This is a screen print (originally from Arroyo et al, a dodecamer i call #51) boosted to 300ppi in photoshop, and a 3px gaussian blur added, minus-8 points in contrast, opened in gwyddion and a limit range filter applied… 100-220.  It might be my imagination, but i can see very clearly the peaks along the collagen-like domain and even details in the CRD which look so like they could be the three arms of the trimer spread apart at those ends of the molecule.  The glycosylation sites (near the junction of the N termini in this dodecamer) have texture and shape as well, possibly providing information on how many of the arms of each trimer are actually glycosylated.

There is a twisted look (as one would totally expect along the collagen like domain that is distinctly present in two of the four trimers.  Fanning out of the CRD in the lower right portion of the image is pretty amazing. In addition, the one tiny little peak that I am hoping to verify on the downslopes of the N termini combined peak is pretty nicely seen on the left middle of the center of the dodecamer.  There is a little stretching on the trimer on the left center….  moving the glycosylation portion to the CRD just a little far from the N termini group… it is possible that is why the minor peak on the downslope of that side is visible.

Also interesting in this image is the relative decrease in peaks in the collagen like domain in the area past the glycosylation peak but before the neck and CRD (where the twisty look is also evident).