f116v squiggle from the multispectral images

[oshfdk]

I think I need to clarify this: if we assume the whole thing is a single feature (both black pronounced strokes and wide blurry strokes are part of the same inscription or stamp), then the blurry parts are actually physically blurry. As far as I understand the whole pipeline from the imaging to processing, no feathering/dithering/averaging at this scale could be introduced at any stage. So, the ink or whatever left this shape was actually spread on the page by something. Also, the fact that the models pick the ink-like noise all over the bottom right of the page could suggest that traces of actual ink happen everywhere, which may be consistent with wiping off some fresh inscription with a wet cloth or something like this. I wonder, if it's possible to remove fresh ink from vellum by wet wiping? Especially if the vellum was already wet in the first place.