I mentioned the hose rabbit-warren, right? What I did not mention is that I am also trying to tag, sort, and order all the images that have not been tagged, sorted, and ordered yet (and no, you do not want to know how many there are). Some are doubles that I can delete, some are easy to sort as a chunk (party photos, for example) and some... well, some are tedious.
One of the images was without a caption, date, tag or anything, and it has been hanging out there since it was used as one of the images in my thesis. The original manuscript is the Codex Vindobonensis 1857, or known today as signature Cod. 1857 HAN MAG in the Austrian National Library.
Back when I was writing my thesis, there was digitisation going on, but it had not progressed as far - so in some cases, the way you'd get to see pictures was through facsimiles. They are intended to be as much as the original as possible. For the thesis, I thus took the image from the book in the library in Bamberg; if I remember correctly, I requested the scan, got it on a CD, and used the image.
So picture my surprise when I wanted to tag this image today and thought the nicest and easiest way would be to find the entry for the now-fully-digitised manuscript in the Austrian Library... and found this:
This is... clearly not the same guy. First thing, my copy apparently got flipped at some point so I have the mirror image. And then - the position is almost the same, but the face is quite different; the hands are different. The folds in the garments are different, with the shirt or braies especially obvious. The leg position is a little different. And the points, the whole point because of which I was using this image, are different too. They are much darker and much more obvious in the facsimile version.
Now I am really wondering. Wasn't the facsimile supposed to look exactly like the original? Was there a second version of the manuscript? But then, why would the images be so close and yet different? How much egg do I now have on my face because of this? Questions over questions. I'm now planning to go back to have a look at the facsimile at some point to see if at least the flip only happened with the scan transferring to me.
And this, dear readers, is why one should never fully trust anything but the original itself, no matter what. Sigh. Also it's why digitisation is The Thing, because it means there's at least a few less steps inbetween - it's digitising the object, and then you can look at it, and if it's high quality you can look at it right down to the level where you see the pores in the parchment. So if you're pointed to a re-drawing*, or a reconstruction without access to the original... be wary.
*Re-drawings, in German "Umzeichnung" used to be a really big thing in archaeological publications, as you don't need a good quality image to use for your print prep and, even more importantly, you don't need to go and ask for publication rights. I don't know whether it was just laziness, or whether it was much more difficult than it is today (it was surely more involved than shooting an email!), or whether it was a cost issue either in printing the image or in procuring the rights, and it was probably a mix of those with varying proportions of the individual ingredients. Anyways this means that a lot of older archaeological publications are full of image references to show things related to the find in context, rendered as a little black-and-white line drawing (which admittedly was much, much easier to print in the text part). If you see these, I heartily recommend trying to find the original before believing all of it. Quite often, a digital image is available today with a little use of the search engine of your choice, and you can see how it really looked... because even if you try really hard, and do really well, a line-drawing will still be an interpretation.

