Ewald sphere correction - reconstruct vs global CTF


I was testing the Ewald sphere correction by running two global CTF refine jobs - one with it set to positive, and one to negative curvature. I then ran a Homogeneous Reconstruct job for each, and was surprised when the results were identical.

I realized that Ewald sphere correction defaults to “off” in this job type, even if the input particles came from a job where it was on. Perhaps to minimize confusion, it would be best to “inherit” Ewald sphere curvature settings from the parent job, if set?

Also, for the standalone Global and CTF jobs, it would be nice to have an optional setting to also reconstruct & calculate FSC, so the user can judge if any improvement has been made without performing a separate reconstruction job.



Hi @olibclarke,

This makes sense – so if we store the curvature sign with the particles, once you’ve reconstructed both + and - curvatures and know the sign, then you can just use the downstream particles from that job and Ewald sphere will always be applied in any job that supports it. Will make a note for this!


1 Like