This is a relatively minor thing but has been around forever and is still there in v5 - the automatically set thresholds in Curate micrographs (which should be set to include all micrographs) actually exclude some micrographs - presumably a rounding error of some kind.
For example you can see here that the default lower bound for CTF fit res is 2.934 - but when this is applied, it excludes some micrographs that are right on the lower bound. This threshold presumably ought to default to something that keeps all micrographs/particles.
This is not a big deal, but it requires the user to remember to manually adjust this threshold, other they can lose mics that ought to be kept.
reading more carefully, you’re highlighting a different related issue in Curate Micrographs and my below discussion is about LIVE preprocessing.
doesn’t this occur when you change the threshold early, say to 2.934 (lowest observed at that time) and 9, and then collect more data that are below the value? For that reason, if I’m ever 1) not done collecting and 2) setting an upper bound, I always choose a very low lower bound like 1Å that won’t be reached. My suggestion to devs in line with yours is that if I didn’t SET a lower bound by changing the number, it should either default to zero or be variable to the lowest value that ever gets collected.
