Eye Magazine

As the demand for print fonts declines, can the emerging digital market fill the gap? Type designer Jack Yan looks at what the future holds for type designers and foundries

Technology has changed so rapidly that the idea that typefaces are developed for print is fast coming to a close. The technological initiates walk around with iPads, we spend more and more time looking at our screens, more people are getting novels on their Kindles, and communication has been supplemented by texts and Tweets.

Yet it has only been in the past year that we have begun to see the mainstream use of Web fonts, where designers can specify typefaces for the screen in a similar fashion to print (see Monitor, Eye 75).

It has been a long time coming, with arguments not just over technology, browser formats and standardisation but intellectual property. Is a Web font not a copy of a font – and what if the designer has not paid for an extra licence?

In 1996, some colleagues and I set up an organisation called TypeRight, dedicated to raising awareness of the intellectual property rights in typeface designs. It had been born out of the early wave of piracy, often thanks to ignorance (‘I got these fonts free with my computer’) and, in one particular high-profile mid-1990s case, the loading, size-changing, and resaving of fonts from Adobe, Emigre and others by an outfit called Southern Software Inc.

Typeface designers so rarely make megabucks from the big licensing deals that a culture of protecting one’s work arose among professionals. There were similar concerns with Adobe Acrobat – isn’t embedding the duplication of a typeface? – an argument that lends itself to the World Wide Web.

The theory was that, regardless of the scheme one employed to get a font on to the screen when it was not installed on the reader’s computer, it necessitated the duplication of a font file. If one used the early WEFT tools (Microsoft’s Web Embedding Font Tools), somewhere along the line, a copy was created. And, today, with Web fonts a reality, usually thanks to the linking of the font through CSS [Cascading Style Sheets] specifications, it is much the same story.

The demand, however, has been there since I began designing Web pages in the early 1990s. Richard Rutter, co-founder of Fontdeck, which specialises in offering Web fonts to designers, is at heart a Web designer, and is pleased at the latest developments: ‘Now we Web designers have such a wide choice of fonts, we’re back on the same terms as other designers when it comes to choosing fonts.’

Thomas Phinney of Extensis, a producer of font management tools and the WebInk Web font service, and formerly Adobe’s font product manager, says the push for Web fonts has come from Web designers: ‘If paying customers weren’t demanding it, most foundries would not have moved on it.

‘The revenue from fonts for print usage is in a slow but inexorable decline. I’m not sure anyone can honestly say whether the new world order will involve more or less total money spent on fonts, but for those foundries who do not come up with a business model that deals with the Web and e-books, they are dooming themselves in the long run.’

A similar sentiment comes from FontShop’s marketing director, Ivo Gabrowitsch: ‘The whole business has been facing a lot of changes for quite some time now: print won’t be dead in the end but will become less important while electronic media continue to gain significance. This will affect font sales as well, of course: the demand for desktop fonts will decrease and the turnover with Web fonts and fonts for mobile devices will get more and more commercially relevant. Whether the decline in the traditional areas can be compensated or even exceeded that remains to be seen.’

But it has been a long road. Phinney traces the font-face tag in CSS back to 1998: ‘Back in the late 1990s Netscape, Bitstream and Microsoft pioneered Web font technologies, but using proprietary formats.’ The sIFR (Scalable Inman Flash Replacement, which attempted to use Flash-based vector fonts to replace text on a web page) and Cufón (which re-rendered fonts using Vector Markup Language) attempted to bring Web fonts to users over the past decade. But, as Phinney recalls, ‘None of these things took hold.’

The sudden rise in Web fonts came after 28 August 2007, says Phinney, ‘when Håkon Wium Lie (Opera CTO and co-creator of CSS) wrote an article called “CSS at 10”. Despite the generic title, the entire article was a call to action on Web fonts. He basically challenged everybody to get off their butts and support font-face linking directly to regular TrueType fonts on Web servers.’

Safari 3·1 in 2008 supported font-face linking to fonts on Web servers, and Firefox and Chrome followed suit. WOFF (Web Open Font Format), the brainchild of Erik van Blokland, Tal Leming and Jonathan Kew, was the next step. According to Phinney, it suits both ‘the browser vendors not wanting anything that even smelled faintly like DRM [digital rights’ management protections]’ and ‘font vendors wanting to add metadata for Web fonts, and wanting a Web font format to be almost anything that wasn’t synonymous with existing desktop fonts’. At the time of writing, WOFF will become the new standard.

Hinting at the future

The foundries have been getting on board now that a technology has emerged. One designer who saw the light early was Peter Bilak, whose Typotheque foundry had a Web font service up before Typekit (see Reputations, Eye 75). Bilak says he was meeting the demands of his customers, ‘and there was no alternative at the time’. Typotheque still claims a broad language support (170 languages at the most recent count), control over OpenType features, and a service that doesn’t require Javascript.

‘We offer one licence for both print and Web, for the same costs as the old print licence,’ he adds, emphasising that the foundry has remained ahead of the game.

It has not all been smooth sailing. Bruno Maag, managing director of Dalton Maag, a London-based foundry that successfully embedded fonts into Web pages in the late 1990s (before moving away from the method for most of the 2000s), says the industry has been caught on the back foot and was ‘forced to be reactive’.

Maag says: ‘The entire debate about a year ago over EOT [Embedded OpenType, one of the formats proposed and pushed by Microsoft] vs WOFF proves that there is no coherent strategy within the type industry in regards to sales and licensing models for fonts.’

Dalton Maag has chosen to offer Web font licences as part of its standard desktop ones. But Maag has a concern: ‘The biggest threat from WOFF and EOT embedding is that users may not care enough to apply simple security measures such as domain binding or font-name crippling, to avoid the font being unwittingly pirated.’

It does appear, however, that the type design game is going to get even more complex. While designers could have got away with 224 characters in the early 1990s, eastern European language support by the turn of the century saw that double for some; and the technical side will only increase as designers tailor their wares to online usage.

Font hinting (in lay terms, getting fonts to appear clearly and legibly on screen) will be the next frontier.

Initial forays into Web fonts have not always gone well. When I told Phinney of my experiences with Web fonts – notably the difficulties getting them to display properly on screen – he sympathised. Windows presents the main problem, on XP, Vista and 7: without ‘significant manual tweaking’, the results may be less than stellar.

‘Manual hinting can take several days to two weeks per font style, and is a particularly specialised skill, even more than typeface design,’ he says.

This will add to the cost of development. Will designers be willing to invest the time? ‘To make the fonts Web-worthy, foundries need to invest further in their products,’ Maag adds. ‘This is not an inconsiderable investment.’

Bilak says Typotheque has developed screen-optimised fonts and will soon launch ‘parametric bitmap fonts, which will remove the need of hinting completely.’

Design aside, Maag says the younger generation consumes type differently from its forebears. ‘The Web, and other screen-based environments [Kindle, iPad, etc.], will inevitably change the demands on type. At present, the majority of users have grown up reading print and would expect the type styles they are used to seeing from print. But already we can observe changes in the way a younger generation reads type, and a younger generation of designers treats type.

‘What used to be the norm for legible type is being changed,’ Maag continues. ‘A simple example is that the RNIB [Royal National Institute of Blind People] recommends Arial at 12 point minimum as legible type, which is contrary to what hundreds of years of experience tell us.’

Nevertheless, Phinney forecasts brighter days ahead. ‘Once everybody using Windows has migrated to browsers that came out in 2011, or later, all this will no longer be such a big deal, because Windows already has rendering technologies that look pretty fabulous. But waiting for the browsers to adopt those technologies, and then the world’s installed base of users to adopt those browsers – well, it will probably be some four to six years before you can count on that stuff.’

The arguments seem vaguely similar to those aired when low-resolution laser printers first came out. Three hundred dpi netted a poor result with typefaces hurriedly adapted for the medium, giving rise to typeface families such as Bitstream Charter, which were optimised to work at low resolutions.

‘Type designers have a difficult choice: invest a bunch of work in manual hinting that will become largely moot in just a few years, or let everybody suffer with crappy fonts on screen in a lot of Windows browsers,’ says Phinney. If Web designers do care about users on Windows XP, he says, ‘it will dramatically restrict their choices for Web fonts at body text sizes, and even a bit for subheads’.

Rutter is much more confident about automatic hinting, especially when used at 14 pixels or larger (forget those point measurements: the language has been changing).

‘Where auto-hinting has resulted in some awkward hints in a few characters, we are seeing foundries [Rutter cites Jeremy Tankard Typography as an example] tweak the outlines to force the auto-hinting to get it right. These bits of manual intervention … still result in acceptable rendering.’ He says only a tiny minority of Web fonts are manually hinted across the various font services.

Maag has his concerns about quality. ‘Despite good tools available today, there is a limit as to how quickly a human can draw a letter, to a good quality that stands up to scrutiny from professionals. I often hear the argument that no one but a professional can see whether something is right or not, and, accordingly, things could be done a bit quicker with a loss of quality,’ he says.

‘It may be that the layperson can’t point the finger at a good and bad curve but the reader will subconsciously pick up the good and bad,’ Maag adds.

‘It’s an emotional response. The crudity of screen-based media may also bring an impoverishment in detailing of shapes which was necessary in print.

For richer for poorer

Despite these quality concerns, the dawn has come for Web fonts. Typeface creators accustomed to designing for high-resolution print outputs will soon need to have a change of mindset almost as great as the one that saw font design methods move to the desktop.

Gabrowitsch foresees a richer type world for users: ‘The extended usage and growing technical standards for fonts will necessitate more innovation and quality in type design.’

If the demand currently comes from the Web and e-books, designers who keep an eye on trends on the Web will be far better placed to deliver the next hits in the type design world. More so if the Web is the core medium for visual communication.

As to the better quality and innovation suggested by Gabrowitsch, the technical skills demanded may well force some level of consolidation in the type industry, or at least greater co-operation with those who specialise in preparing fonts for online usage.