Glacios + selectris + falcon 4
Added by Morgan Beeby about 4 years ago
Hi Anchi,
Well, I'm excited to report that after seven years, Imperial will upgrade from our venerable F20+falcon 2 (with which we've collected tens of thousands of tilt series using Leginon) to a Glacios with Falcon 4. I am hoping to also raise money for a selectris energy filter.
I wondered: is this likely to work "out of the box" with Leginon? I see you have an open ticket for Falcon 4; is the selectris something you might support too?
I also wondered if you think there might be any chance of implementing a protocol like Borgnia and Bartesaghi's recent "BISECT" beam/image shift tomography multiplexing? (https://www.biorxiv.org/content/10.1101/2020.09.24.294983v1).
thanks,
Morgan
Replies (2)
RE: Glacios + selectris + falcon 4 - Added by Anchi Cheng about 4 years ago
We have a working Falcon 4 implementation with MRC frame saving, but not EER. TFS is still working on the latter to be accessed through TEMAdvanced Scripting.
Currently David Mastronarde and I are working with TFS regarding access to needed functions for selectris.
Regarding the third question, if I am reading it correctly, the three things that they did are
1. Use sum of cosine-stretched lower-tilt images to increase correlation signal. This should not be hard to add. We are already using cosine-stretched single image in the correlation. Just need a cache to save the accumulated one.
2. Using the first few tilts to estimate the offset of eucentric to provide a better goniometer model and make defocus correction accordingly. This is already in leginon tomography.
3. At each tilt, many targets are taken with beam-image shift. This scheme makes it look more like RCT or MSI-Tilt rather than MSi-Tomography. This will take more time and planning to combine with 1 and 2.
Overall, yes, it is possible to do this. I don't know if I will have time to start this by end of year, though. My first priority right now is documentation for release 3.5
RE: Glacios + selectris + falcon 4 - Added by Morgan Beeby about 4 years ago
Hi Anchi,
Excellent. That is great news that we'll be able to hit the ground running with Leginon immediately; it will be delivered in the early spring (regardless of whether we have the selectris or not) so hopefully you will have that code in place before we sign off on it.
Yes, those are the crucial points of the Bartesaghi approach. I suspected it would be feasible. I guess there's not much we can use it for, yet, but I'd be keen to get back in touch in early 2021 to see if this is something you think you'd have capacity to work on!
best,
Morgan