Different parchment color through Herbal section?

[Jorge_Stolfi]

The new method was to take "400 random patches of clean parchment, about 2 mm square, from anywhere on the page. A patch only counted if it was uniform and at least about 4 mm from any ink or paint. That’s about 48,000 patches in total." The results matched, and became even a bit sharper.  ...

The two sides of the vellum are different, even though the finishing by the maker tries to make them as similar as possible.  Vellum is very light when freshly made, but gets darker (browner) with handling and exposure to light, moisture, and air.  Some of the chalk paste that was used to fill the hair pores and smooth out the surface will fall off as the vellum is bent.  Dirt and mold will accumulate in the microscopic crannies, and the vellum will  become impregnated with sweat and grease from fingers.  The pigments from paints and ink will come loose with friction and some of them will contribute to the dirt.

This browning will be different on each folio of a bifolio, and each side of the same folio, and each place on the same page.  You can see this variation on any page.  For example, when a folio is a bit wider than one of its neighbors, the exposed edge is usually much darker than the rest of the page. Also You are not allowed to view links. Register or Login to view. is much darker than f1v, f8r, f8v; likewise You are not allowed to view links. Register or Login to view. is much darker than f116r, f103r, and f103v.  Also f2r+f2v are darker than f7r+f7v,  presumably because they have been pored upon ten times more often.  

Since you now have a slave that you can abuse without scruples, you could perhaps ask him to redo the scatterplot with a separate dot for each panel (each side of each simple folio, or each section of each side a fold-out folio).  Maybe with lines connecting the two sides of each folio and/or the two or more panels on the same side of each bifolio. 

To get around the "dirty edge" problem, you could ask the slave to take, from each panel, as many samples of blank vellum as possible, and use the one that is lighter than (say) 90% of the others.  That should avoid the darker edges and the occasional lighter outliers.

However, it seems that the Beinecke color images were taken with a consumer or "professional phtographer" camera.  It is possible that the camera and/or the post-processing did some overall color and brghtness adjustment, independently for each image.  These adjustments would have affected even the "clean" vellum areas, depending on the amount of "dirty" vellum and text+dtawings visible on that page.  if that is the case, comparison between different pages will be problematic, to say the least.

The Multispectral images were taken with a technical camera so that the value of each pixel is strictly proportional to the intensity of the light scattered by the object towards the camera.  The sensitivity and exposure time were set separately for each image , but each image includes a reference white patch ("Spectralon") and a gray scale, which allows different images with the same wavelength and direction of light to be meaningfully compared.  Unfortunately there are only 10-20 pages in that collection (and, like the McCrone samples, they were carefully and perversely chosen to exclude precisely the pages that I most wanted to see   )

All the best, --stolfi