeneous refinement. In the end-result of homogeneous refinement, I see the GCFSC resolution plots with and without auto-tightening. Without auto-tightening the “corrected” and “tight” plots are deviating very much at the horizontal line around 0.15 correlation coefficient. But they are not much deviated in the auto-tightened plot. I had plans to do 3D variability from this result. I am not sure of the mask. The mask is not as I expected as well. I feel that it is cutting some features, though preserving major features of the molecule. Can I please have your advice to go about it?

Best regards,
Iswarya