Hi,
I have done a local refinement job according to the case study. And all the jobs run successfully, but the final local refinement map seems weird with the subtracted region still exist. Although the map I got have a better resolution in local refine region, the left region which supposed to be subtracted also showed with a same signal strength as the local refinement region. Did I do something wrong? And what I should do ?
Thanks!
Best Wishes!
Wendy
Hi @Wendy,
Thanks for reporting. It’s possible that the scaling of the mask isn’t calibrated to the range [0, 1]. Did the mask come directly from a Volume Tools job, or an Import 3D Volumes job? In either case, would you be able to attach a screenshot of the mask plot that was plotted in the source job?
Best,
Michael
Hi Michael,
Thanks for replying! Mask came from Volume Tools job after Import 3D Volumes job. Is that the screenshot you mean?
Hi @Wendy,
Thanks for attaching the screenshot, the scaling of the mask looks good. For best results, low-pass filtering the reference to around the reference’s gold standard FSC resolution is likely to help (this can be found in the final iteration of the source refinement job). More importantly though, the mask used for subtraction should be softly padded. The mask in the screenshot appears to have a hard edge, and this can cause strong ringing in fourier space when the mask is applied to the reference, which might distort results. In our Mask Selection and Generation guide page, we recommend that soft padding is applied when working with masks. This can be done by clearing and re-running the Volume Tools job, and setting the “Soft padding width” parameter to something around 7-10 pixels. Let me know if you are still having issues after trying with a soft mask!
Best,
Michael
Hi Michael,
Thank you very much! Your suggestion works fine! 
Best，
Wendy
