When the first issue of Communication Arts rolled off the presses, magazine article texts were typically composed on hot-metal typesetting machines. Those from Linotype and Intertype cast whole “lines of type” at once, while the Monotype cast whole texts letter by letter. The machines had keyboards—no one thought about going back to the days when each word in a magazine was set by hand, one letter at a time.

Let’s define the difference between typefaces and fonts. A typeface is a design—lead letters have “faces” that are printed from—while fonts are delivery mechanisms. A metal typeface might havea dozen fonts or more per style. Akzidenz-Grotesk Bold, for example, had fonts in 4pt, 5pt, 6pt, and even up to a 60pt size or larger. Akzidenz-Grotesk also had Light and Condensed versions, etc. Yet, hot-metal machines were on their way out by 1960. Letterpress printing was replaced by offset, for which new typesetting machines based on photography were developed.

For centuries, each font size had been produced on its own; compare approximately 10pt and 12pt texts printed by John Baskerville or Giambattista Bodoni, for instance, and you’ll find subtle differences between the letters—the larger ones weren’t just “scaled up.” While hot-metal typesetting machine manufacturers didn’t always design each typeface size their machines composed with from scratch, most of them had master drawings for at least three sizes, which they scaled up and down to manufacture the metal-font matrices for all intermediary sizes.

As printers and typesetting houses switched to phototype, they had to buy typesetting machines that created the photographic positives on transparent film used in offset printing. As in hot-metal, these phototypesetting machines often had a few sizes of each typeface available, but they also allowed for fonts installed inside the machine to be resized. To save money, some businesses only bought one size per typeface, knowing they could always scale it to other sizes.

ONE TYPE STYLE, MANY FONTS
Type design moved from making many fonts for a single typeface to creating one font per typeface. Like today’s digital design applications, phototypesetting machines could artificially slant type, condense or stretch it, and even make letters artificially bolder or lighter. Those alterations might be used more often than real italic, condensed, extended, bold or light fonts from the same family, which would have been drawn with those slopes, widths and weights in mind. This was the typographic downside to leaving letterpress, and some typographers see machine-made changes to fonts as being almost criminal.

Phototypesetting technology had upsides too. According to Briar Levit, whose 2017 documentary Graphic Means: A History of Graphic Design Production documents the transition from analog paste-up work to digital desktop publishing, “good results depended on the design itself. Things could go wrong if the right darkroom techniques weren’t used while developing the film, or if the quality of the equipment wasn’t high, but many folks I spoke to who worked with phototype mentioned how beautifully sharp it was, and how tough it was getting used to digital output later.”

ONE TYPE STYLE, ONE FONT
The switch from multiple size-specific fonts to single fonts scalable to many sizes was significant. When phototypesetting machine manufacturers converted their font libraries to the PostScript Type 1 digital format—the font format developed by Adobe and used most often in desktop publishing—only one font per size usually made the transition. Helvetica, which, like Akzidenz-Grotesk, had many sizes per style in letterpress printing, was reduced to just one font per style in PostScript Type 1, an “error” only corrected in 2019 when Monotype released the OpenType Helvetica Now family. Adobe wasn’t satisfied with such a limitation, so in 1991, it expanded PostScript Type 1 to support “Multiple Masters.” In essence, this packaged several fonts in one file. The drawings for every letter inside each master were mathematically compatible, allowing users to interpolate any weight and width on the fly inside applications like Adobe Illustrator and QuarkXPress, giving them the ability to fine-tune their typesetting for the intended print size.

During the 1990s, type designer Luc(as) de Groot did corporate design work at BRS Premsela Vonk in Amsterdam and then at MetaDesign in Berlin. “Multiple Master fonts were awesome,” he recalls. “My colleagues and I experimented with them a lot, beginning with the very first ones—Adobe’s Minion and Myriad. Later, I created Multiple Master fonts for my teammates. They loved using the fonts, but they didn’t work with every printer, or with all Linotype imagesetters.”

Desktop publishing applications’ Multiple Master user interfaces were not intuitive enough for a lot of designers either, and layout applications eventually dropped Multiple Master font support. Type designers continued to rely on them when making new font families. Multiple Masters never left font editing software. “When Multiple Masters didn’t work, I’d create individual font instances for designers, but I kept using Multiple Masters to develop all my typefaces,” de Groot recalls.

ONE FONT FILE WITH MULTIPLE TYPEFACES
In 2016, Adobe, Apple, Google and Microsoft updated OpenType, the font format that superseded PostScript Type 1, which itself—long story!—was an extension of Apple and Microsoft’s TrueType format. Since then, OpenType version 1.8 has supported “variations,” which in principle are another attempt at implementing Multiple Master technology. OpenType Variations allow for the use of variable fonts, which package different typeface styles in a single file. As of this writing, variable fonts are still limited to use inside operating systems and websites; despite support in Adobe Illustrator, Photoshop and the latest beta build of InDesign, they play almost no role in print design since the “real” InDesign has no interface for them yet.

On the surface, variable fonts offer similar functionality to what large font families already provide. Indeed, they’re an alternative way of packaging existing font families, at least for the moment. I’ve selected two new typefaces to illustrate this: a serif design named Antonia and a sans named Grtsk. A seven-weight and six-width family, Grtsk was designed by Ilya Naumoff and Benjamin Blaess. Black[Foundry] in Paris, France, published it. Each weight and width has three styles—upright, italic and back slant. According to Naumoff, “the variable font format fits our approach well. It gives freedom to creatives and empowers them to experiment in new ways. Installing one font file instead of 126 is also easier, and it saves space, which is important online.”

Antonia comes from Typejockeys in Vienna, Austria. Its four optical sizes are cleverly named H1, H2, H3 and Text. Each is essentially a subfamily containing between eight and fourteen styles. Michael Hochleitner, who codesigned the fonts with Franziska Weitgruber, explains the naming system this way: “I’ve always been involved with web development a bit, and spent more time with CSS and web typography while building the new Typejockeys website. H1, H2 and H3 is the hierarchy you use to structure text online. It’s a logical system that makes sense for any kind of structure—even offline—because it’s easy to understand and language independent.”

VARIABLE FONTS FOR A RESPONSIVE FUTURE
Naumoff echoes Hochleitner’s sentiment, stating, “We wanted to design a toolbox for designers. Grtsk’s design is relatively neutral and can work in many contexts. Our approach was to create a workhorse and let designers experiment with the options.” Jérémie Hornus, Black[Foundry]’s director, adds, “Most designers will find what they need for a specific project in a couple of static fonts. Others seeking a more versatile tool for multiple projects might use Grtsk Variable instead.”

Variable fonts allow for dynamic and more responsive usage. Hochleitner reminded me that “while it’s nice to adjust weight and optical size even more finely than large font families made possible in the pre–variable font days, what’s even better is that you can control how the interpolation axes in variable fonts are used online with programming. Over the next decade, variable fonts will enable responsive web typography. A site’s CSS can send a specific optical size from a variable font to a particular device, so viewers on iPhones see the result of different variables than readers on laptops. You could also manipulate the variables so headlines and body copy could be set from the same file, but each kind of text would be tailored to work the best at its relative size. More weight would be given to smaller-sized texts, less to large headlines.”

In the future, variable fonts may be used for more than just web text on two-dimensional screens. Naumoff says, “Video, motion and animated content are more present in public spaces than ever before, and online experiences are becoming more immersive. Variable fonts will be able to truly shine when they are integrated into AR, VR and three-dimensional media.” Hornus agrees, adding that “in an era where networks and streaming are becoming ubiquitous, we see real usage cases for the new format. Graphic designers will have to move beyond the traditional vision of their craft and embrace those new media.” The technology will probably be used in media we can’t even visualize yet. While dynamic changes to font appearance might have been foreseeable when Multiple Master was pioneered 30 years ago, this all would have sounded like science fiction in 1959. Naumoff hit on something that I think makes a great closing point: “In ten years, responsive design utilizing variable fonts will be much more common than today. It’ll be fun to imagine new ways of implementing variability in fonts then too.” ca