Leginon

Similar look does a digital zoom at the same dimension as full camera but at lower binning if possible.

For example, if your align preset is 10x lower in magnification then the reference preset on the left. The full camera image will cover 10x area and a feature on it is quite small.
Let's say that this image is at 1024x1024 binned by 4.

With similar look activated, it will try to take an image at 1024x1024 binned by 1 centerred at the middle of the camera in this case since it is not possible to take a binnined smaller than 1 to get 10x zoom.

Because the image is at different dimension/binning/offset than the one specified in the preset, Leginon complains that it could not find references. It should still taken an image at lower bin, though. Is this on K2? K2 binning works different from other camera, I will need to investigate further.

Thanks for the explanation! For some reason it works now. I was also wondering about 2 other things:

1) I am consistently 2um more under focus than what I set it to. If I try to correct it decreases with 2um more, and so on. I do set eucentric focus and rotation center in the beam tilt node before starting the session.

2) It is really hard to do the align presets between ed and fc (4700x and 9700x on microscope, which is ~28000x and 58000x on k2). There are no mags in between, we do not have the EFTEM package for the GIF mag correction yet. Any suggestions?

Thanks!

(1) The offset may come from the fact that your autofocus and actual exposure are not at the same magnification, although that should be corrected easily if you give the exposure the same amount of offset in the opposite direction. The other possibility is that autofocus algorithm can not tell it apart between 0 and 2 um defocus. This more likely happen if your fa preset mag is on the low side.

If you can get reasonable autofocus result at the mag for the final exposure, you can move your fa preset there, then there won't be the offset.

(2) Preset alignment between ed and fc is not super important. As long as it is not so bad that you end up focusing inside a hole that you plan to image when you aim for the carbon support, you can just accept the same image shift that your procedure alignment gave, i.e., the same 0,0 user image shift.

Thanks, yes, I had to change image shift for ed to 0,0 and it worked. Now all presets are aligned. However, it seems that the exposure and focus target that I set end up being 90 degrees off. I target specific features in the hl image, but when I check the hl image afterwards I can see the burn areas for exposure and focus targets being 90 degrees turned. I know there is a 90 degree rotation between 350x in LM and 4700 in SA mode on the scope. How do I correct for this?

This is an important question so I started a different thread http://emg.nysbc.org/boards/6/topics/2059