Sunday, November 22, 2009

Virtual Unrolling


I started this blog because of a single idea. But, I'm very happy to say, someone else published the idea before me, and actually did something about it. The right thing.

In 2005, I saw a preview for a PBS documentary touting new technology used in reading the Papyri at Herculaneum -- I'd hoped it would show some non-invasive technologies. But, it turned out, those researchers, though using multispectral imaging to better see inks, were still likely to destroy the integrity of remaining ancient scrolls, carbonized by the eruption of Vesuvius, using specific restoration techniques that had proven inadequate for centuries. Full texts would continue to be reduced to mere fragments.

As you can see from my 2005 post, I thought I'd coined a new phrase: "virtual unrolling" (or its inflection "virtually unroll") ... Luckily, for the literature of the ancient world, W. Brent Seales had already used the phrase, in a 2004 paper at the 4th ACM/IEEE-CS joint conference on Digital libraries, already far along on his quest to do exactly what I described a year later.

Now, I think any computer programmer with graphic, modeling or other experience in computational geometry would have similar thoughts if they look at this Herculaneum scroll unroller at the National Archeological Museum in Napoli (well worth the visit: an extraordinarily collection from the ancient world). It just makes you wonder: "could we do that virtually?"

But by 2004, Seales, a computer scientist at the University of Kentucky, was already doing projects a geek archaeologist could only dream about: for example, going to Venezia to 3D surface scan the oldest complete manuscript of Homer (here's a lovely, romantic mini-documentary, the Wired article, and a dreamy online browser for the Venetus Manuscript).

Right now, Seales' team is right in the middle of decoding and modeling their first Herculaneum papyrus scroll (here's a video of the high-res CT scan). The stoa.org blog is a good place to keep track of developments with the EDUCE project, but they've put together a very nice documentary about the research so far.

So, what is on these scrolls?

We don't know, which is quite exciting!

It's a library, apparently selected by an Epicurean philosopher and poet, Philodemus. I'm very fond of Epicureans, the gentle, radical idealists of the ancient world, so I have high hopes for the quality of scrolls there. Works of Epicurus himself have been found among these scrolls: unfortunately fragmented by the old process. But with hundreds of scrolls still to be virtually unrolled, and perhaps more buried still, we could in the future find lost works -- Aristotle's dialogues? The lost plays of Sophocles? -- in their complete ancient form, thanks to the persistent efforts of these Greek-loving geeks.

Sunday, November 04, 2007

Scanning artifacts

Digimorph is an online repository of CAT scans for paleontologists.

Why hasn't something similar been done for human artifacts? Breaking open an oddment is the surest way to lose important information about it.

How about field CAT scans of sites? We could 3-D map entire sites before excavation.

Wednesday, April 06, 2005

Virtual unrolling of a papyrus scroll


The House of Papyri is a library of burned Roman scrolls of epicurean philosophy, found in a palace at Herculaneum, orginally thought to be charcoal, eventually unrolled, and found to contain new quotes from poets as far back as Sappho. A new show on this subject just aired on PBS, funded by BYU.

The show highlights the use of UV to reveal ink, which is otherwise invisible, on the charred pieces of papyri. But the text from these papyri are all impossibly discontinuous, wih more than half of the material missing, because of the damage during unscrolling. The stuff just turns to dust under any stress. This makes normal reading impossible, in the sense of sitting down with a coffee & a cat on a comfy chair and reading your fave philosopher. The epicureans would be furious.

I wonder why conservators unscroll them ... Isn't there some high-resolution CAT scanner that could make a model of a unscrolled roll? The model could be "virtually unrolled", with no text lost. There are many scrolls unopened still. Anyway, I googled CAT scanner & papyri, and only found a thread about reading papyri after the ink has disappeared, through depression scanning (which is interesting). But it looks like no one is trying to "virtually unroll" these priceless Herculaneum scrolls. Of which there might be thousands more buried at the site.

Since the ink & the carbonized paper react differently to different frequencies of light, it must react differently to x-rays. If one can adjust the spectral absorption of a CAT scanner (can one?) then one could get the model of floating letters in a single session. I'm pretty sure there must be someone with this kind of equipment already ... just can't quite find it.

Had to tell someone this idea -- nerdy ebullience can't be capped. I'll get back to work.