I’m in the process of picking a typeface for the book. I have a very strong contender that seems to have the right characteristics for letterpress printing and for the paper I believe I’ll be using. But one of the deciding factors is cost.
If you haven't had to buy fonts, you may not realize how expensive they are. We’re in a golden age of high-quality “free” fonts, because Google, Microsoft, Apple, and other commissioned and paid the $100,000s or many millions to develop faces that they allow general use of. Google is over five years into the Noto project. (Which I wrote about and then had the piece killed when Trump was elected—the publication didn't have room for it, then it perished as it grew old before it ran. It may be in the book: it’s pretty apropos.) Noto comprises over 100,000 glyphs (characters of all sorts across all script types around the world) and will ultimately be even larger. It will eventually also come in both serif and sans serif versions.
But those freely available faces, some of them quite excellent, are a drop in the bucket of the type universe. Most of the time, when you’re planning a project, you rack your memory of faces you like and look through samples online or in print to find candidates that convey the feel and fit the medium. If you have need of particular typographic features, like small caps or swashes, that focuses direction, too.
(FYI, while there is some variations in how type historian, type designers, and graphic designers talk about it, a typeface typically refers to a family of letters designed to work together across multiple weights, and a font is the instantiation of a typeface. So Garamond is a typeface that encompasses Garamond in fonts of Roman, italic, bold, and so forth. Thomas Phinney ran a poll to find the most popular definitions of these and related terms, and it was fascinating.)
In my case, I want a serif face that hearkens back but doesn’t feel like a copy of a face from the past. I want something highly legible and without terribly thin strokes, which don't always hold up well in the plate-making process nor in printing nor stand out from the paper. A more even stroke will make it easier to produce pages that have an even and equal impression.
However, if I were to purchase such a family outright, it could easily cost $750 to $1,000 to purchase the license that covers the right number of variations (Roman, italic, and bold in text weights and then some versions tweaked for larger sizes) and have the right to use both in print and include in a single ebook. I don't think this is outrageous: font design and development isn’t cheap, and it’s very speculative. No type designer ever really knows how much they’ll sell of a face sold commercially, as opposed to one designed for and paid by a client.
Because of this project’s modest budget (about $12,000), I can’t purchase a set of fonts, nor is a particular typeface critical to the project. I turned to Adobe TypeKit, a service I’ve been using for years, since it was an independent company, and one of the first good ways to deliver Web fonts reliably to browsers, so you could actually design with a typeface in mind and it worked everywhere. After Adobe’s acquisition, TypeKit expanded choices and its purpose and license. While terms vary somewhat, the subscription (which I have as part of Adobe Creative Suite) lets me pick among many typefaces, sync across my computers, and then use in print projects, ebooks (that lock usage to the book), and often on Web sites. You don't need an active subscription to keep using the fonts in ebooks, just on associated Web sites.
I won’t reveal the face yet, but I’m very pleased all around about it so far, and will be making more experiments. In the nearish future, I’ll prep some variants in size and leading (vertical spaces between lines) and get a photopolymer plate made so I can do test printing onto small sheets of the paper I expect I’ll choose. This will let me test for type color (the relative density) and legibility, and make sure I'm making the right choice.