Coming into this project we did not have any experience in creating "archival
images." Our experience was in taking digital images with which to create a
visual record of items in Beloit College's Wright Museum of Art. Some of the
questions we had to answer before digitally photographing the Nuremberg
Chronicle include: What is the best camera we can buy on a fixed budget?
How to light the book and what type of lighting is "best"? Where do we take
the photographs? What file format do we save the images in? How do we support
the book while photographing? Our solutions to these questions, and how we arrived
at them, follows.
The camera we purchased was the Olympus C-5050 Zoom along with a one Gigabyte CompactFlash memory card (we also purchased a second CompactFlash card for reasons I'll get into later). We also used two sets of rechargeable NiMH batteries.
Learning to use this camera was not the challenge it initially appeared to
be. The camera did not come with a manual, but a CD-ROM from which you can print
the manual's 200+ pages. Once we spent some time with the camera we were able
to get it to do what we wanted. One thing that is nice about this camera is
that it has ten memories (called my mode) for saving various user-controlled
settings (e.g., white balance, macro mode, etc.). This was used so we could
quickly set the camera up for a particular type of shot as needed.
A nice thing about this camera is that it came with a remote control, so we did not have to touch the camera (and risk unknowingly changing our settings or camera aim) to take a picture. We would get the next page ready and then use the remote to activate the camera.
Each day we were photographing, the book had to be retrieved from the Morse Library Archives. We always handled the book using clean, white cotton gloves. At the end of the day's shooting, the book was returned to Archives.
Lighting the book and where to photograph it are actually parts of the same question. While all of the photos are taken with the camera mounted to a copy stand, we initially wanted to photograph the book using natural, but not direct sunlight. We had several reasons for doing this: to avoid using the camera's flash (it was never used), to protect the book from the excessive heat and bright light of direct sunlight which might damage the book, and to have more accurate color reproduction. We initially tried a location with large north facing windows in the basement of Morse Library using natural light, light from a 300 watt halogen lamp, and the two quartz lights attached to the sides of the copy stand. The copy stand lights were aimed up to bounce their light off of the ceiling. After viewing several test shots, we thought the lighting was too harsh and caused the pages to look washed out.
We then tried a location in one of the study rooms on the Library's second floor where we used only the halogen lamp and copy stand lamps. While we could control the light better in this location, we found that the walls (and ceiling) were not actually white as we had thought, but a very light pink. This was reflected onto the pages and picked up by the camera causing the images to have an undesirable pink tint.
We decided to move back to the original location, but with some changes. We would use large sheets of glassine paper to mask the windows directly adjacent to the copy stand. This was to soften the indirect sunlight. The copy stand lights would again be aimed up to reflect off of the ceiling. The halogen lamp would not be used. This lighting solution produced very consistent, "true" colors. We did find that on extremely overcast days, our picture color was not as accurate as when the sun was shining. On those days, we did not photograph.
When photographing a five-hundred year old book, one encounters unexpected problems. Our copy has been rebound at least once and has one break in the binding (at folio Cv). The book also does not open "square." As you look at the two-page spreads, one can easily note how the book slightly changes "shape" as you progress from front to back.
Another problem is that because of the age of the binding, the pages do not lay flat. To ensure a level surface, we had to support the book so that for the two-page spreads both pages would be level with each other. This required the use of supporting material under the book halves to raise the lower half to the same height as the opposing page. Some leveling was also done while shooting individual pages and images, because the book is thicker at the binding and did not lay level. The book was placed on a sheet of light blue/grey foam board with the supporting materials hidden from the camera beneath the book. To make sure there were no shadows, we stood on the side of the book away from the windows when a photo was taken.
The book was placed on a copy stand with the camera at a height of 37 inches for the folios and 34 inches for the single pages. The height was adjusted as needed for the individual images. The zoom feature was also used at a scientific "two clicks out" so the camera's all-purpose lens was not at either extreme of wide angle or zoom.
We tried various f-stop and exposure settings with the camera, but found that the "automatic" settings produced the most consistent (and best) images. Despite this, there were some settings which we wanted to control. The ISO was set at 64. The Macro setting was used for the individual page images. The Super-Macro setting was only used for the individual images.
Initially we photographed the folios in High Quality JPEG format, but we later re-shot the folios in TIFF format which produces a superior quality image. All of the images on this website that aren't two-page spreads were taken in TIFF format. One problem with TIFF files is that each image was very large (14.1 megabytes on average). Once the camera took a photo, it had to save it to the CompactFlash card; and with files this large, that process took about 40 seconds. When the camera was done saving to its memory card we often would have the next page ready for shooting.
One unavoidable problem is some camera lens distortion. Research showed us that all lenses have some flaws. We found that the lower right corner of our images tended to be slightly fuzzy (because most of the images were actually taken upside-down and rotated, the distortion is in the upper left hand corner). To compensate for this while filling the frame to take advantage of the camera's resolution, we shifted the image away from the part of the lens where it would distort. For the close-up images, we were able to avoid using this portion of the lens.
To avoid moving the camera, the CompactFlash card was removed from the camera while it was still in the stand. It was then put into a CompactFlash reader connected to the computer by a USB 1.0 port.
Downloading the images from the camera to the computer was the only bottleneck we ran into throughout the whole process. The problem is one of Compact Flash card speed and the speed of the USB 1.0 connection. It took c. 40 seconds to save the image to the CompactFlash memory card, and it took about the same amount of time to transfer that image from the card to the computer. Using the 1 gigabyte card, we would take about 48 images, filling the card and then download the images into the computer. We used a CompactFlash USB 1.0 connector to connect the chip to the computer. Downloading took about 40 minutes for 1 megabyte.
To speed up the process and not have us sitting around unable to work, we purchased a second CompactFlash memory card of 512 megabytes. While one card was being filled, we would download the other. This greatly improved our efficiency.
Once all of the images were saved to the computer, each file was renamed with the appropriate folio number and put into a folder with the other images from its Age. These folders were then copied onto both CD-R's and DVD-R's using Apple's iPhoto.
Initially, the two-page folio shots (JPEG file format) were more than 2000 pixels wide. In order to make them fit the screen of an ordinary computer with 1024x768 resolution, the images were cropped down to the size of approximately 950 x 678 pixels each. The entire procedure was done in Adobe Photoshop 6.0 using Action Tool and employing lighting and sharpening effects as well in order to make the images appear more vivid. The latter was done to facilitate an understanding of the graphic design of the pages.
Individual pages (TIFF file format) were cropped separately. Each page was then converted to Adobe Acrobat (PDF) format. Besides significant decrease in the size of the file, the PDF files make it easy for a visitor to view details of the page, zoom in and out, and print the page in high resolution to a standard 8 1/2" x 11" sheet of paper.
A select group of individual images was photographed as well. We decided to
keep the original size of the images (approximately more than 1500 pixels),
but at the same time offer a preview of the image in smaller scale. Thus there
are two options in the Images
section, "View", which provides a smaller image that fits within the screen,
and "Full Size", which describes the size of the photographic image (one that
is often many times larger than the original image in the book!) and is larger
than the screen.
Camera: Olympus C-5050 Zoom, 5 mega-pixel camera
Compact Flash memory - type I: 1GB and 512MB cards, USB 1.0 reader
Computer: Apple G4 800GHz, OS 10.2 with a 15" 800x600 VGA monitor
Imaging software: downloading the images from the CompactFlash cards into iPhoto and burning them to CD-R's and DVD-R's