The first one’s the hardest

by Jeremy Mickel

I remember clearly the day I was waiting for the 6 train at 33rd Street and Park Avenue in New York. I had taken pictures of type on the street for some time, but there was something here that caught my eye. There was a plastic sign on a door with letters and numbers routed out of plastic, and I noticed a couple of characters in particular: the way the 8 curved back into itself, the charming tail of the a. And then I realized that the lowercase e’s were all different. This had been done by hand and therefore wasn’t an existing typeface. I knew then that I could actually make this into a font.

router inspiration

This moment of inspiration in a subway station was the beginning of the year-and-a-half process of designing my first typeface, and the start of an obsession that would permanently change the way I experienced the world. I thought I knew a lot about type as a graphic designer, but I soon realized that the art of typography went much further and deeper than I could have imagined.

Diving in

early draft

Having photographed the subway sign, I fired up Illustrator, and later FontLab, and started putting down points and pulling away at the beziers. One thing that really helped early on was opening up some of my favorite fonts in FontLab. I would look at how the points were arranged and the forms were constructed, and would flip and rotate the drawings to see what sort of optical corrections had been made. I found it intriguing to see that many glyphs that appeared symmetrical, like the O in Gotham, were subtly lopsided.

Though I was learning a lot on my own, I needed something more structured. I enrolled in a type design class at Cooper Union, but just days before it was to start it was cancelled. So, I contacted the instructor, Hannes Famira, and started taking private lessons. Hannes had studied at the Royal Academy in the Hague, where he was exposed to the Dutch school of type design. He taught me about Gerrit Noordzij’s Stroke of the Pen theory, the idea that there are two fundamental categories of lettering based on the broad-nibbed and pointed-nibbed pens and that serifs are just an expression of contrast. I did calligraphic exercises based on these ideas as well as drawing experiments with Frank Blokland’s Lettermodel modular type system.

This is how I first learned to evaluate drawing in type design. I went to Hannes’ once a week for the better part of a year, spending two and a half hours with him at a time. Every week I would show Hannes the progress on my typeface, which I was calling Router. He’d get out his red pen, and we’d go through and mark up the drawings. We’d talk about the things he was seeing, and the things I was trying to do. I was initially focused on some of the quirkier aspects of the subway sign, like the narrow f and t, and the strange r. I eventually abandoned those elements as I realized that Router was about something else. As other type designers have noted, self-editing is a crucial part of the process.

But then a funny thing happened. I kept correcting and correcting, and all of a sudden I had sanitized the font and there was almost no personality left in it. What I was left with might as well have been VAG Rounded. In a very early draft, I had played with the idea of exaggerating the swellings in the strokes from the original sign. Now I resurrected that, and found the true character of the font.

terminals swelling router


In addition to working with Hannes, I reached out to other type designers whose work I admired. One of the first people I contacted was Chester Jenkins at Village. I liked that his company was set up as a co-op, run by designers who all did great work.

I also posted my designs on Typophile, and was encouraged to submit my font to FontFont. I worked feverishly toward their May 15 submission deadline, preparing the book weight in roman and italic. I also showed Chester where my work was heading, and he suggested that there might be a home for Router at Village. The royalty percentage at Village was higher, but total sales would probably be less. I heard back from FontFont, and they were interested. I took a few days to think about it, but then let them know that I was going to proceed with Village and release my font in their new foundry, Incubator.

What made the difference was that Chester was willing to work with me and give me the time and attention to help make Router the best it could be, offering drawing feedback and sharing production tips. FontFont is a much bigger organization with a lot more releases, and they would have put me on a much faster timeline.


router complete character set

Once I had decided to release with Village, I started working toward a full draft of the family. Chester sent me files of Apex to show me the character sets they required. At first I was overwhelmed. I didn’t think I could possibly draw a font with about 1000 glyphs in each style. Fortunately, a lot of those are accented characters which can be created through components (linked references to other glyphs), but there are still several hundred original drawings in each of the styles. I had a lot of fun drawing some of the peripheral characters and finding ways for everything to follow the rules of the font, asking questions like “How can a triangle be routed?”

I spaced the font using the control characters ‘H O n o.’ Once I had the correct sidebearings for these characters I applied those values to all the glyphs that have straight or round sides (with some adjustments). Then I placed each character in the string ‘HHAHOAOO’ for uppercase and ‘nnanoaoo’ for lowercase, where ‘A’ or ‘a’ is replaced with the current glyph. Spacing the italic was a counterintuitive mystery — the only character that has the same sidebearings on both sides is the lowercase o, and you build everything off that.

In order to generate the in-between weights (extralight, light, and medium), I made interpolation tests. I did all of the drawing of the thin, book, and bold in Fontlab and then used Robofab (python-enabled Fontlab) to generate UFO masters (Unified Font Objects). Then I opened the UFOs in Prepolator to make sure all of my characters were interpolatable and used Superpolator to do the actual interpolation. Finally, I re-imported the UFOs back into Fontlab and checked each glyph for errors. It’s a lot of steps, but it’s superior to using either the Multiple Master or ‘blend’ action within Fontlab because it’s much more flexible and precise, capable of generating instances that require minimal redrawing.

During this time I kept showing my work to Hannes, as well as having meetings with Chester. I also got great feedback and guidance from Village member Christian Schwartz.

Time to deliver

router specimen on vllg.com

I could have kept fiddling with it forever, but a date had to be set. I was moving to Providence, R.I. in the summer and didn’t want to take an unfinished font with me.

The last minute drawing and spacing changes were very important. The final steps were to build the accented characters, and then use Metrics Machine to kern the six masters (thin, book, and bold in roman and italic). I finally sent the files to Chester. There was some back and forth over the next week — he caught some mistakes and I rethought a couple of things. Then he mastered it and launched the Incubator foundry on Bastille Day, July 14, 2008, coinciding with Village’s three year anniversary.

A year and a half after I first started, and a world away from my first draft, I had finally published a typeface.

What’s next

I’ve worked on a few custom type commissions since Router’s release as well as projects with my friends at The Design Office, a collaborative office of independent designers here in Providence. Most importantly, I am working on more uncommissioned designs that will eventually become commercial releases. But it hasn’t been as easy as I thought it would be. Just because I have an idea doesn’t mean I can pump it out right away. As Chester told me, “you still have to go through the same snakes and ladders.” The production side of type design is much easier though, because I’m not figuring it out for the first time.

I’m trying to push myself and expand the ways I think about letterforms. In addition to drawing type, I’m taking a calligraphy workshop and later this year a stonecutting class. I’ve been spending a lot of time at the Providence Public Library where they have D.B. Updike’s collection of type specimens and books on printing. And I’m still taking lots of photos of signs and things I see on a day-to-day basis, and reading all the books I can find.

But no amount of research or theory is a substitute for the most important thing: drawing as much as I can.

Bonus: links & thoughts

Several designers have told me how important it is to have a specific use and point size in mind. The idea is that if you try and design a font that’s good for everything, it might not be REALLY good at anything. But if the font works really well for one specific use, then it can probably work well for lots of others. I’ve heard the example of J.K. Rowling writing her books for her daughter. If she tried to write books that everyone would like, they might be too general to connect with anyone.

When you’re drawing a character and it’s taking up the full size of your screen, it’s easy to forget how it will look when you print it at 12 pt. In order for your decisions to have a real impact, the drawing has to be a caricature. It’s okay for details to disappear in text. But only by printing specimens at different sizes can you see the real effect of your actions. And only by looking at the individual letters in words, sentences, and paragraphs can you understand how all of the glyphs work together.

Draw the black and the white shapes. Many designers will say that the white shapes are more important to the overall harmony of the letterform. Just like turning the glyph upside down or looking at it sideways, concentrating on the white shape lets you understand the form from a different perspective. And balancing the white with the black helps you with the spacing and understanding the overall weight of the forms.

Optical corrections are key. It’s been said that type design is the art of making unequal things appear equal. Noordzij’s theory of the Stroke of the Pen is apparent even in monoweight sans-serifs. Flip Helvetica’s A, V, or W sideways, and you’ll see that the diagonal strokes are slightly unequal. Rotate the O in Futura, which I was always told was a perfect circle, and you’ll see why that’s not true.

futura o


The most helpful sites for basic type info:
Underware’s typeworkshop
Briem’s Notes on Type Design (disregard the site design — the information is great)
Hannes Famira’s www.typedu.org

And the most informative (and often out-of-print) type design books:
A Book of Type and Design, Oldrich Hlavsa
Counterpunch, Fred Smeijers
Dutch Type, Jan Middendorp
Fontographer: Type by Design, Stephen Moyer
Adrian Frutiger — Typefaces
The Stroke: Theory of Writing, Gerritt Noordzij

Many thanks to Chester and Tracy Jenkins, Hannes Famira, Christian Schwartz, and everyone else who helped me along the way.

previous: Malabar type family released

next: Yes, we kern

Typeface Categories

learn greek type desgin
learn greek type desgin
font deals of the week
fonts on tape