Reference images saved as a different instrument than normal acquisition

[leginonbot]

---

Author Name: Anchi Cheng (@anchi2c) Original Redmine Issue: 2516, https://emg.nysbc.org/redmine/issues/2516 Original Date: 2013-09-11 Original Assignee: Anchi Cheng

---

This problem happens at FSU where Krios has different magnifications at different kV (or is it just energy-filtered (EF) and non-EF?) Scott made two instruments and switch them instrument.cfg from what I understand.

http://emg.nysbc.org/redmine/boards/6/topics/1667 describes a problem where the references can not be found when Leginon is restarted but o.k. if references were acquired in the same start-up.

This means that Leginon can find the references while they reside in memory but can not find them through database query.

I looked at their database, and found that the problem is limited to the instrument named Tecnai but not the one named TecnaiEFTEM. The references saved while as Tecnai (these are at 120 kV HT) are incorrectly recognized as for TecnaiEFTEM (seems to be those at 300 kV).

Now we know what the pattern is, we will need to find a solution

[leginonbot]

---

Original Redmine Comment Author Name: Anchi Cheng (@anchi2c) Original Date: 2013-09-12T02:53:19Z

---

Scott,

Now that we know why references could not be found, I will need to know what you changed, if any, in pyscope, leginon, and what exactly you do to trick Leginon to think they are separate instrument.

Let me know if I should get these details from Duncan, not you.

[leginonbot]

---

Original Redmine Comment Author Name: Scott Stagg (Scott Stagg) Original Date: 2013-09-12T11:06:02Z

---

Hi Anchi,

I've added my modified tecnai.py and gatan.py to this issue. In the presets, we use Tecnai and Gatan for non-filtered imaging and TecnaiEFTEM and GatanGIF for energy filtered imaging. This system has worked for quite a while. I'm not sure why it stopped working. I did recently update Leginon, but its hard to say if the problem cropped up after that. The problem is happening consistently now, though.

[leginonbot]

---

Original Redmine Comment Author Name: Anchi Cheng (@anchi2c) Original Date: 2013-09-12T22:51:00Z

---

I tried it on our F20 but by using subclass

class TecnaiTest(Tecnai):
    name='TecnaiTest'

instead of copying the whole code. It worked fine even if I leave both classes in instruments.cfg

Since I can't reproduce the bug, I will you or your student to try out the following:

The instrument configuration settings gui in Correction node often get stuck on the previous settings. If you go into the settings, and go through the routine of selecting a different camera (None in this case), and then select the one you really want, it will get it right. It is a timing issue. See if by faithfully doing so if you still have the problem. Check the result with this python script instead of restarting Leginon to save time:

form leginon import leginondata
normdata = leginondata.NormImageData().query(results=1)[0]
print normdata['scope']['tem'].dbid, normdata['scope']['tem']['name']

If the name corresponds to what you want, then it does work.

[leginonbot]

---

Original Redmine Comment Author Name: Anchi Cheng (@anchi2c) Original Date: 2013-09-13T04:32:37Z

---

fixes for bug #2517 might fix this, too.

[leginonbot]

---

Original Redmine Comment Author Name: Scott Stagg (Scott Stagg) Original Date: 2013-09-13T14:53:16Z

---

Your suggestion worked. We manually changed the scope to None then back to Tecnai and the camera to None then back to Gatan, and it subsequently fixed the db query. Your bug fix for ref#2517 permanently fixed the problem. After updating, we no longer had to switch the scope and camera to None and back.

[leginonbot]

---

Original Redmine Comment Author Name: Anchi Cheng (@anchi2c) Original Date: 2014-01-13T02:40:15Z

---

No one to review. Tested.