When I started out my mapping career, I had a dream of eventually making my own atlas. And, in the following years, I was fortunate enough to get to work on a few, including the Atlas of Design, and the Ecological Atlas of the Bering, Chukchi, and Beaufort Seas. Of these various atlases, my latest one is certainly the most “my own” of any of them. I’m pleased to share with you An Atlas of Great Lakes Islands.
While my earlier projects involved me producing some digital files, then handing them off to be printed, bound, and shipped, this one was a little different. Though I did design the maps digitally, I printed each one by hand using using the cyanotype process. I also manually trimmed each one down to size, and sewed my own binding. It may not be the most visually impressive, but it has much more of me in it than probably any other book I have worked on.
There’s a lot to talk about here, so let’s go through the process in rough chronological order.
A Starting Point
Some months ago, I decided one evening to see how much information I could cram into a black and white map, with no shades of grey in between.
It was just a fun challenge, but I really liked the look that came out of it, and I ended up doing a couple more of Great Lakes islands. If you browse through my portfolio, you’ll already know that I make a lot of my maps about my Great Lakes homeland. I feel like it’s part of my cartographic mission to articulate, in my own way, the beauty of this part of the world. I’d already made one poster of Great Lakes islands before (in a black-and-white style, no less), so this was a sort of unintended continuation of my previous work. I think of them partly as an exploration of geographic forms: the interesting aesthetics of the shapes nature has laid down.
I also was definitely inspired by the work of Heather Smith, including her map of Iturup, though it took a little while before I realized it. I often absorb ideas from others and then, months later, they pop up again without me realizing that I’m drawing upon them.
Around this same time, I was getting into cyanotype printing, a nineteenth-century alternative photography technique. You can follow that link if you want to dive a bit into the chemistry, by the idea is straightforward: (1) coat some paper with chemicals; (2) put a photo negative or some other light-blocking stuff on the paper, and then expose the paper to sunlight or another UV source; (3) dip the paper in water; and (4) watch the paper turn blue in areas that were hit with sunlight, and stay white in the areas that were shaded.
Cyanotyping is cheap and easy to get into, so I started printing up some maps to try it out. I made negatives on transparency film, and copied a few maps I’d already made.
But I was looking for an original idea, and that’s when I thought to wind together two threads: black-and-white maps, and cyanotyping. I decided that I would produce a series of cyanotype maps of Great Lakes islands.
First off, I needed to digitally design the maps that I would eventually be printing. The process here is pretty much my standard GIS-to-Adobe Illustrator workflow. A few highlights:
- Gathering data was a challenge. I needed equivalent information for places in the US and Canada, and oftentimes had to mash stuff together from a lot of different sources. Here’s an example of one map’s sources:
- My data sources were often in conflict, too. Streams would be shown running in different areas, for example. Given the size of these islands, that’s not surprising: many of these streams are probably small trickles that aren’t well-documented. Sometimes roads or trails would have spotty data, and I’d have to correct or supplement things by looking at satellite imagery.
- There was also some disagreement over coastlines due to things like tides and whether wetlands counted as water or land. There’s definitely room to quibble with some of my choices.
- I tried to fit as many different data layers as I could, given the limitation of using only black and white (which would eventually print up as blue and white in the final book). To keep everything looking clean, there are a lot of knockouts. Roads knock out streams, streams knock out land cover, text knocks out everything, etc.
This helps a great deal with legibility, I think. Without those buffers around every feature type, the map would be a mess:
- To keep those buffers from hiding critical data, I often rerouted streams and roads to make sure that a coastal road didn’t hide the outlet of a stream.
- I’m quite pleased with the forest texture. Instead of a repeating dot pattern, I gave it a little bit of randomness. The dots vary slightly in size and position. They’re almost regular, but not quite, which I think adds some organic-ness to the whole thing that ties the symbology into the idea of vegetation. (Edit from years later: I made a tutorial.)
- Because I pay way too much attention to tiny projection details: each island is on its own Lambert Azimuthal Equal Area projection, with a center point at the centroid of the island.
- I set all the type in Mostra Nuova, which is a typeface that will perhaps be familiar to longtime readers. It’s a favorite of mine, and based on Italian Art Deco posters. It is, admittedly, not really connected very well to the atlas subject matter. But I still liked how it looked with the stark monochrome cartographic design. Mostra Nuova is a geometric sans, and the cartography is all about highlighting interesting geometries, so they sort of work, I think. One limitation: it doesn’t come in italics, and most its various weights would be hard to tell apart in cyanotype. So, I didn’t develop different label styles for different feature types, for the most part (except settlements and airports, which are given in heavy type).
Add up everything above and these maps took a lot longer than you might expect just by looking at them. In a process that took dozens of hours, I made maps for the largest thirty islands in the Great Lakes system (including some of the rivers connecting the Great Lakes). I had planned to stop a little sooner, at maybe twenty maps. However, I wanted to include at least one island from every Great Lake, and Pelee Island, the largest in Lake Erie, was the thirtieth-largest in the system.
Once I had my maps designed, I printed off negatives on transparency film and set about cyanotyping.
First off, the paper is coated with sensitizer (ferric ammonium citrate and potassium ferricyanide). When I started learning this process, I used watercolor paper. Later on, I switched to kōzo, a Japanese mulberry paper. It’s handmade, thin yet strong, has no particular grain, and has a really nice natural, organic feeling to it. I got mine from the Awagami Factory.
Then the paper is dried in the dark for a while. When it’s ready (about a day, for the kōzo), I can put a negative on it, sandwich it between some glass, and put it out into the sun. In Wisconsin, sunny days are rare in the winter, and so I usually had to drop everything when one came around so I could get some prints done. While later on I got good enough that I could do exposures on cloudy and overcast days, it was nice to have an excuse to go out into the sun and get some Vitamin D.
Once the print is exposed (anywhere from 5 minutes to 2 hours, depending on conditions), a little water will cause the image to develop.
The prints vary quite a lot, as there are a myriad of tiny things that affect the outcome: the angle of the sun, how tightly the negative sits on the paper, the pH of the paper and the water used to wash it, how evenly and thickly I coated the paper, how long it was washed, how long and under what conditions the sensitizer dried before exposing to the sun, and more. Below you can see two very different-looking outcomes from the same negative. One is very dark and overexposed, and the other very light and underexposed (the more exposure to the sun, the more blue pigment develops).
Controlling all these variables is almost impossible, and each image is a little different. I picked up a lot of small tricks along the way to try and solve imperfections, but many remain: spots on the print, odd wrinkles or smudges, blurring, etc. I’m an extremely detail-oriented perfectionist who zooms in to 64,000% in Illustrator, so this project has been good for me to practice letting go. This whole process is very, very analog, and only some of it can be rigorously controlled. The imperfections also add to the charm, and express how the print was handmade.
The first draft of the atlas was printed on heavy watercolor paper. To get 34 decent prints (thirty maps plus a few additional pieces of front & back matter), I had to go through over 100 sheets of paper, as I still had a lot to learn about how to get good prints out of the cyanotype process (and how to accept imperfections on the non-ideal prints that came out of the process). You can see in the image here how many different shades of blue there are; just another reflection of the variability of the process.
After I finished up the first draft (and bound it up into a book), I made a second draft of the maps, printed on kōzo. This time I was much more efficient, both at making good prints and at letting go of my perfectionism. I got 99 good prints out of 110 attempts, enough for three atlases of 33 pages (one less than the first draft; I dropped a filler page).
These books would have just been a bunch of loose pages without the guidance of my colleague Caroline Rose. She has experience in book repair and bookbinding, and was kind enough to teach me some basics and loan me a bunch of supplies. With her pointing me in the right direction, I bound up my maps into actual atlases.
The first draft of the atlas, printed on watercolor paper, was given a simple accordion binding. I simply attached the pages end-to-end with strips of kōzo glued on the back, acting as hinges. I dyed the kōzo blue first, using the cyanotype process, just to make it look a little nicer.
The first atlas was the culmination of dozens of hours of work (and dozens of dollars of supplies, mostly watercolor paper), but I found it only mildly satisfactory. Among other issues, the book had some warping problems. Despite trying many, many things, I could never get the pages to dry flat once they’d been soaked in water. But this was all fine, because I had planned it, from the beginning, as a first draft, with the understanding that I’d do better the second time around.
The book’s kōzo hinges were my first time working with (or hearing of) the material, which Caroline introduced me to, and I knew I wanted to work with it more. So, I decided that my second draft would be printed on kōzo, rather than just using it as a supporting material. After wrapping up the first book, I printed up the 99 maps on kōzo that I showed above, and then prepared them for binding. First, I hand-wrinkled each one. You see, during my test prints on kōzo, I accidentally made a print that was somewhat wrinkled (due to not handling it super-carefully: I wanted to try to see how strong it was). But I found that I liked the wrinkled look, so I decided to make sure that every print was sufficiently wrinkled. It’s sort of wavy, and organic, and aged. It gives the prints some suppleness.
After I spent a while wrinkling each print, I tore them all down to size. I thought about cutting them to give them a clean edge, but I liked the softness of the torn edge.
The whole stack of pages reminds me a bit of worn blue jeans, which makes sense. Blue jeans are often dyed with Prussian blue, which is the same pigment that the cyanotype process produces. And blue jeans are made from cotton, which is mostly just cellulose. Kōzo, like other paper fibers, is also mostly cellulose.
Once I had my soft, wrinkly organic sheets, I wanted to give them a better binding than I had the first atlas. Something more book-like, instead of a giant accordion fold. Caroline pointed me in the direction of Japanese stab bindings, which are simple to execute and designed for single sheets. So, I poked around some tutorials, bought some supplies, and got to work. I used an awl to poke holes in the pages, then sewed them up.
The thread is marbled: I used the cyanotype process to dye parts of it, but not others.
Unlike the first draft, I also gave the book a cover material. I used grey and black Japanese linen cardstock. I also gave it a string and button closure, using paper buttons I cut from the cover stock and leftover binding thread.
Kōzo is handmade and doesn’t have much of a grain, so there’s no particular warp direction this time around. It’s also much thinner than watercolor paper (while still having a lot of wet strength). These are two reasons I had wanted to switch to using it instead of watercolor paper. However, the book is very puffy, probably due at least in part to all those wrinkles I added.
I hadn’t planned on drawing so much from Japan when I started this project, but it just sort of happened that way. I just happened to learn about and come to enjoy working with kōzo; I found a good cover material that happened to have been made in Japan; and I needed to learn a simple style of binding that beginners like me could handle, which happens to have been Japanese. So, it’s a book about the Midwest, with an Italian typeface, made with Japanese materials and techniques.
Significance and Sharing
I produced one copy of the first draft (on watercolor paper), and three of the second draft (on kōzo), though I’ve still got to get around to binding the final copy of the second draft. It took a huge amount of time (and a fair bit of money), but I’m pretty satisfied with the results. As I said, there’s a lot of me in this work: it was hand-printed, hand-wrinkled, hand-torn, and hand-bound (using hand-dyed thread). There’s not much more possible influence I could have on a single book. And it’s of a subject that has deep personal significance to me: my Great Lakes homeland (even if I’ve never been to these particular islands — yet).
Despite its unassuming appearance, I count this book as one of the more significant accomplishments of my career, because it sort of serves as a microcosm of so much of the rest of my practice: it’s in monochrome, it’s about the Great Lakes, it hinges on a lot of obsessive detail work, it’s something nobody asked for, it required learning a bunch of new skills, and it’s a little off the beaten path. It may be the most Daniel P. Huffman-like of all of my works.
As I said, there are only three copies. But, maybe you’d like to see one? It so happens that I’m sending one copy on a world tour, where it will be shipped to whoever wants to borrow it for a few days. If you’re interested, you can read more here. Note that the tour is pretty full, so you could be in for a long wait.
A lot of my purchase of cyanotype materials was made possible by the generous support I receive from the map-loving community. If you want to support my future work, you’re welcome to use the links below.