The Edge of Logos

When I got into photography as a teenager, I had a 35mm SLR that was great for emulsion photography. Because of that, I gradually developed an archive of thousands of print photographs over the years from art projects, international travel, parties, weddings, and so on.

Then I bought my first digital camera in 2002 and abandoned 35mm photography. So a that point I began a new and rapidly growing archive of digital pictures. The advantages of digital photography are legion and at this point, well known I think to everyone. The marginal cost of taking and sharing digital photos is zero. Digital photos are nonrivalrous; I can give you an identical copy of my photo and still keep mine. They can be edited, modified and enhanced in Photoshop, and the creative possibilities are vast; they can be integrated into video or blogs, for example. The photo quality does not degrade over time. They take up no physical space. And so on. So I wanted to unify the two collections into a single comprehensive digital archive containing every picture I owned. In short, I wanted to scan those old photos.

Jesse Bushnell, my maternal great-grandmother. Denver, 1924.

So in 2002 I started thinking about the best way to digitize the prints, but I kept the project on the back burner. I knew it would be a big, time consuming project, and I wasn’t wrong about that. (It’s 2009 and I just finished!) Plus, I had concerns about the image quality of the digital photography of the day, which I assumed would improve over time. And while it was important to me, it wasn’t high priority. I knew I’d get to it someday.

Then, in 2004, my dad approached me. He had a giant collection of family photography, thousands of photos, mostly snapshots, spanning multiple generations over a century. He asked me if there was some way I could convert the collection to digital format. I told him there was and that I’d start working on that, because I saw it as the best way to solve a few important problems:

• Most of the photographs in that collection are unique and irreplaceable. The negatives for many of the pictures are gone, and if any prints were destroyed, it would be a total loss. By archiving them digitally, everyone in the family could have a copy of the entire archive, thus making destruction unlikely.
• While black-and-white emulsion photography stands up to the test of time fairly well, color pictures decline noticeably in just a few years, especially if left in sunlight, and especially if the print employed an older chemical process. Digital photography can’t lose quality unless the picture itself is damaged by physical data corruption or buggy software. Any decent backup strategy will prevent that.
• Photographic prints have poor metadata, so to speak. Most pictures don’t even have a timestamp, let alone any description of the event or people involved. If you get lucky there might be a terse description on the back. As time progresses, much of this information is forgotten and lost, especially for the dead. Photographs become mysterious or even meaningless.

These needs overrode my last objection to going digital: the possible reduction in image quality. So I decided to move forward. In starting this project, I sought to achieve these objectives:

• To have a complete digital archive containing every worthwhile photograph
• To make the scanned photos at least as high quality as the photos my digital camera can produce
• To capture as much relevant metadata for each picture as possible
• To store the pictures in an open, unencumbered, widely supported, high quality, compressed image format
• To preserve the photographs over a multi-generation (say 100-year) horizon
• To do all this in a reasonably timely and cost-effective manner

To do this project, I would need only three tools: a decent computer, a flatbed scanner, and image editing software. At first I used a Umax Astra 2100U scanner, though that scanner eventually died and I replaced it with a Canon LiDE. Later I used the scanner on an Epson all-in-one printer. The computer was a flat screen iMac running Mac OS X, although I went through several computer upgrades to what I now use, a MacBook Pro. I started with Photoshop 7 to post-process the scanned images, now using Photoshop CS3. This solution works well… but it doesn’t scale, not the way I needed it to.

Jesse MacKrell, my maternal grandmother. 1917, age 2 months.

I took that approach, which I’ll describe in detail below, to digitize thousands of photos. After working on this project in dribs and drabs for four years, however, I began to realize how time consuming it had been, how dreary and repetitive the work was, and that I had not yet gotten half way through the photos! This wasn’t supposed to take a decade. And I began to realize the opportunity cost I was paying. So in 2008 I decided to pay a company called ScanCafe to scan the remaining photos (2000 or so) for me. It cost a few hundred dollars but it was absolutely money well spent. I was pleased to discover the scans were just as nice as mine.

But I’d like to describe what I did for the majority of the photos, which produced excellent results. Here’s what I did. When dealing with a large collection of prints to digitize, it’s good to break the overall project into two iterative phases: one for scanning the source photography, and one for post-processing, to clean up, adjust, and optimize the scanned images. I generally kept the two phases in separate sessions. The two tasks are completely independent anyway. Think of the scanning phase as producing raw images, and the post-processing phase as producing final images.

Now on to the first phase: scanning the photography.