The Development of Benson Script

Benson Script started as a rejected piece of lettering from a logo job in early 2011. Now that I look at it, it’s pretty obvious why it was rejected. The logotype, for a monument company in Santa Barbara, was an exercise in what chiseled script might look like. Knowing nothing about the art of chiseling, I could only imagine that, as much as possible, all letterforms would adhere to a ridged chisel.

I lettered many logos in this style long after that job was done. It became very attractive to imagine a script that was desperate to be anything but a script. As if it was straining against the author’s insistence that scripts were more curvaceous. As the author, I often questioned my decision to adhere to those modern, techno like, horizontals. No doubt I will one day release a font that is a revisiting of these ideas, with kinder curves.

Many of the early ideas that derived from that sketch lacked unity and the modernist, anti-calligraphic, look I was going for.

It was difficult to decide which conventions should stay, and which should go. This lead to a lot of experimentation—leading Benson Script to become a type family with a lot of variety. At first I hoped to whittle down all of my ideas to a single, worthwhile font. Eventually I realized that its strength was in its ability to provide variety within tabular space, and released it as such.

Specific Solutions with Opentype

As I discuss this stuff, keep in mind that your “Character” and “OpenType” windows in Illustrator should be open. When I tried to explain some of this stuff to a veteran Illustrator user, they were unaware of these totally crucial windows. So forgive me if some of my explaining seems a little dumbed down, while simultaneously being TMI.

When using Benson Script, as is the case with just about every font I (and most type designers) design, keep the ligature button checked. Clicking that little button keeps things nice and b-e-a-utiful. Especially in the case of the lowercase /r/.

This little fix came thanks to my pal James T, of Wisdom Script fame. The trouble with the /r/, in a script, is that you want it to look good for those who don’t turn on that ligature button we talked about. So the default you design is one that connects to most letters. BUT, if you should start a sentence with a lowercase /r/, it looks awfully gross. So the solution below essentially dictates that you replace it with the more conventional /r/, no matter what, then swap it back to the regular /r/ only if it is appropriate.

Creating alternates for the end letters is a little more common, but a solution I hadn’t thought much about until I was spending some time in the land of connections and ligatures. Especially with the odd, awkward on the end, connections on numbers 15, 25, and 35 (see the /w/ and /v/ below). This solution says that whenever one of these awkward ends is put up against a space, or a character where a connection doesn’t make sense, it swaps for an end character.

Last, but certainly not all, is the fractions. I’m mentioning this because it’s an opentype feature that a lot of people don’t talk much about. Just click that little button, and normal numbers with slashes between them become beautiful little fractions.

That’s all folks!

Pick up Benson Script over here, and feel free to email me with any questions or comments you might have while using it.

The Development of Tide Sans

Tide Sans Blog Post-01

Tide’s In! Started as a Painted Sign Revival. It was the First Sans I Designed, perhaps a bit too obviously so.

The Tide laundry detergent ad below served as the primary inspiration and namesake. But it provided very little by way of the lowercase set, or the crucial letters /HGMBV/ for reference.

Bezier Curves

Being extremely new to the sans genre, and new to type design in general, a lot was misunderstood about the appropriate shape of letters, and how to they should interact in the first few versions.

Most recently, I released V3 of Tide’s In at a pretty transitory point in my life. I had just started interning with Thomas Jockin, and had learned enough to know that if I showed Thomas my final version of Tide’s In he would have told me not to release it yet. I had spent the last two months working on it, and I didn’t want to be shut down right in the moment of my triumph. So, I released it and no more than two weeks after releasing it I had learned enough to realize that it was awful, and not well tested. So I pulled it from all of my distributing sites before the ugly scar I put in the design world was beyond repair.

Now, this was only four months ago. I’m under no misunderstanding about how far I’ve come since then. I realize that a year or two from now I’ll be criticizing the release of Tide Sans for having just as many issues. But this post isn’t about that. Many of you have emailed me, asking to buy Tide’s In or asking why it was no longer for sale. Many of you still use it, and will continue to use it after I send the update. This post is about a brief explanation of some of the important things I’ve learned since then. The tiny things that have made Tide Sans a HUMONGOUS, 1000 hour, project that at first glance might not seem justified.

Some obvious things that were changed include the addition of a few fractions, added weights, adjusted spacing/kerning, small caps, ligatures, and ordinal characters, but that’s not the stuff I’d like to talk about. I’d like to talk mostly about using bezier curves for drawing. I know you all think you understand how those work, cuz Illustrator and stuff, but you’ve got it wrong. There’s a bigger, tastier, picture here to be seen.

The colorful examples shown of each letter below are provided by the bezier examining tool SpeedPunk. I had this tool long before I designed Tide’s In, so it is not the reason for the improvement, but it does help illustrate the balance and momentum of each curve. I exclusively used the 700 weight for my demonstration, as it is the only weight that has the similar stem widths between old and new.

Tide Sans Blog Post-02 Tide Sans Blog Post-03 Tide Sans Blog Post-04 Tide Sans Blog Post-05 Tide Sans Blog Post-06 Tide Sans Blog Post-07

Good Taste and What it Means

When I released Tide’s In I didn’t have a sense for good taste, and simply traced it because it was old. I still struggle with what taste is and how to bottle it up and sell it, but I the breakthrough I had with Tide Sans came when I set up this site to use Tide Sans Condensed as the body text. I cringed, not because it looked bad, but because they weren’t my style. I set the css to access the stylistic alts and it all looked much better to me.

This is the biggest stylistic jump for Tide Sans, and it was made only a few days before the release. Why would I release a design that I liked hidden behind a design that could potentially fit a niche? I guess I figured there weren’t a lot of perpendicularly cut off leg/arm terminal sans out there.

I still do use that style, but certainly not as much as I was going to use the stylistic alts. Taste is deciding to call something good more regularly than you would call its neighbor. I used to think that taste was something more absolute, that you could call better all of the time.

The Development of Millie

Millie Blog Post-05

On a plane ride from New Hampshire, Millie was born.

When I first drew up the Millie alphabet, all I wanted to do was create a cool font, and as a result have people think I was cool. It was the logo for Milwaukee Tools that got me started. However the logo doesn’t begin to have enough data for a complete font. For a budding type designer, this lead to some problems that I was not ready to solve – or comprehend.

Last month I decided to re-approach Millie in the same way I have approached my more recent type designs. This necessitated that I be very deliberate and methodical. This post is a summary of those deliberations.

No Objective

Before the creation of Millie I had only ever used design to solve problems. I had never asked myself what kind of project Millie should be equipped for. Which meant that I wasn’t sure of the full expanse of the character set. This kind of “preemptive strike” design was an entirely new idea to me.

Millie’s objective was to fill the void where a tough, straightforward, and durable script should be. Stylistically speaking, I hadn’t learned enough to pursue the goal of making Millie a script. So I ended up with an amalgam of all the off thoughts I’d had on typography—made geometric. I was somewhat aware of its lack of script-ness, but found that a lot of the script conventions contradicted the straightforward, tough, and durable goals of Millie. I just didn’t have the skills to negotiate a compromise.

The script genre is quite expansive, and spans from simple hand painted styles, to more formal styles with swashes that make letters nearly unreadable. Here are four examples from MVB, James Edmondson, Laura Worthington, and Parachute that demonstrate significant differences in scripts I often use.

Popular Script

With those conventions in mind I set some firm guidelines this time around. There are three big rules that helped me draw Millie into her proper spot in the script universe:

  1. Text must be clear and easy to read in headlines and titles.
  2. Each letter must connect, unless connecting impedes on rule #1.
  3. Each weight must hold its space as if it were as heavy as the heaviest weight.

Drawing a script letter is not so complicated. The madness lies in ensuring each letter will properly connect to the letters on either side. Sometimes that even means setting a criteria for when two letters should not connect.

When I designed Millie Light, I had more space for experimenting with connections. This lead me to draw many more variations, and design an almost completely different alphabet. Even still with this weight there was confusion on when to connect. There was no criteria for why two letters should connect.

Millie Vs. ExtraLight

In this latest release Millie will essentially always connect. Some connections were removed if those connections impeded clarity. In hindsight, the solution seems simple, but this solution required work that I just wasn’t able to stomach at that time.

The most obvious example of these connection difficulties is the letter o. The o script requires that negative space become aesthetically scrambled. This was the problem that I avoided because the optical solution was beyond my comprehension. It’s also complicated because it’s one of the few ligatures that doesn’t connect to the next letter via the base. So I crafted a simple opentype solution that essentially says: if it makes sense to connect, do it. Otherwise, keep your counter untouched and beautiful. Here’s what it looks like:

o connections

From Logotype to Just Plain Type

What works in a logo, doesn’t always work in type. Some big things that didn’t transfer well included:

  1. The slant.

    I increased the slant of the type by about 4 degrees. The main objective in doing this was to maintain motion, and the sensation of straightforwardness and strength when the type was put on a slant.

  2. The case contrast.

    I decreased the size of the uppercase set by 200 units (out of 1000). You can see in the comparison below that I kept that bigger size in v1.0. I ditched the 200 units to keep with the objective of appearing durable. For the same reason, I increased the x-height by about 50 units.

  3. Certain angles.

    For example, I tried to keep the angles found in the letter /k/, but eventually decided that it cluttered the continuity.

Caps Comparison


A terminal is often decided based on the way that a calligraphic pen would release the stroke. You’ll notice in the example below, that Millie’s terminals are not decided that way. That decision was made when I felt like the angled terminals of the pen detracted from the modern, strong, look I wanted.


Attention to Detail

The original drawings took place in Illustrator, which is a poor environment for attention to detail. This time around, everything was built out in Glyphs using a guide to ensure that every stem and turn was at the right incline and weighted correctly.

Lowercase Comparison

I would love to illustrate the angle differences, but they are too minor to effectively show. Before I had a purpose for Millie, slight differences didn’t mean anything. Nobody is measuring, right? With my new objective of being durable (and by extension: precise), small details like that were being counted on by users.

They weren’t measuring, that’s my job.

Millie V5.0 is Alive

This deliberate design, tailored to new demands makes Millie a piece of software. It is prepared for your problems. Millie has been through a troubleshooting process that makes it road ready. For this reason, V5.0 is the last update Millie will go through. From here on out all of my type families will be on par with that scale of consideration—I’ll prove it by accompanying each one with a post similar to this.

Enjoy, and as always, thanks for your support!