Leginon

We often setup RasterFinder as "Hole Targeting" node in now a day because of the use of much lower magnification in hl preset. It works well in avoiding overlap even if we might miss a few on the edge of the hl image. For screening, we just use one single target at the center of the image with this node. There is enough randomness of square target accuracy that we do get a bit of variation relative to the grid square.

All TargetFinder nodes have random sampling feature now, so you can also let it find more places and then sample one from the ones passes ice thickness criteria. There is also a new option to "apply Ice thickness filter to the template-convolved acquisition targets" under Target Template section. The latter option comes handy to pick non-fully blocking region which is problematic for Ptolemy.

You can of course use the traditional JAHCFinder (Template Hole Finding). We now typically make sq image to cover the whole square in LM. I can set template diameter large enough that it has a peak at the center of the grid square as a result. I then set convolution template from there.

Thanks Anchi!