Notes

January’s font of the month: Klooster Thin

One of the joys of this club is having the opportunity—but not the obligation—to revisit and expand upon my previous work. It feels like a natural way to let some fonts grow into multiaxis families (Roslindale, for example) while others can simply remain one-offs (ahem, Pappardelle Party).

I always thought that Klooster was one of those one-offs: a font that I never expected to revisit, and certainly one that nobody has asked me to revisit. I sent out Klooster three years ago as a single heavy weight inspired by a bookplate I saw while browsing a friend’s book collection (shown below). I don’t think anybody has thought about it much since then, and hey, that’s fine with me...it’s not like there are a lot of contemporary use cases for a wacky bold uncial!

But this typeface has always been a personal favorite of mine because it was one of the first club fonts that really got me drawing outside of my comfort zone. It gave me the opportunity to research the uncial hand, a phase of the Latin script I knew relatively little about, and forced me to consider what the alphabet was like before the lowercase had fully evolved.

A typeface is a system of shapes, and a type family is a system of those systems. It can be taxing to track the consequences of every little change across a family, and I’ve written before about how this requires a different kind of thinking in interpolating variable fonts, which demand a certain level of compatibility across variations.

But with nobody very invested in Klooster’s original weight, I felt free not to worry about fealty to it or compatibility with it. Instead, I decided to let the new Thin style lead me in new directions. I know designers can champion “systems thinking” but this is a perfect example of where too much of it is a bad thing; I cannot tell how refreshing it was to focus on the problem in front of me and not sweat the rest!

While the original Klooster had a vertical axis, this time I ended up with a diagonal axis more closely connected to the calligraphic origins of the uncial hand. And while the original Klooster had tails and terminals on letters like R and E that gradually gained weight, this new style is dominated by extraordinarily long wedge-shaped vertical serifs that abruptly protrude from the letterforms. And because this is not a variable font, I was able to easily change minor details that I felt made more sense in this new weight (the triangular period, for instance, or the serif on the middle stroke of the M and W).

Every time I opened up the font to work on it, I kept on making those vertical serifs longer and longer until they became arguably the most important feature in the new design. I liked how the rigidity of the vertical wedges contrasted with the broad, round shapes that are characteristic of the uncial style, adding some snap and elasticity to the curves and giving the letters an untamed energy.

Having so many of these serifs opened up a host of spacing and kerning issues. There’s a bit of tension every time two vertical serifs appear next to one another...I think of it like the tinge of feedback you might get when two microphones get too close. After testing out ligatures and alternates to help them blend in more seamlessly, I decided to let the weirdness stand...some dissonance felt appropriate for this typeface, and I hope it doesn’t bother you too much!

December’s font of the month: Gimlet Sans Weights

Whether I like it or not, a typeface will always be defined by its relationship to typographic convention. Classics from 50 or 500 years ago cast a long shadow in how we think about and talk about type, and often the most effective new type designs are those that can offer something new within a familiar space. (I don’t think it’s a coincidence that my two best-selling fonts happen to be the closest things I have to Times and to Helvetica.)

In May’s mailing about Gimlet Sans Black, I wrote about how sans serifs can be exercises in distillation, as there’s a relatively smaller number of details available to set the typeface apart. With narrowing options in an ever-expanding field of sans serifs, it’s easy to feel pressure to add eccentricities that are, for lack of a better term, quirky for the sake of being quirky (QFTSOBQ?).

This month, as I moved Gimlet Sans to lighter, more conventional weights, my challenge was to sort out which of the typeface’s many quirks were QFTSOBQ, and which were manifestations of some central logic for the playful quirkhorse family.

I don’t remember consulting Georg Trump’s Schadow much while doing my initial pass on Gimlet Sans Black...it felt too far removed. But now that I’m tightening things up and paring the vocabulary of shapes down to the bare essentials, I figured it made sense to take a close look at Gimlet’s source material for the first time since 2016.

The squarishness of the curves, the underbite, and the super-closed terminals...I couldn’t bear to part with them. They all feel too central to what Gimlet is.

On the other hand, the thin/thick stroke contrast and the wider-than-usual proportions felt like elements that could be tamed in the lighter weights in order to make the family a bit more versatile.

Gimlet Sans sits at the midpoint between the superelliptical rigidity of Eurostile and the off-kilter playfulness of Ad Lib. And now that there are more weights in the mix, it’s funny to see how it moves across that spectrum.

Take for example the lowercase a: in the Black weight, its curves are mechanical and its tail short and stubby so it doesn’t interrupt the letterspacing too much. In the Light, however, the curves are more organic and the tail is allowed to stretch out.

I’ve also taken the opportunity to align certain horizontal terminals in the Light, even when there wasn’t room to do so in the Black. As you can see with the e and s in the image below, I created breakpoints in the designspace so that the terminals would either definitively align or definitively not align, never falling in between.

I’ll keep looking for more opportunities to tighten up this design, and if you mix the weights together, I’ll be curious to hear how you how you think the original Black weight clicks with the lighter end of the spectrum!

November’s font of the month: Club Lithographer

Like many other type designers, I have numerous typographic “sketches” that sit around in a folder on my computer. Sometimes they’ll contain a full alphabet, but usually they’re just a handful of characters...just enough to rough out a developing idea. Every couple months, I’ll crack open one of those files and noodle around for a bit, trying to see if I can get something in the design to click. And if it doesn’t, then usually I’ll forget about it for a while.

This month I’m sending you Club Lithographer, a sketch I’ve been playing with on-and-off since 2017. It’s a wide-set italic with elongated serifs, blobby outstrokes, and an unusually steep 24° angle. (My italics tend to be in the 10°–15° range.)

As I am wont to do, I started this sketch after seeing something I liked in a Victorian-era specimen book. This time, the spark was Lithographic Italic, credited to Andrew Little and published by A. D. Farmer & Son in 1873.

Ever since the angular chancery cursives of the Renaissance, Italics have tended towards the narrower end of the typographic spectrum. In contemporary use, this helps create a contrasting rhythm between Italic forms and their wider Roman counterparts when they are set together in a block of text.

I think this is why Lithographic Italic feels so refreshing to me: with no Roman counterpart to speak of, there is no need to squeeze. There’s no hint of choppiness or angularity, and you can feel the laconic freedom that each letter has to take up as much space as it pleases.

While Club Lithographer follows its Victorian predecessor closely in its overall proportions, it diverges greatly from the rigid and spartan drawing style of the original. In my rendition, I tried to take advantage of all the extra space to make something curvier and more free-flowing. I played up the expansion contrast present in this style of lettering, letting the weight quickly swell up in the downstrokes in a way that’s reminiscent of the expanding nib of a pointed pen when pressure is applied.

I admit I know very little about pointed pen calligraphy (and most of what I know comes from videos of Erik van Blokland’s pointed pen tool). But what I do know is that serifs and outstrokes in this style tend to be hairlines, where there’s very little pressure on the nib. I decided that the imagined calligrapher behind Club Lithographer should do the opposite, punctuating the beginnings and endings of strokes with expressive blobs.

You might have already noticed one conspicuous omission in this font: capital letters. If you want to think of this as a high-concept modernist reduction à la New Alphabet, I’m not going to stop you. If you want to think that I wasn’t happy with the capitals I drew, and was simply more interested in the lowercase’s asymmetry and syncopated rhythms, that might be slightly more accurate.

Either way, I hope you’re able to think of it as a fun constraint in your designs—a challenge to create an unexpected pairing of typefaces, or a challenge to sidestep the need for capitals altogether. 😉

October’s font of the month: Megabase Open

After the big Roslindale update I sent out in September, I figured I would go easy on myself this month and cross off something fun from my to-do list. Megabase has been one my favorite projects to work on this year, and ever since I sent it out in February, I’ve been dying to give it an inline companion.

Megabase Open involved relatively little drawing; I used some tools to help me carve out the space inside the thick strokes while matching the weight of the existing thin strokes. Then I manually cleaned up a few shapes (those tricky corners on A and M, for example), and voilà, it was pretty much good to go.

I could have stopped there; maybe I should have stopped there. But even the most straightforward project provides opportunities to trip and fall into a typographic rabbit hole, and fall I did.

It irked me that Megabase Open doesn’t distinguish between the counterforms of the letters and the interior “cores” of the horizontal strokes. They are all just negative space. I felt I needed something more complex than black and white could offer, and I looked to color fonts to help fill the void...both literally and figuratively!

My first instinct was to make a a color font with no color at all. Megabase’s “Fade” style shown above simply plays with the opacity of the interior shapes, letting whatever is behind the type show through the letterforms veiled in a semitransparent grey, simultaneously interacting with the background and sitting on top of it.

Then, a couple weeks ago, longtime club member Agyei Archer happened to mention gradients to me during an unrelated chat, and it struck me that they could be the key to taking Megabase Open to the next level.

There’s a big problem with gradients, though: color fonts come in a few different flavors, and not all of them support gradients. For example, I know it is possible to include gradients in a SVG image and then embed it into an OpenType-SVG color font, supported in some Adobe apps. But I don’t know of a way to specify a gradient in a COLR/CPAL fonts that builds up layered glyphs by stacking up a font’s native glyph data. (There is a proposal to add gradients to COLR/CPAL, but it has not been implemented.)

COLR/CPAL could really use a catchier name, but it’s the only flavor of color font supported by all modern browsers, and as we’ve discussed in the past, it’s the only flavor of color font that can also be variable. I got a bit carried away running experiments on how I could approximate a gradient in a COLR/CPAL font using only outlines and flat colors.

I settled on an approach that effectively rasterizes the gradient, turning it into a series of narrow stripes of subtly varying colors. With every tiny stripe now stored in the font as an alternate glyph, a font with 800 glyphs turned into 8000. And then, in the spirit of just trying things out, I added a variable slant axis (and I actually kinda liked it).

Was all of this a good idea? Probably not, but you never know until you try. Even though I probably won’t include these styles in the final Megabase family, I’m excited to share them.

Chris Lewis helped me build a demo where you can try out the color variable gradient font, and users with the fonts can use my Color Font Customizer to download their own versions. The randomize button (↺) can yield some pretty intense results! 💈

September’s font of the month: Roslindale Display Widths

I’m learning that this club is a great way for me to start new projects, but way-less-great for finishing them. The final stretch of a typeface’s development involves a decent amount of drudgery that doesn’t exactly lend itself to monthly installments. After all, type design doesn’t follow a traditional narrative arc; there’s no exciting build-up to the climax.

During the early stages of type design, I’m figuring out a system of shapes through exploration and trial and error. Once things begin to solidify, my process becomes more about managing the execution of that system through interpolation. This is essentially where I’m at with Roslindale, so I figured it was as good a time as any to send the club an update of the newly-expanded Roslindale Display (v2) and see what they think.

Interpolation is just math… it’s easy enough for a computer to interpolate between a Light outline and a Bold outline and spit out a Medium. Whether or not that blended result looks any good is a totally different story, and much of my time is spent futzing with outlines trying to get the different parts of the space to play nice with each other.

For me, a large interpolating family like Roslindale is a circus tent: a massive canvas propped up by tent posts in order to cover a certain amount of space. Math is the canvas, the connective material that unites the space, but it’s the strategic placement of the tent posts (not to mention their structural integrity) that keeps the tent aloft.

Up until recently, Roslindale Display was anchored by three tent posts: Condensed Bold from June 2017, Condensed Light from May 2018 (subsequently pushed further to Extra Light), and Ultra from last November. In anticipation of this month’s mailing, I’ve added three more: Extra Light, Bold, and Condensed Ultra.

Fewer tent posts means less work, less complexity, and (for a variable font) less weight. But it also means that the fabric can start to sag in the middle, where the result is rarely catastrophic (barring a technical error) but it can get soft and gummy and just a bit “meh.”

This is precisely what happened between my new Display Extra Light and the existing  Display Ultra, which I needed to create for the lightest + widest corner of the space. The thick parts of the letters got thick too fast, and the thin parts of the letters didn’t get thick enough fast enough.

My solution was twofold: first, I went back and subtly overhauled the way strokes taper in the Ultra weight. Next, I stuck another tent post in the middle (Display Bold), which gave me the additional support I needed to manage the progression of thicks and thins.

Interpolation doubtlessly involves compromise… it is, of course, a literal compromise between different outlines. But I try to draw shapes in a way that every instance within that interpolating designspace “pops” as if it is the only one.

Even if all of this interpolation talk was a bit dry, I still hope that you enjoy the circus that is Roslindale Display! (I’ll be curious to hear if you think the new Wide styles retain enough of their De Vinne flavor.) I plan to keep working on these right up until the family’s full release, so feel free to re-download the fonts here at any point if you want to make sure you have the latest.

Roslindale Display is September’s installment of Font of the Month Club; this month only, you can get all 24 styles sign up for as little as $24.