Heldane design information

Heldane is a contemporary serif family inspired by the renaissance works of Hendrik van den Keere, Claude Garamond, Robert Granjon and Simon de Colines. Rather than emulating a specific font, Heldane amalgamates the best details from these sources into a cohesive whole. The classical typographic foundations of Heldane are refined with rigorous digital drawing.

2,522 words by Kris Sowersby

Over a decade in development, Heldane comes in two families: Heldane Display and Heldane Text. The fine detail and tight spacing of Heldane Display is best expressed at large sizes. Heldane Text thrives at smaller sizes, taking stylistic cues from the display cuts, but with an additional focus on optical functionality.

I consider Heldane a third generation garalde typeface drawn from secondary sources. The first generation are 16th century works from the likes of van den Keere, Garamont, Granjon and De Colines.¹  The second generation are 20th century conscious metal revivals. By conscious, I mean the concerted effort to revive a specific style, whether the source was accurate — like Stempel Garamond; or not — like American Type Foundry Garamond No.3.²

The third generation are those made since 1955, after the re-discovery of the Plantin-Moretus archives and subsequent scholarship. These are types like Sabon, Galliard, Adobe Garamond and Renard. The designers of these faces skilfully exploit modern scholarship, disambiguation of punchcutters, and trace accurate lines to their primary sources.

I used secondary sources because I didn’t have access to primary specimens or to original renaissance books when I started drawing Heldane. I used facsimile prints, scans and images found online. During the 10 year development process I acquired primary specimens of van den Keere and Garamont, but they didn’t change the fundamentals of Heldane.

Modern scholarship distinguishes between Garamont the man and Garamond the typefaces.

In 1926 Beatrice Warde (pseudonym Paul Beaujon) published an article in the Fleuron clarifying that the “Garamond” types in the French National Printing Office were actually Jean Jannon’s. Some type foundries inadvertently had issued “authentic” Garamond revivals based on Jannon’s.

Hendrik van den Keere’s 2-line Double Pica Roman. Scan from Type specimen facsimiles II.

Heldane Display’s Roman lowercase draws from van den Keere’s 2-line Double Pica Roman, reproduced in Type specimen facsimiles II. The lowercase is arresting: dense, sharp and powerful. It’s elegant and classy without feeling fragile or haughty. In the context of the specimen book, it’s quite unique. I felt compelled to start drawing it immediately.

Left: 2-line Double Pica Roman. Right: Heldane Display Regular.

The a is damn near perfect. It’s the first letterform I really noticed in Facsimiles. A decade later, I still struggle to articulate why I find this letterform so compelling. The angular curves, sharp tail and delicate head serif make a dramatic combination. But there’s something else. Something ancient and elusive about its form I wanted to capture within the rigid confines of digital vector drawing.

2-line Double Pica Roman.

Heldane Display Regular.

The modulation of widths — like the narrow n and wide o — is quite different to Garamont’s regularised proportions. Coupled with tight spacing it makes a graphic impression in a text block. Other details like the subtle clockwise lilt of s, the broad t and the structure of g set the tone and direction for the rest of Heldane.

Firstly, in looking with Mike Parker at the early trial drawings of Galliard Roman, he would comment that some letter must be wrong because Granjon would never have cut it that way. I could immediately point to a case where Granjon had cut it exactly so; but Mike was essentially right because the perception of style is subjective; it must be assimilated and re-created as a whole, and not defined by its eccentricities.

— Matthew Carter, “Visible Language”, vol. 19, no 1.

The Facsimile didn’t have many examples of van den Keere’s roman capitals.³ Initially frustrated by my lack of resources, I eventually resolved to make the best of what was at hand. It was more important to make a cohesive family than a slavish digital reproduction of his work. Conceptually, this allowed me freedom to look at van den Keere’s contemporaries for inspiration.

When I finally acquired an original specimen of van den Keere’s 2-line roman capitals I was disappointed. They were perfectly serviceable, but rather boring and stiff. They didn’t have the same visceral power of his lowercase. I’m relieved I took an interpretive instead of literal approach.

The first result of such an assessment must be, I am sure, to confirm the stature of Garamond, but to see him no longer as a solitary eminence but rather as a first among equals.

— Matthew Carter, “Visible Language”, vol. 19, no 1.

Looking at related fonts from a period for inspiration isn’t a new idea, but it was certainly liberating for me. I settled on Simon de Colines’s Roman Titling on Two-line English, reproduced in Hendrik Vervliet’s Paleotypography of the French Renaissance. The scholarship is thorough, but the reproductions are poor. The image quality indicated basic, gross shapes of the letters. Despite this, it allowed me to supplant the gross shapes with my own imagined detail.

I took the curvaceous bracketing of E and L for Heldane. This stroke transition from vertical to horizontal is common in garalde letters, but it’s rather pronounced in De Colines’s caps. The sharp sweeping leg of R is compelling and rather different than the straighter construction of its contemporaries. Like E and L, its stance is wide — which makes sense in the original titling context. I tamed its breadth to work with the lowercase and kept the leg. The C illustrates what I imagine the “true” shape could be: a pointy top serif anchoring a sheared, sharp exit stroke below. G is another example of imaginary development, the spur is sharp and delicate to harmonise with its siblings.

Top: Great Primer Roman. Middle: Heldane Display Regular. Bottom: Sabon.

Heldane’s numerals are informed by two asynchronous sources: Granjon’s second Great Primer Roman (1566) and Tschichold’s Sabon (1964). Like De Colines’s capitals, Granjon’s numerals were poorly reproduced in Paleotypography. I fancied his 3 — it looks like two distinct pen strokes, the top stabbing into the bottom. This gestural construction made sense within the emerging style of Heldane’s other letters, like G R and a. The tail of 3 and 5 are taken directly from Sabon, along with the open construction of 6 and 9. Angled stroke terminals within Heldane’s 3 6 and 9 retain vestigial pen traces, anchored by the horizontals strokes of 2 4 5 and 7. And, finally, the 0 is geometric and circular. This is my personal preference for the genre — I’ve always considered it the perfect typographic form for the concept of zero.

During the renaissance matching regular and bold weight typefaces didn’t exist, but is now expected for modern families. The 2-line Double Pica Roman had much darker capitals than the lowercase, an unusual contrast by today’s standards. Vervliet writes, “Plantin used them as capitals for van den Keere’s 2-line Double Pica Textura.”

Van den Keere’s roman capitals with his textura lowercase.

He continues describing the roman, “in the design of this face [he] kept to the regional tradition of bold, fat-faced Romans with a big x-height, comparable for weight with Gothic letters…” I read this half-way through designing Heldane, and it made a lot of sense. I love van den Keere’s texturas. I can feel the influence of them within his roman forms: they’re both narrow, dense and sharp. It’s like he was channeling salient blackletter qualities into his romans, rather than abandoning them altogether like his contemporaries. As Vervliet explains, “it looks as though he wanted the Roman to give the same optical effect of blackness as the Gothic.”

It is one thing to lay out a multi-weight family schematically, it is quite another to draw a workable Ultra Black with personality of its own, a process close to caricature — an exaggeration of normal features to shape forms at once massive and articulate. Without careful design, boldfaces lose definition and look as though they have been dipped in chocolate.

— Matthew Carter, “Visible Language”, vol. 19, no 1.

Van den Keere’s dark capitals indicated a possible bold weight. I needed a practical guide for a weight limit that made sense within the conceptual framework of Heldane. In theory, the vector design-space of digital typefaces is almost limitless. I can easily arrange the point structure in Heldane to generate ultra black weights. However, there is a point where a black style drifts towards parody, lowering the tone of the entire family rather than reinforcing it. I was careful to make Heldane’s Bold retain the dark sharpness of its Regular. This was achieved by borrowing articulation strategies from Lexicon’s heavier weights — essentially thinning the joins between stems and arches to maintain a feeling of sharpness.

Garamont’s Vraye Parangonne. Scan from Type specimen facsimiles II.

Van den Keere’s Reale Romaine. Scan from Type specimen facsimiles II.

Garamont’s Parangonne. Scan from Type specimen facsimiles II.

The even texture of renaissance text typefaces is sublime. Their calm, assured clarity is almost transcendental. Five centuries later they’re still fresh. That first step away from the calligraphic influence of Jenson’s type and towards something more ‘designed’ solidified their status as classic for all subsequent generations of typographers. They’re not classic because they’re old — they’re classic because they’re good.

It’s easy to fuss while drawing a letter, endlessly adjusting a few units here and there. It’s important to remember that many details will disappear at 16px on a mobile screen, but the texture will remain. Texture is what I’m actually designing typefaces for at small sizes. I’m simultaneously designing the letter, the space in and around the letter, the space between the letters, and the space between the words. The interplay of black and white determines the paragraph’s texture and is how a text typeface is judged successful or not.

Heldane Text Regular, Italic & Bold.

Heldane Text is optimised for small sizes. I prefer a darker texture in text, having explored dark and soft with Feijoa, my first retail typeface. Heldane Text retains the dark and sharp atmosphere of the Display cuts, but tempers some of the details that fall apart at small sizes. The widths of the letterforms are even, serifs are longer and the overall proportions are closer to Sabon. I tried to avoid the light elegance of Sabon and other Garamonds in general, I wanted Heldane Text to feel denser and more substantial, like Stempel’s legendary Garamond designed by Dr. Rudolf Wolf.

The value of van den Keere’s contribution to typography lay in his Black Letters, Romans, and music. Excepting an adaptation of an existing fount, he cut no Italic, so far as is known.

— Hendrik Vervliet, “Sixteenth-century printing types of the Low Countries”

The fact that van den Keere didn’t cut any italics created an interesting challenge for me, opening up interpretive possibilities for Heldane’s italic styles. Most contemporary renaissance-inspired typefaces freely combine roman and italics from different sources, and Heldane follows this pattern.

Heldane Text Regular Italic.

Stempel Garamond Kursiv.

Heldane Display Regular Italic.

Stempel’s Garamond italic was the perfect starting point. The construction of n — a fluid, curvaceous in-stroke and a long linear out-stroke — set the foundations for the lowercase. I paid particular attention to the a and k, modelling them closely. Once Heldane Text italics were settled I applied a similar structure for the Display italics, but increased the sharpness and contrast. I wanted the sharp fluidity of the italics to resonate with the tense curves of the romans.

Antoine Augereau

Heldane Text Regular Italic

Late in the development of Heldane I visited the National Library to view several renaissance books. At a practical level I saw a swash M from Antoine Augereau, which I immediately copied for Heldane — it was much better than my version. But on an aesthetic level, it helped reassure me that Heldane, as a whole, felt cohesive and natural.

After this visit I remembered an issue of National Geographic magazine that I read as a child. In it some scientists and artists were trying to recreate the head of a Neanderthal man. They dug up his skull and meticulously stuck on plasticine to recreate the muscles, sinew and skin of his face. I was fascinated by the mix of science and art.

Even though it was a fabrication — informed guesswork on their part — it was presented as fact: “this is what he looked like.” I remember feeling sad, wondering if that was true — what if he looked nothing like their creation? Because now, for all intents and purposes, that’s exactly what he looked like. We don’t have anything else to compare it to.

In Something Old, Something New Ben Kiel says, “as a typeface designer, you go grave-digging.” This is a fabulous metaphor for contemporary typeface design, minus the tedious excavation in harsh African deserts. The internet has pushed everything to the surface, it’s a virtual smörgåsbord of typographic skeletons for us to pick over and flesh out.

Like the Neanderthal face, Heldane is a hybrid, a bastard, a fabrication. I vultured my way through history picking the bones from old fonts I like to make something new. I hesitate to call it original, but it is new. But only in a strict temporal sense — that is, this exact typeface hasn’t existed before. I used to say Heldane is “my Garamond” as a shorthand explanation, despite having very little to do with Garamont’s work. But it’s very much in the garalde genre. I’ve drawn fuzzy golden threads from his contemporaries to weave my own texture.

To me, this is the point: high resolution, high quality primary sources are complete and finished. Low-quality poor images can be of more value. Secondary sources have more potential because they’re unfinished, their very nature allows room for interpretation and imagination.

And it’s not like his ghost will send a recriminating fax to the 1993 editor of National Geographic magazine. That’s the best part about working with fragments of the long dead — you’re pretty much home and hosed to do what you like. The real concern with digital “revivals” is the reverse chronology. If a designer’s first interaction with “Garamond” is Adobe Garamond, that becomes their reference point. For all intents and purposes it is Garamond: “I used Garamond for this book”. Taking the forms is one thing, taking the name is another. Adobe Garamond thus becomes the Neanderthal face: it’s presented, marketed and discussed as the Garamond instead of a version of Garamond.

Notes & references

Modern scholarship distinguishes between Garamont the man and Garamond the typefaces.

In 1926 Beatrice Warde (pseudonym Paul Beaujon) published an article in the Fleuron clarifying that the “Garamond” types in the French National Printing Office were actually Jean Jannon’s. Some type foundries inadvertently had issued “authentic” Garamond revivals based on Jannon’s.

When I finally acquired an original specimen of van den Keere’s 2-line roman capitals I was disappointed. They were perfectly serviceable, but rather boring and stiff. They didn’t have the same visceral power of his lowercase. I’m relieved I took an interpretive instead of literal approach.

And it’s not like his ghost will send a recriminating fax to the 1993 editor of National Geographic magazine. That’s the best part about working with fragments of the long dead — you’re pretty much home and hosed to do what you like. The real concern with digital “revivals” is the reverse chronology. If a designer’s first interaction with “Garamond” is Adobe Garamond, that becomes their reference point. For all intents and purposes it is Garamond: “I used Garamond for this book”. Taking the forms is one thing, taking the name is another. Adobe Garamond thus becomes the Neanderthal face: it’s presented, marketed and discussed as the Garamond instead of a version of Garamond.

More blog posts

Domaine Sans design information

Domaine was an exploration into the unpopular Latin typeface genre; Domaine Sans delves into another unpopular genre: sans serifs with contrast.

Darling design information

Darling was designed for The Darling Hotel. Inspired by a 1920’s French typeface, the design process gently updated Darling via the 1974 classic ITC American Typewriter.

Mānuka design information

Mānuka is new growth from old wood. With deviant details pilfered from Teutonic timber type, Mānuka grafts a contemporary antipodean aesthetic onto 19th century German root-stock.