biochem_fan

You can use the classic intel compiler that comes free with the oneAPI HPC toolkit. The new compiler (based on LLVM) should work as well but was not tested. Or,...

> Add low and high pass filtering options to the GUI This is on GUI when opening a single micrograph. I don't think this is often used with STAR files...

I don't understand your question. The shifts are in floating point numbers, not integers. So working in the original pixel size or the binned pixel size does not really matter,...

Are you aware that `data_local_shift` table from MotionCorr is never read by Polish but `data_local_motion_model` is used?

At the moment I don't have time to investigate this carefully so below is my random guess. Patch shifts are normalized to the image size, e.g. `const RFLOAT x =...

Aha, OK, this indeed sounds suspicious. I will investigate this later. Meanwhile, please don't worry too much. Bayesian Polishing is not very sensitive to initial values. When you use UCSF...

Brief notes on my investigation: - Indeed SVD is called in binned pixels when `early_binning` is ON (i.e. the default). - But if we change that, we have to fix...

Can you show us the full command line of `relion_project`? The program should not project beyond the 2D circle touching the box. This is fine because corners are masked in...

I reproduced your issue. The artifact is caused because the filament occupies volumes outside a sphere touching the box (see the first slice). If you make the box larger (e.g....

The above screenshot is with your command, `relion_project --i emd_22220_6apix_120px.mrc --rot 45 --tilt 90 --psi 45`. > I might hack the subtraction to pad the box in real-space Or you...