Leginon

Try disabling "Correct image" in the Acquisition nodes (Ones with a camera icon). However, I must warn against this practice for any MSI application. MSI applications calculate correlation in many steps and Leginon's correlation uses separate gain normalization for the two images that are correlated. This switch need to be managed by the database through Leginon. When you use external single gain reference, the results may not be as good when the correlation peaks are weak.

Anchi

"anchi" wrote: Try disabling "Correct image" in the Acquisition nodes (Ones with a camera icon).
Anchi

At least in the manual application, when I do so, I get the truly uncorrected image, even when gain/bias correction is turned on in the Tecnai UI and using Digital Micrograph directly would yield a corrected image.

(The uncorrected Leginon image looks exactly as the result in Digital Micrograph when I turn off gain/bias correction in Tecnai UI.)

We have never tried to use the corrected images from DigitalMicrograph, but you could experiment with it. This would require modifying the file: pyScope/gatan.py
Find all instances of "AcquireRawImage" and replace them with "AcquireImage" (there should be three)

You will also have to do as Anchi said and tell Leginon not to correct the images. The images in the database will be flagged as uncorrected, but I am not sure if this will have any bad effects anywhere else.h

Let us know if it works and we can make this an option in some future version of pyScope.
Jim

It works, thanks! Of course, the correction node prints error messages and some nodes in the calibration application do not even allow for uncorrected images to be taken.

-M

I hope you don't me hijacking this thread:

When we were obtaining our reference images in the correction node (under the calibration procedure) all of our bright references and some of the dark references had the CCD honeycomb pattern that normally shows up when you don't do any gain/reference correction in the Tecnai UI. Even though the images we got in the Tecnai interface were corrected normally (no strange artifacts were seen at normal exposures of ~300+ ms), the Leginon ones showed this pattern even at very low exposures (~40 ms).

We tried getting the reference images with the artifacts anyway, and it seems to be fine for acquiring data (no pattern after Leginon corrects them). I just wanted to make sure that this was normal since the manual says to pick an exposure where you don't see overexposure. As I mentioned before, this meant going to very low exposures, so low that the corrected images had higher standard deviations than the raws.

I'll see if I can get a picture to you at some point if it would help.

Thank you and cheers,

~Carlos

Carlos,

What camera is this? Dark reference is taken with the camera shutter closed but at the same exposure time as the bright. so its mean intensity should be lower than the bright reference and, as you point out, void of the honeycone pattern of the fiberoptics. It looks like that your camera shutter does not work right when dark reference is taken. We had that problem on an Eagle camera, I think. If it is a Gatan camera, you can watch the shutter LED light on the camera control box to see if the shutter still opens when the dark reference is taken. If you know that your camera shutter operates right in normal imaging (meaning that when you double the exposure time, the mean intensity doubles), but the camera won't take the right dark reference, you can close the column valve while you do this step.

Anchi