Distorted raw micrograph from Counted bin2x collection

I collected my data in ‘counted’ mode, bin2x, 130K magnification. When I import the raw micrograph (without gainref), it appears distorted-the image is divided into several stripes and the left side and right side are switched in every other stripes.


Also, I receive this error message:

[libtiff error] TIFFFillStrip: Invalid strip byte count 0, strip 12

May I kindly ask does anyone know what happened here and is it possible to correct it? Would it be because I should not bin2x under counted mode? Thank you so much for you help.

When you say “without gainref”, do you mean gain correction is already pre-multiplied into each micrograph? Or that you do not have a gain reference?

Because that micrograph looks like the latter, but the error suggests that the TIFF file is corrupted somehow. Also, I’ve never seen a K3 gain pattern that looks like that… the pattern is usually vertical from the datasets I’ve seen, which further makes me think the TIFF is corrupt.

Welcome to the CryoSPARC forums. :smiley:

Thank you so much for your reply! For ‘without gainref’, I mean this is the raw micrograph and I didn’t apply gainref to the micrograph.

If the TIFF is corrupt, is it an issue during saving the micrograph? In this case, is there any way to correct?

Thanks so much for your help.

You can use something like sum_all_tiff_files from cisTEM, or similar to make a gain reference from all frames of the micrographs. That may correct that odd pattern. But the TIFF error reported stiff concerns me. Could be a bug in saving, could be memory corruption, could be storage failing. Sorry for wall-of-text, using a phone right now. :sweat_smile:

Thank you so much for your advice! I also tried to apply gain reference to the micrograph in cryosparc, but it doesn’t help-the micrograph is still distorted. It might be as what you’ve said, that the TIFF file corrupts. In all the micrograph I’ve collected, I only see very very few ‘normal’ micrograph. May I ask would you have any idea of the possible way to recover these corrupted micrograph? Sincerely thank you so much for your help : )

Sorry, I’m not sure how robust multi-frame TIFF is on that front. One of the devs might have an idea or two, though…?

You file is divided into 12 “strips”, as also indicated in your error message, and for some reason every second one seems to be shifted by half the image width. To get back to the real image, each half of the affected strips needs switch position. Here’s a quick image edit:

Gain refs wont fix it. To me, it seems more like the file format is causing problems.

What is the rest of the output from the Import Movies job?
What format did you choose for the movies in EPU?
What size are the movie files?
Can you import the movies in Relion or any other software without getting the same offset error?