Perhaps the greatest challenge of creating revival typefaces is finding source material. Once upon a time, it was a daunting task but, lately, it has become a lot easier. I consider myself fortunate because I live in the Washington, DC area, and have access as a Registered Researcher to the vast holdings of the Library of Congress, the world’s largest library. I can do searches in the Library’s online catalog, compile a list of promising suspects, then take the train into the city on a Saturday morning and spend time poring over potential revival prospects.
Of course, there’s another library of a kind available, potentially even larger than the LOC, and that’s the internet. There are an astonishing assortment of websites with graphic resources of all kinds posters, luggage labels, travel brochures, sheet music, signs — and it’s always a pleasant surprise to discover new ones.
Sometimes the source material is an outright gift — complete alphabets from old handlettering books or type specimen catalogs, which makes the task of creating a font simply a matter of time spent creating vectors. Sometimes the source material only offers hints — a few letters here and there, which require a lot more prep time in analysis and interpretation. In this latter group, perhaps the two most demanding projects I undertook were ITC Scram Gravy and Picayune Intelligence BT.
Previously, I discussed the development of ITC Scram Gravy with Allen Haley,
in an interview in the November/December 2002 issue of STEP Inside Design magazine, so let’s take a look at Picayune Intelligence.

I came across a 1920 poster for Riquet Candies by the estimable Ludwig Hohlwein in Steven Heller and Louis Fili’s "German Modern: Graphic Design from Wilhelm to Weimar," and liked what I saw .

Introduction to Riquet

I then did some analysis of what made the typeface "tick" and came up with a simple set of rules:

1. The uppercase letters were basically mapped to a square originating from the baseline — that is, they were fairly wide.
2. The typeface had a large x-height — the lowercase letters were almost as tall as the uppercase.
3. The descenders and ascenders were equal in height.

4. Some lowercase letters had serifs, and some did not. I recognized this rule, but chose to ignore it: the typeface I would create would apply serifs consistently.
5. Finally, the curly tail on the uppercase R dictated that this typeface would be playful.
With this set of rules in mind, I went about the task of creating a complete font upper and lowercase letters, numbers, punctuation and symbols. As it turned out, the only rule I had to alter slightly was Rule #3 — the very short descender made the lowercase g and y a little too cramped, so I decided to increase that length slightly to give me a little more breathing room on these two characters.

I believe that the finished product retains all of the charm of the original handlettering, while adding a few tasteful and exuberant variations to it.