Project

General

Profile

Returning Tecnai to z=x of parent image xyz...

Added by Matthias Brunner almost 14 years ago

Hi there,

I select for instance 30 squares in the atlas, submit them in order to queue up subsquare targets. (So that I can check each square at sq magnification and handpick z-focus points).
To speed things up, all of those sq images are acquired at the same z height (usually in a way that the first square picked is at eucentric height).

Of course, I pick the squares in a way the stage can move from neighbour to neighbour, minimizing stage movement...eucentric height difference between neighbouring squares is minimal (usually a z change of 1-2um is required).

Unfortunately, the subsquare node "remembers" the z height at which the sq image was acquired for queueing up targets (Returning Tecnai to z=x of parent image xyz...). This is not the eucentric height of the current square but was just the z height the stage had when queueing up the targets. This causes an unnecessary zig-zag z movement of the stage, leading to much more z movement.

How can I tell Leginon not to recall the z height of a parent image (in this case when returning to queued up subsquare targets)?


Replies (2)

RE: Returning Tecnai to z=x of parent image xyz... - Added by Anchi Cheng almost 14 years ago

With queuing, Leginon always returns to the parent image z height because if the grid is bent and the parent image is at high mag, it can miss the area the parent image was taken at completely during the target adjustment step. This is why z-focus node is important. When most people use queuing, they do not carefully pick only neighboring square as you do, and the grid can be very bent, so we did not have an option to not use parent z-height because it can go wrong easily and badly.

If you want to rely on the z-height of the last subsquare z-focus, you should not use queuing. In this case, the sq image will not be taken in advance so that it is less likely to drift away in both xy and z so Leginon assumes no target adjustment is needed and uses current z for the imaging.

On the other hand, unless you intend to skip z-focus for all subsquare targeting, whether the square is at the parent height or the current height does not impact the data collection efficiency since the procedure they go through are the same. I would stongly discourage you to skip z-focus because if the squares you selected have an overall 20 um z difference, your procedure would force your defocus adjustment to accumulate that much, the beam will be shifted at the
higher magnification and you may not be aware of it at all. We have seen that during early development.

RE: Returning Tecnai to z=x of parent image xyz... - Added by Matthias Brunner almost 14 years ago

Thanks for the detailed explanation!

I really like looking at the square in sq first before submitting that is why I have to use queueing. Found a solution: Whenever I will submit many squares, I will hit "simulate target" in the z focus node, say, every 5th square...This should also help.

Thanks again.

    (1-2/2)