Classic Computer Magazine Archive ANTIC VOL. 5, NO. 3 / JULY 1986

By Charles Mazel


Restoring rare documents with 800XL and Computereyes

Blending ancient binding techniques with modem technology: The Floger
Shakespeare Library restores the crumbling pages of rare literary manuscripts
with the help of an Atari 800XL computer.

For Frank Mowery, the master bookbinder and chief conservator at the Folger Shakespeare Library in Washington D.C., book preservation combines both art and science. He must blend a deep understanding of the historic methods of papermaking and binding techniques with up-to-date knowledge of modern chemistry. Even so, many of the conservator's tools are the same traditional ones that have been in use for centuries. Now, an Atari 800XL computer, teamed up with an inexpensive video digitizer, is making the process faster and more accurate.

I was in Washington on business and dropped in on Frank, an old friend, at his laboratory in the library basement. He was working on making a new binding for a first edition of the complete works of Shakespeare, valued at somewhere around a quarter of a million dollars.

Many of the old books and documents that come into the possession of historic libraries are in poor condition to start out with. Over the centuries, paper will deteriorate if not stored under ideal conditions of temperature and humidity. Even when stored with care, bookworms may attack, leaving holes throughout an entire volume. Any handling of the fragile, crumbling paper could produce even more damage. Before the delicate documents can be made available to researchers or put on display, they must be restored.

Leaf casting, the specialized process of preserving centuries-old books and documents, is a complicated art practiced by only a few preservation laboratories in the United States. The leaf casting process involves "casting" new paper to replace deteriorated gaps and holes, or to add a new border to stabilize a fragile page. The leaf caster itself is basically a large tank fitted with a filter screen. The tank is filled with water to a level that covers the filter screen, and the "leaf" (or page of a book) is immersed in the water and placed on the filter. To create a frame to mold the new border, plastic sheets are used.

The conservator than adds a slurry of paper pulp made from recycled handmade paper. The slurry is mixed into the water over the paper, and a suction pump is turned on, drawing the water out of the tank from the bottom. The plastic sheets and the page are impermeable, however, so the water cannot pass through them. All the water is drawn through the holes in the page and through the border, carrying the suspended paper pulp with it. Trapped by the screen, the pulp is forced to accumulate where it is needed, and the fibrous pulp bonds to the edges of the old paper. After the water has been drawn out of the tank, the page is removed, placed between sheets of absorbent paper, and subjected to high pressure in a press, squeezing out most of the water. After the page is has dried slowly in the air, it is safe to handle or rebind.

The most critical phase of the process is adding the right amount of paper pulp slurry to the tank. If too little or too much is used, the "new" paper will be a different thickness than the "old" paper. The ideal thickness is typically four to eight thousandths of an inch.

In order to add the right amount of pulp, Frank must know the volume of paper he wishes to create. He can accurately measure the thickness of the page with a micrometer. But his problem was in measuring the area in square centimeters. With a few holes, a ragged edge and a reasonable sized border, he might need to produce well over one hundred square centimeters of new paper for a single page.

Frank's old method was to place a transparent centimeter grid over the page and count the number of squares that were over holes. This was inaccurate because of the uneven edges, which meant a lot of guesswork and approximation had to take pace. More importantly, though, it was an incredibly tedious process. To treat an entire book, it would have to be repeated, by hand, hundreds of times. This is where the Atari comes in.

Frank is always looking for better and faster ways to do things. Though he designed the leaf caster himself, he doesn't know much about electronics or computers. He had an IBM PC at the lab, but used it only for word processing.

While I sat back at a respectful, and safe, distance and watched Frank restore the works of Shakespeare, I started thinking of a way to speed up the process. First, I thought about paired arrays of light emitting diodes (LEDs) and photoreceptors placed above and below the page. The page would block off some of the LEDs, and it would only be necesssary to count how many receptors were turned on. But Frank was hoping to discover a way to treat pages up to about 60 by 90 centimeters. This would take quite a few components to achieve any accuracy-not to mention the wiring time.

I left Washington with Frank's request in the back of my mind. Sometime during the next few months I hit on the idea of using a video digitizer to capture an image of the page. I figured that if you put the page and a border on a light table and produced a high contrast image, you could probably adjust sensitivities so that the page would be black, and the space between it and the border would be white. It would then only be a matter of counting white pixels and comparing this to the number of white pixels produced by digitizing the border alone without a page to be treated. The digitizing process would not have to be very exacting, since I only needed a two-level output and was not concerned with shades of gray.

So, I started looking for video digitizers for the IBM PC. They would do more than I needed-1/30th of a second frame grabbing, 16 levels of gray-and they were expensive.

Then, at a computer show in Boston, I discovered a video digitizer aimed at hobbyists-Computereyes by Digital Vision, Inc. It was slow- taking six seconds to capture an image-but those Shakespeare manuscripts had been around for hundreds of years and weren't going anywhere fast. Best of all, the system was cheap, and would run on low-cost computers, including the Atari 800XL. For half of what the cheapest IBM PC digitizer would cost, the Folger Library could buy a Computereyes digitizer, video camera, Atari computer, disk drive and monitor-all the hardware necessary for a stand-alone system. I wrote a proposal to Folger which was accepted.

I was able to use the non-protected Computereyes software as the basis for a customized system. The main program is written in BASIC, with machine language routines for the actual image digitizing. This made it easy to modify the menu and add subroutines for performing the final calculation of needed slurry volume.

For the area-measuring application, it was necessary to perform one more step not normally done by Computer-eyes-counting the number of pixels in the digitized image. With the Atari's 320 x 192 high-resolution display, 61,440 pixels must be accounted for. A BASIC routine for counting the bits would have been painfully slow, so I wrote a machine language routine that takes about one second, and added it to the Computereyes routines.

The result is a stand-alone area-measuring system that was delivered to the Folger Shakespeare Library in October 1985.

I used the Computereyes system's ability to store and retrieve pictures to disk to add a nice touch to the program. When you first load the Folger leaf casting program, a high-resolution, 8-level image of William Shakespeare slowly fills up the blank screen. To the conservators-magic.