Berlin Papyrological Congress Presentation (15 August 1995)
by Robert A. Kraft, University of Pennsylvania

Manipulating Digitized Images of Some Unpublished Papyri
(University of Pennsylvania, Center for Judaic Studies;
 inv. 3.6 + 1.8 + 3.11 and 11.3b)

The primary purpose of this presentation is to illustrate how 
computer assisted manipulation of papyri images can be used to 
great advantage. 

The subject matter consists of four fragments discovered among a 
small miscellaneous collection of Greek and other papyri at the 
Center for Judaic Studies (formerly the Annenberg Research 
Institute, and prior to that, Dropsie College/University) at the 
University of Pennsylvania in Philadelphia, USA. Since the 
individual fragments have already been mounted in a manner that 
is not conducive to juxtaposing them physically, they are 
excellent candidates for the present purposes. Furthermore, since 
the entire collection of which these fragments are part has been 
photographed and placed onto Kodak Photo CD-ROMs by my students, 
it is relatively simple to isolate the relevant "digitized" 
images and manipulate them, much as a photographer would do in a 
darkroom. With the right computer equipment, this presentation 
would be in full color and the results could be transmitted to 
other suitably equipped locations throughout the world.

Three of the fragments clearly were once joined together to form 
a piece approximately 8.4 cm wide by 4.0 cm tall in its largest 
dimensions. Portions of 5 lines, apparently from a list of some 
sort, are written along the fibers on this reconstructed piece, 
with left and right margins represented. The fourth fragment, 
measuring about 4.0 wide by 3.5 tall, contains portions of 5 
additional lines in the same format, but does not seem to join 
the larger reconstructed piece. It might be possible to determine 
more accurately the relationship between these unjoined pieces if 
the fiber patterns on the verso were more clearly visible, but 
the current mounting employs a thin translucent sheet that covers 
the verso, frustrating careful examination by eye or camera.

When I first discovered, by examining photographs, that the four 
fragments were related [exhibits ## 0.1-0.4, colored photos (not 
reproduced for distribution)], I proceeded by making xeroxes from 
the photographs to cut and paste into the desired reconstruction 
[exhibit #1, xerox mockup]. This procedure was perfectly adequate 
for my immediate purposes, but it was awkward and inconvenient 
(in addition to being quite "low tech"), so I made plans to 
pursue this project more thoroughly on the computer as time 
allowed. 

One of the first problems I encountered in using the xeroxing 
approach was that the photographs were not all printed to the 
same scale. Each would need to be resized in order to display the 
joins and relationships accurately. A second problem was the 
difficulty of creating images that were not overly dark, or 
light. The xerox machines at my disposal were relatively limited 
in their abilities to adjust such features.

On the computer, however, such tasks as resizing and adjusting 
the visibility (as well as the colors) promised to be relatively 
simple, once the capabilities of the chosen software were 
understood and mastered. Many computer programs are available for 
such tasks, but I was fortunate enough to have access to Adobe's 
PhotoShop 3.0 for most of the work, as well as the shareware 
programs PaintShopPro and LView 3.0. Each had certain advantages. 
The appearance of Armin Lange's manual on Computer Aided Text-
Reconstruction and Transcription [exhibit #4 (not reproduced for 
distribution)] was encouraging, as was the short piece by Jim 
Marchand on "The Computer as Camera and Darkroom" in my OFFLINE 
37 column [exhibit #5 (available on the InterNet)]. 

Unfortunately, I cannot actually walk you through the entire 
process here, since appropriate computer equipment is not 
available. But I can explain briefly the stages, show some of 
the results on a color screen, and answer questions: 

1. Extract desired photos from the CD-ROM [exhibits ##3.1-4, 
black and white, lightened for display]. 

2. Align positions (rotate as needed) and sizes of each piece.

3. Create a layered computer file in which each piece can be 
manipulated individually as well as displayed in relation to the 
other fragments.

4. Preserve the desired final arrangement [exhibit #2.1-2].

5. If desired, import the images into text documents, or add 
textual materials to the images. 

Of course, the ultimate goal will include decipherment and 
publication in some appropriate form (perhaps electronically, on 
the World Wide Web). My efforts have not reached that point, and 
any suggestions you may have about deciphering the document 
would be appreciated greatly; probably the left column lists 
quantities, and the right has names. One of the obvious 
advantages of placing such things on the WWW is that many sets of 
eyes will be able to see what would otherwise be relatively 
inaccessible, and to contribute to its accurate transcription and 
presentation. 

//end//