Leginon

There is an ace2.exe in appion/bin directory which would have been installed to /usr/local/bin with standard installation. If it does not work, the instruction is at

http://emg.nysbc.org/projects/appion/wiki/Install_Ace2

Beam Tilt Image node can function without ACE2.exe, though. You will still be able to create the Tableau just without defocus labels on them. You will still be able to click a point on the image display to move the beam tilt.

I assume that you are doing this in preparation for image-shift-coma calibration. Unfortunately, I will have to say that it might fail very easily. The calibration needs so many rounds of ctf estimation on images with least amount of imaging area (typically 512x1 or 512x2), a few failure in the process can give you total garbage calibration. I need to write something to remove bad ctf estimation, not just doing average, as it is now. Until then, I would recommend against trying it unless you can verify the results.

Hi Anchi,
Has this issue been resolved so that we can install ACE2? It would be a valuable tool to be able to create a Tableau. Thanks.
Bob

Bob,

You could have install ACE2 even before and then have the estimated defocus for Beam Tilt Image Node. The part I recommended against is the "image-shift-coma calibration".

Tableau in Beam Tilt Image Node should be able to create tableau without that calibration. Did Beam Tilt Image Node not work with and without ACE2 installation I mentioned in the last post?

Hi Anchi,
Sorry I meant to ask if it is possible to calibrate image-shift-coma with some level of confidence. Have you created something to remove a bad calibration. We will try installing Ace2 and wait for your verification that the appropriate removal tool is available before calibrating.

This is embarrassing, but I have not managed to do anything about creating a tool in Leginon to remove the bad measurements used in the image-shift-coma calibration. Our quest to high resolution (4A) has not gave us significant improvement even when I knew I had a proper calibration and correct the beam tilt induced by the image shift.

Looking back your original message, you got the error message in Beam_Tilt_Image Node, not Beam_Tilt. I guess you were not planing to do image-shift-coma calibration but just wanting to do coma-free beam tilt alignment using the tableau. Am I right? My misunderstanding on which node your message is about shifted our conversation to image-shift-coma calibration, a much more complicated issue.

I tried doing a beamm-tilt coma calibration in the beam tilt node of the Calibration application and got the following error
Calibration failed: can not concatenate tuple (not "float"} to tuple. I am not sure if this is related. This is on our Polara with a TVIPS 415 camera.

I can't tell where the error is from. Could you turn on debug logging for Beam_Tilt Node (under the menu bar "Other") and then give me a screenshot of the logger and the image it fails on? The algorithm I have in there is quite picky and has not been tested on many scopes. If it fails, it would be because it is out of range of what it is capable of doing. What I have it working before, the first node of CTF was ~1/3 Nyquest with the image at 512.

I tried a few CCD, defocus, sample and Magnification configurations and still get the same error reported. See: Beam-tilt Coma calibration failed thread. I placed the screen shots there. With our F20 with a Gatan camera I can get it to work but not with our Polara with a TVIPS camera.

Let's close this thread as it is no longer an ace2.exe problem.