Roger Black

The Last Blog

• Authors
  • Roger Black
  • Rob Hunter
  • Alexander Probst
  • Theo Fels
  • David Berlow
  • Walter Bernard
  • Emmy Huang
  • Amid Capeci
  • Arthur Hochstein
• Categories
  • Industry Commentary
  • Point & Counter
  • Question & Answer
  • Visual Commentary
• Archives
  • April 2008
  • March 2008
  • February 2008
  • December 2007
  • September 2007
  • August 2007
  • July 2007
  • June 2007
  • May 2007
  • April 2007
  • March 2007
  • February 2007
  • January 2007
  • December 2006
  • November 2006
  • October 2006
  • September 2006

RogerBlack.com > Blog > Fuzzy thinking and sharp type

Fuzzy thinking and sharp type

2007-06-22 by David Berlow

OR MANY YEARS now I've been aware that people designing for the web are interested in type being "sharper" on the screen. Customers have contacted our shop saying that logos, buttons, and whole fonts needed "sharpening". I think, they think, there is some magic hole that one can put a typographic image into, turn a handle, blow off the dust, and the fuzziness will be gone.

If there was a such a magic hole, it'd be built into the operating systems of today's computers, or browsers would have them, or the internet itself would take care of the problem. To take an outline of a letter, logo or graphic, and "decide" what would be the best pixels to remove, and where precisely to add pixels to create a "sharper" image for a particular use for any user, of any font in any computer — what a great thing!

But "a particular use" has a range of possible appearances through the combination of operating systems, browsers, languages, scripts, and screen displays, with user preferences sprinkled throughout, before reaching the url to be seen and becoming those pixels that form letters. And though the typographic can be made to work uniformly sharp in a consistent user environment like a corporate intranet, taking type to the world at large with sharpness and fidelity is another thing.

I've had thread encounters where I hear one user complain bitterly about fuzzy blurry type, while another user lauds the same type at the same url — only to find, that one sits at a machine with 144 dots per inch while the other is sitting with a CRT. The range of display possibilities for typographic material in the real world now has become vast; from MacOS7 to Leopard, DOS to Vista, 72 dpi to 220, millions of fonts and billions of users and most of it is various kinds of fuzzy.

But what is sharpness to begin with? In print, over the last 200 or so years, sharpness is taken for granted, i.e. the physical edges are "sharp" between what's printed and what's not, or between one color and another. Only in the days before real paper, in today's cheep flyers, grunge style design or badly registered news printing do we see fuzziness or blurriness in print. So what's with these screens, these fonts, these screen font makers?

In the early years of the internet, graphic designers complained bitterly about having to manage the quality of their web type in Photoshop as a graphic. They set the type, it didn't look good, they went to work pixel editing and were not pleased. Everyone knew why: Photoshop didn't even kern type, and if you were a dedicated pro you'd be doing that kerning anyways in your larger sizes of type, so what's a few pixels in the smaller uses of type. . . . That wasn't it, though, it was type designers who wouldn't make their fonts right that caused it all.

Despite the resistance to quality thus posed by type designers, graphic designers learned to maximize the contrast, make the letters repeat and space well, (by bitmap font editing), and this sharpening was passed on to the next generation of graphic designers as an entry level web design position. Some learned that you could put the type in different locations on a Photoshop page, get different pixels result, cut & paste the letters around to get a good word or two, and be done faster.

Designers also learned that you could use TrueType fonts, in fact many fonts bundled with operating systems composed TT typographically sharp in HTML, so you could skip Photoshop if you didn't mind limited typographic expression. Only some TT fonts worked well and this number really never grew much. With unlimited typographic expression as one of the underlying conditions of free enterprise, sharp unlimited typographic expression over a broad network of users is an issue now facing free enterprise users en masse.

Sharpness of type is a thing relative to a lot of variables. But what is clear is that the options font developers have, and thus the options font users have, are limited by a feudal attitude towards type technology. Who isn't fuzziness a problem for? The way this sharpness of type worked out is that the big font software developers: Adobe, Microsoft and Apple made fonts fuzzy for free to make things easier for themselves, each for different reasons.

If one uses pdfs, the use includes zooming, so one can adjust fuzziness by zooming the type bigger, perhaps fine tuning their view by moving back slightly, or by zooming the type smaller and moving forward, one can at least change the fuzz. Microsoft, has so many solutions to so many options, it's impossible to tell all the reasons it's types are not sharp. Apple has been following Adobe's lead on the Mac (except without zooming), sharpening some fonts for their OS and personal devices by hand tuning.

So, all of the solutions exist in outline font design (see link on RB.com), hinting (see MS/typography), or bitmap editing (see yourself doing so soon), to make everything sharp all the time and the only thing standing in the way right now is delivery. During discussion of yet another sadly inadequate embedding scheme, as in, something that just wants to call the name of a font, and not tell anything else about what the user's need's are, someone from "guess where?" said I just wanted better embedding 'cause I want more $$$$$! or was it $$$$$$$$$, I forget.

Well, properly specified embeddable fonts are the hole into which fuzzy fonts can be put and sharpened. Any fonts will work, any type designer can make them, any url can call them, and if only one more user can see them, and read them, well, isn't that worth it? It'd also, by the way, cause a market to open to competition for me, and anyone else willing to study the issues and technologies, design well, and work hard, which is more often than not, not the way things are now between user and type design provider.