Phinney on Fonts

Category: Typography

  • About .notdef: the symbol (not emoji!) that is often an “X” inside a tall rectangle

    I have fielded many versions of this question: “What does the emoji with an X in a large rectangle mean?” or “When texting, what is the meaning of the symbol with an “X” inside a tall rectangle?”

    That it is an X-​in-​a-​tall-​rectangle is just the most common rendition, although not the only possible one, for the “.notdef”. A variety of font-​specific notdefs are shown above.

    It is not an emoji, but you can get it instead of an emoji or other unusual character. A notdef (undefined glyph) is what gets displayed when a character is specified in text, but your current font does not support that character. Most commonly it happens with a newer emoji, but it could be in an unusual language or a new currency symbol, or… well, something unusual.

    This problem is not unique to texting, but applies to all kinds of displays of text on all devices, whenever the needed character is not available in the available font(s).It is an indicator for a “missing character,” which stands in for an emoji or any other character that your device or current font doesn’t have a glyph to display it with. This makes it a very special symbol.

    (A particularly hefty notdef, that I designed for one of my typefaces. The font is quite bold, so I made the outer box of the notdef quite bold to match. How font-​specific should a notdef be? As I get older and hopefully wiser, I gravitate towards what seems to be a near-​concensus view that it should contrast with the font, so users realize something is wrong. That means matching weight is actually a bad idea. Oops.)

    This is more common when you are receiving rather than sending a text, since you are unlikely to enter a character that you don’t see correctly. But it can happen when you try to view a web page, or copy text from a source and paste it somewhere that it comes out in a different font.

    Often there is some kind of font fallback available; your phone (or computer) tries to display unusual characters and emoji in some other font, that supports those characters. That is why in some situations you can see a name and some unusual character in the name is displayed in a different font than the rest of the name. But phones have limited storage space, and whether it is a phone or computer, there are over 150,000 characters defined in Unicode, with more added every year.

    So when your phone (or web browser or computer) runs out of ideas on how to display a character? You get a notdef.

    If you copy and paste a notdef on your computer or into a text or email you are sending, you will probably be copying some specific emoji or obscure character. That means that some other person who receives that (or later views the same file on their computer) may well see something else entirely!

    Even if they do see a notdef, it may look different, depending on the font they see it in. Here are the most common/​standard approaches to the notdef, as defined in the OpenType specification.

    The thinner box creates a very different appearance compared to the X-​in-​a-​box approach, doesn’t it? Note that these are general approaches, not precise glyph outlines that a font maker would use directly.

    The plain rectangular box is the style of notdef that was common in PostScript Type 1 fonts, back in the 1980s–90s. But it has fallen out of favor. I believe the main reason for this is that as a shape it is not as distinctive, and hence more easily missed by someone looking over some text.

    That plain box is the origin of the slang term “tofu” for the notdef, over some notional idea that it resembled a piece of tofu. I originally thought Google staff invented this slang, as the first time I remembered seeing it was in publicity for Google’s “Noto” universal font set (“Noto” being short for “no tofu”!). But my font colleague Denis Moyogo Jacquerye pointed to this thread on the Unicode mailing list in spring 2009, and says it was one of a number of references around that time. John Hudson seconds encountering the “tofu” term in Unicode circles, so I may have been hasty in assuming it was a Google invention. 

    The question-​mark-​in-​a-​box is used in many of Microsoft’s fonts, such as Calibri. Note how the question mark inside the notdef is in the style of the font—it isn’t just a generic one. This style was invented by John Hudson during the development of Calibri and the other so-​called “ClearType fonts,” that shipped in January 2007.

    “The ART [Advanced Reading Technologies] group had used a spiral .notdef in Palatino Linotype, but it had caused confusion because it wasn’t recognised by users as a missing glyph indicator. For the C* fonts, I suggested that a) a box of some kind was necessary, and b) a question mark would indicate uncertainty: there’s a character here, but we don’t know how to display it.”

    John Hudson on Mastodon, 17 Dec 2024

    A plain rectangular box was the style of notdef that was common in PostScript Type 1 fonts, back in the 1980s–90s. But it has fallen out of favor. I believe the main reason for this is that as a shape it is not as distinctive, and hence more easily missed by someone looking over some text and trying to spot errors and glitches. When I was at Adobe, we went from the empty box to the X-​in-​a-​box style as part of our transition from PostScript Type 1 to OpenType, from 1999–2003.

    HOW DO I GET THE RIGHT CHARACTER INSTEAD?

    Updating to the latest OS for your phone (or computer) usually also updates your Unicode and emoji support and system fonts. If the problem is in an app that has its own Unicode/​emoji/​fonts, then updating that app may help.

    Many apps and OSes will use “fallback fonts” when the current font does not support a needed character. In that case, the above advice is good: you need better support from some core system font.

    (This was originally written for Quora, but as Quora continues to turn to garbage, one of my answers on this, despite having the most upvotes, was made invisible by the system for unclear reasons. So I have merged my answers to two similar questions into one, and posted it here.)

  • “What does a design brief for a new typeface (font) look like?”

    Many of the same questions could reasonably be in play, whether one is choosing an existing typeface, commissioning a typeface, customizing an existing typeface, or designing a new typeface oneself. There may not even be a “typical design brief” for a new typeface—but there are certainly elements one should include and things to consider.

    A design brief is sometimes neither written down, nor clearly developed. I encourage both aspiring type designers and clients of custom type design projects to go through the same process: write it all down. It will be helpful, often immensely, to articulate questions and goals clearly. It sets everyone’s expectations and creates reasonable limits.

    Even in a solo project, sometimes there is a temptation to allow “goal creep” and more gets added to the project in small pieces, with a final scope that is considerably more than originally intended. Writing out a design brief can help prevent this.

    Many clients don’t know what questions to ask, so the design brief is something that usually gets developed in collaboration between the type designer and the client. Or, when there is no specific client, it means asking the questions of yourself, to better focus the design process. Being specific is restrictive, but this is likely to result in a more successful design outcome—even if the final fonts are used in ways beyond what was originally intended (consider Bell Centennial, originally designed for telephone books).

    A design brief may be a living document, revised over time during the early stages of the project as it unfolds. There may be a first round brief written in the early exploratory stages, and a later “final” brief to guide the full execution.

    In any case, when taking on a new typeface design project, some good questions to ask might be:

    Who is the client, or target customer?

    Hypatia Sans: Myself/​graphic designers. But I also wanted something Robert Slimbach would say was good, original, and versatile enough to be an “Adobe Original.” In case you are wondering what it looks like… this entire blog is set in Hypatia Sans, both body text and headlines.

    Extensis logo redesign: Software company Extensis (my employer at the time)—makes font management & digital asset management apps.

    Is it replacing a current typeface? If so, what does the client like and dislike about the current typeface? What is motivating the change?

    Neither project was approached as a font replacement. For Extensis, I was specifically trying to ignore the previous logotype (which I disliked very much), just starting over. But along the way, I made a full basic-​character-​set font.

    If they considered off-​the-​shelf options, what did they consider and what did they like about each of them? What did they dislike about each of them? Why did they not go with any of them?

    Hypatia Sans: Although not replacing a single existing typeface, I was trying to differentiate it from other geometric sans. Futura was too cold, but the classic proportions of the caps were good. Instead of imitating them directly, I instead looked to the same source, classical roman caps (e.g. Trajan) for proportions. The Futura lowercase was too cold, but Avenir lowercase was too bland.

    Extensis: I found this part incredibly helpful in the process of creating a new logotype recently for a font software company, Extensis. We looked at a bunch of specific typefaces and rejected them for a variety of reasons. In the end I took an existing typeface, Adelle (by Veronika Burian & José Scaglione of TypeTogether), and modified it quite heavily—with their permission, of course! But I used the knowledge of what my internal client and I liked about other typefaces to guide what I did to the pre-​existing typeface. The logo is wider, a tiny bit lighter than the Thin weight of Adelle, and 5 of the 8 letters have significant design tweaks. But it did start with Adelle.

    What is the typeface a vehicle for? What is to be communicated with it? In what way should it flavor the message? Is it intended for a particular project or product?

    Extensis: We wanted it to feel modern and somewhat techno, yet warm and approachable. We had a very playful graphic for the logo—it was almost wacky. We needed the font to be playful enough to not clash with the graphic, but still be serious, to ground it all. It was a balancing act.

    Is there a specific target usage? 

    E.g. “advertising headlines” or “body text in all publications and online.” Even if not…. What sizes will it be used at? In what media? How will the type be reproduced (imaged, rasterized)? On screen? For web pages? In print?

    Extensis: The logo needed to function at pretty small sizes, as logos often do. Some of the typefaces we had considered were dropped because their weight got too spindly at small sizes on screen… they were not holding up well enough across all use cases.

    Hypatia Sans: Originally I intended it for display usage. I imagined it being used for product packaging, maybe some logos. Then I found it worked surprisingly well even at larger text sizes. So I revised my plan and spaced it so it was OK in larger text sizes (like 12-​14 pt in print). So, moderate amounts of body text, through to larger display sizes. Should look good on screen, but with details that will be interesting in print.

    What else is known about the desired design category?

    Extensis: We had decided we wanted something in the line of a slab serif typeface, something in a realm defined by typefaces such as Archer, Donnerstag, Vista Slab, and Adelle.

    How many styles (individual fonts) are desired? 

    Regular, italic, bold and bold italic are four fonts right there (and no, you can’t get reasonable quality results by just using algorithmic slanting and bolding.) More weights, more widths, or other variants (eg different optical sizes) can all add up. Families of 8–20 fonts are common. The largest family I know of is Kepler, comprising 168 fonts!

    Hypatia Sans: I wanted a wide dynamic range of weight, and ended up with six weights and their matching italics, from extra light to black.

    What kind of language coverage is required? 

    Any other particular character set needs (e.g. particular symbols, math capability, whatever). There are a variety of semi-​standard character sets and language groupings, but the whole matter is a bit fuzzy around the edges. A basic but complete western European character set might include over 200 glyphs. With central/​eastern European accented letters (“extended Latin”), you would end up over 300. 

    Each of these choices involves either choosing to adopt somebody else’s pre-​packaged language coverage definitions, or extensive research of your own. And some choices are more complex than they first appear: if you do Greek, do you also do polytonic Greek? If you do Cyrillic, which languages do you cover? (Cyrillic character sets are almost as complex as Latin.)

    For Hypatia Sans I was completely out of control. Latin, extended Latin, and even more obscure. Cyrillic, extended Cyrillic… I ended up further formalizing and extending Adobe’s character set standards for Latin and Cyrillic because of it! My manager stopped me when I was considering Norse runes (I am not making this up, I swear). Still, it was too much and I regretted it later, when what seemed fun for one style became a ton of work, for the full range of weights and italics too. Plus, the project became so big and slow that I advanced massively in skill before I was done, and found myself redesigning some things, or just seeing things at the end that I wished I had done differently. In retrospect, I could have advanced my skills more efficiently/​effectively by doing multiple smaller projects.

    What kind of typographic extras (characters/​glyphs) are required, or might be desirable? 

    Arbitrary fractions, both lining and oldstyle figures in both tabular and proportional widths and the five f-​ligatures – fi fl ffi ffl ff – are now “basic” for me. But others might think of them as extras. I think of small caps as extras, especially if there is a large language support requirement. Superscript and subscript numbers? A full set of letters for ordinals? So many possibilities!

    Create a glyph set definition 

    Now that you know what you want, consider documenting the glyph complement /​ character set fully, perhaps with a spreadsheet. If there are common characters not covered,  that too should be mentioned or highlighted some way, either by the spreadsheet or in accompanying text.

    Hypatia Sans ended up with something like 2700 glyphs per font ( 3000 after it was updated to match later character set standards). That is why it took for-​bloody-​ever to complete. I hope everyone learns from my errors! Not that you shouldn’t ever do a huge project, but just perhaps not as a first (or second or third) typeface.

    Many of these things essentially multiply together. For example, if you need ‘real’ small caps, you should probably have them for all the supported languages, and in all the fonts in the family. This kind of extension of features to the full font is often assumed, but it is best to be explicit about it, so it can be part of a delivery checklist. It is even more important to be explicit if there are inconsistencies either within a font (small caps only for un-​accented Latin?) or between fonts in the family (small caps only for the upright styles but not the italics?).

    For reference

    Adobe character sets: Latin (5 levels), Greek (2 levels), Cyrillic (3 levels). These do not include “typographic” extras such as small caps, oldstyle figures, or additional ligatures beyond the most basic (fi and fl). But they are fairly comprehensive for language and basic symbol coverage.
    Thomas’ page of type design resources!

    Note

    This is a much edited version of what was once a Quora anwer. Special thanks to Dave Crossland for edits and input when we used this in our Crafting Type classes! Any errors or omissions entirely my fault. Also see discussion about typeface design briefs on Typedrawers.

  • More of my fonts/​typography answers coming here!

    Back around 2013–2021, I really liked Quora. I could go through questions people asked, find interesting questions that I was especially qualified to answer (mostly about fonts and typography), and write up an informative answer that people would read and upvote. Comments and further questions in the discussion to each of my Quora answers would help me refine and improve them.

    I could similarly find and read interesting thoughts and analysis on darn near any topic of interest. I accumulated favorite writers and would often read Quora as a leisure activity, just for intellectual interest. It was a social network focused on ideas. 

    Quora has since mostly turned to junk. At this point, I don’t even have any confidence as to how long the site will even continue to be up. Or if it is still up, will the content be freely available? Findable? So I have been revising and saving my best answers here, to trickle out over time.

    Why did Quora turn to crap? This is the process Cory Doctorow calls enshittification. They were getting participation and views, but without making money. The site needed to figure out how it would make money. So they experimented in a bunch of different ways to try to increase “reader engagement” and reduce costs… and in the process lost everything I liked about it.

    In the case of Quora, some particular things included: 

    • Making it harder to find the answers to a question. The default when you click on a question is now to show “all related” answers instead of the answers to the current question… which means you have to read more and click more to get the info you actually want.
    • Rewarding people for asking questions (through the “Quora Partner Program”) instead of focusing on getting good answers. Worse, specifically rewarding questions that upset people as long as they got responses. Responses pointing out problems in the question itself are still responses, so… you can see where that path leads.
    • Allowing anonymous questions. Yes, there are totally legitimate reasons for question-​askers to want to be anonymous. But the proportion of junk and trolling skyrocketed after this change.
    • Doing all the above while reducing their staff of moderators.
    • Most of the above factors contributed to spam, misinformation, and low-​quality content
    • At first, it was just more bad content. BUT, the spam, low-​quality content and user-​hostile interface decisions drove away many of the best contributors (myself included, obviously). So now the fraction of “good content” is much worse, not only because the denominator grew out of control, but the numerator also shrank.

    Quora’s “Top Writers” program lasted from 2013–18, coinciding with peak Quora. There weren’t any huge perks, just a little recognition, a badge on your Quora profile. But still, it was nice. It was not a big enough deal that it made it onto my c.v.—but I did link to my Quora answers. I just deleted that link on my c.v., because being associated at all with Quora seems like a negative.

    But now Quora has been overrun with spam and fake questions plus hate speech and bullying. I won’t yank my existing answers (already often revised and polished over time), but I am polishing and further revising the best ones, and will post them here. I have already copied a couple dozen of my best answers to draft posts here, edited and posted the first couple, and done some light editing on most of the others (with more to come). I may set them to auto-​post periodically.

    That all makes me sad, but at least I was able to slurp up a whole bunch of my existing content and plan to re-post.

  • Wired mag off-​base on Roboto typeface

    Wired magazine’s puff-​piece on Google’s Roboto typeface revisions is really bothering me. I thought if I held off, I could just do a few sarcastic tweets and be done with it, but no.

    I am not a huge Roboto-​hater like some folks in the type community. I just object to uncritically publishing quotes that make blatantly false statements.

    “UIs [user interfaces] are crafted from images and type,” Matias Duarte, Android’s head of design tells WIRED. “But the idea of having a typeface that’s thought out as a UI typeface—that’s not been done before.”

    Well, that’s pretty much simply false. (UPDATE: Duarte says he thinks he was misquoted, basically he was trying to just say UI typefaces are hard, and Roboto had a particular challenge in needing to work in a wide range of contexts and types of devices.)

    [Perhaps not Duarte, but apparently the Wired author was] unfamiliar both with an obscure operating system called “Windows” and its typefaces Segoe UI (introduced in Windows 7) and Tahoma (introduced in Windows 95), both of which were specifically designed/​intended for UI usage. Not to mention Chicago, developed for the original Mac OS back in 1984. (UPDATE: Plus, there is Prelude, designed by David Berlow and Font Bureau as a UI typeface for the Palm Pre operating system—when Duarte himself was in charge of UI for the Pre. Not to mention Android’s own Droid Sans, also designed as a UI typeface.)

    A slightly weaker argument could be made for Lucida Grande (the Mac OS X UI font), which is only slightly tweaked from Lucida Sans. Of course, Lucida Sans itself was specifically designed for low-​res screens and the like. Designer Chuck Bigelow got a MacArthur “Genius” award for his work on the family.

    There are seven substantial paragraphs to the article, but both the people quoted are on the Android team. Thus it avoids mentioning the most famous thing anybody has said about Roboto, ever: Stephen “Stewf” Coles calling it a “four-​headed Frankenfont” in a strong attack on the design philosophy behind it.

    This is also why there is so much puffery throughout the article emphasizing how the typeface is designed for performance rather than aesthetics. Such choices do certainly explain most of the changes from v1 to v2 of Roboto, but “performance over aesthetics” is clearly false as a general proposition about the typeface. My big problem with Roboto is that the choice of closed counterforms for many letters and numbers (35CGSacs) is an inherently anti-​legibility choice. Yes, they had more of these before the revision, and some (5) have been slightly improved, but they need to finish the process of transforming it into a different typeface if they want it to be an outstanding UI typeface.

    Roboto compared to other UI fonts

    Indeed, I would argue that such closed shapes are stupid bordering on criminal in a user interface typeface. There is a reason that most other typefaces specifically designed for user interfaces have used open counters, and that is because there is massive evidence that tells us these shapes are more legible (see for example the research cited in Sofie Beier’s book on the subject). Legibility should be a the paramount concern for a user interface typeface.

    Roboto designer Christian Robertson explains the mix of open and closed shapes as saying that they create an appealing texture in body text. Which is lovely and all, but in a user inferface, that is not as important as legibility.

    That said, to be fair, Apple is doing a much worse thing in choosing Helvetica Neue as their UI typeface, first for iOS and soon for the next version of OS X. They too have gone to lengths to declare publicly how they are optimizing it for legibility, which is rather like trying to polish a turd. Helvetica is inherently anti-​legibility. The only way to make it otherwise would be to change it so much that it doesn’t look like Helvetica any more. Sadly, that is not what Apple is doing. (Update: Apple later changed to San Francisco, which has the same “modern” look as Helvetica, but with just slightly more open counters… rather like Roboto in that regard.)

    Aside from the business of being first with a dedicated custom UI font, if Google and Apple were to explain that they are making their UI font choices for design reasons, that’s fine. But when they (or Wired) start touting the awesome legibility and functionality of their choices, I have to call them out on it. Nonsense.

  • How to Tell If a Font Sucks

    Are you a user of fonts who needs to tell if a font is well made, or an aspiring novice type designer? The March–April 2014 issue of Communication Arts features my article on evaluating font quality, “How to Tell If a Font Sucks,” on p. 24—now online as well!

    It looks like it is hard to see the subtleties in some of the graphics in the down-​res web-​ified version of the article, though the print mag looks great. I will see about posting a version with high-​res images in PDF.

    I’m really pleased with this article. My new editor Robin Doyle at CA did a great job helping me clarify some points and figure out where more graphics were needed.

    That said, there are some corner cases and subtleties around this discussion that I didn’t have time or space to get into in the article, which was already long and involved. But that is what blogs are for. 🙂

    Although I stand by everything in that article, typefaces that are deliberately naïve/​unsophisticated are one place for legitimate exceptions to some of the guidance I give in the article. For example, I had a lovely discussion with some folks who made a typeface based on some classic road signs. The original signs did not use optical compensation at stroke joins (point 5 in the article), so they didn’t do it in the typeface either. Although I might rarely be interested in going that way myself, I have to agree that it was a perfectly legitimate design choice, given the origins of the typeface as a signage revival—even though in many another context I would be calling it crap!

    Optical compensation at stroke joins is also specific to certain typographic traditions. Certainly for Latin-​based fonts (English, French, German, Hungarian, etc.) it is nearly universal, as it is for Cyrillic (Russian, etc.) and Greek. But some writing systems do things differently, such as Devanagari (used for Hindi, Marathi, Sanskit).

    Non-​western writing systems can also change other assumptions. For example, the idea that straight-​to-​round transitions  (point 6 in my article) should be very smooth is very much not the case for Thai.

    Anyhow, check it out and let me know if I can clarify anything else!

  • Fonts and Typography for Writers and Non-Designers

    What are some good resources for non-​designers, who perhaps write, edit or publish professional documents? Somebody recently asked this in the comments to my blog. There are quite a lot of resources I could suggest, but given limited time, we should limit the complexity/​depth/​scope of the resources. So given that….

    Before getting into the depths of font selection, teach typography. I think Matthew Butterick’s Practical Typography is a great place to start. Short, straightforward, no-​nonsense, useful, and little I could disagree with.

    After that, for an intro to selecting and combining fonts, this article from Smashing Magazine is good.

    At the next level of complexity, there are plenty of good longer introductions, mostly aimed at designers. Ellen Lupton’s Thinking With Type is a good start here. Nothing wrong with reading Butterick first, before moving on to this, btw!

    For more advanced thought, the closest thing to a typography bible remains Robert Bringhurst’s The Elements of Typographic Style. It is better as a reference book or to read a chapter at a time, rather than try to take it all in at once.

    The original query from a business writing teacher at the University of Colorado, Colorado Springs:

    I teach business writing at a university and we have a document design unit. I try to get the students to understand fonts, but don’t have a good exercise, video, material, etc. about effectively using fonts. DO you have any tips, links, etc. that I might be able to use with the students to help them discover fonts beyond Times Roman and Arial and understand how to use them effectively? Thank!

  • Save $400M printing cost from font change? Not so fast…

    I am really bummed that the idea trending hot online now, popularly represented as “the US government could save $400 million dollars a year by switching fonts,” is a bit off-​base. It is not the change of design that saves toner; it is that their chosen font is smaller at the same nominal point size than the comparison fonts. Not to mention that the $400 million figure being bandied about is not actually the main number suggested by the kids, which was $234 million. Unfortunately, those fonts that use less ink/​toner at the same actual size are generally less legible.

    That said, it is great that middle school kids (the study has two authors, although one has gotten all the media attention) are doing creative problem solving and applying scientific thinking! No sarcasm intended. It is not their fault that non-​obvious aspects of the problem mess up the idea.  (Readers of my blog may remember that point size and font size have a rather nominal relationship.) Garamond* lowercase is about 15% smaller than the average of the fonts they compare it to, while its caps are only about 7.5% smaller. So it is no surprise that it uses less ink at the same point size.

    This is why most scientific studies comparing typefaces first compensate by resizing the fonts to eliminate differences in the lowercase height (called “x-​height” by us font geeks). This study failed to do that. As a result, they actually get results that are the exact opposite of other studies. Century Gothic has a very large x-​height, so printed at the same nominal point size it uses more ink than Times. If it were instead printed at the same x-​height (as in other studies), due to its relatively thin strokes, it would use less ink.

    Setting any font 15% smaller would save 28% of its ink usage. This is because the font letters are two-​diemensional, so the ink usage is based on the square of the size:.85 x .85 = .7225. Of course, there are some caps in the texts as well, which would make the savings a bit less. Interestingly, this is pretty exactly much what the study found. So, you could just as easily save ink by setting the same font at a smaller point size.

    For a moment though, let us pretend that the study did in fact equalize the x-​height, and found that a typeface change saved noticeable amounts ink. With a “normal” typeface such as Garamond, this would mean that the strokes making up the font were just thinner at the same size (“stroke” is a virtual thing here; modern digital fonts essentially trace the outlines of the letter). If that were good and useful, why not go further? Why not make the strokes even thinner? Maybe there is no font bundled with common operating systems and software that would meet these needs, but one could just commission one. Even a master type designer could do a basic four-​member family for $100K or so, which is a lot less than the hundreds of millions at stake. Make it razor thin and save even more!

    But any of those changes, swapping to a font that sets smaller at the same nominal point size, or actually reducing the point size, or picking a thinner typeface, will reduce the legibility of the text. That seems like a bad idea, as the % of Americans with poor eyesight is skyrocketing as our baby boomers (and even their children, like me) age.

    Aside from that, the reduction in toner/​ink usage probably would save less money than claimed in the study. The claim is based on the proportion of total cost of ownership of a laser printer that goes to toner. There are sadly two big problems with the idea that using less ink (or toner) will save that amount of cash, based on that proportion.

    First, large offices that use printers and copiers do so under a maintenance agreement that includes the cost of toner. They pay per page printed, and actual toner consumption is generally ignored. In such cases, a font change will only save based on the page count, not the toner. (Certainly, smaller fonts can also use less paper—I will get to that.)

    Second, the study makes the interesting claim in a footnote: “Ink and toner are used synonymously in this study. Even though traditional ink is more expensive than toner, a focus on determining the percent savings in cost rather than the magnitude of the cost obviates this difference.” Urm… how? They are assuming that the percentage of printing cost ink or toner accounts for is the same for all classes of output.

    This is untrue. Many of the documents that account for a substantial percentage of the government’s overall printing costs are printed on a printing press, using offset lithography. For offset printing, the percentage of the cost of  that is associated with ink is in fact much smaller than for laser or inkjet printing. But it isn’t a fixed percentage, either, due to the large proportion of the cost that is associated with setup. It will be a higher percentage for short runs, and lower for long runs. Additionally, because of the huge cost of owning printing presses, many or most offset litho jobs will be printed out of house, using third-​party printers.

    So, for in-​house printing-​press printing, the savings will be a much smaller proportion than the quoted 26%. For outside printers, they will not charge based on minor variations in ink usage; they just check things like whether it’s a page of text vs graphics. Either way the savings will be less.

    There is a different way an effectively smaller font will definitely save money: by allowing multi-​page documents, especially long ones, to take fewer pages! So maybe it all works out—if you don’t worry about legibility.

    There is another practical issue with Garamond in particular. The version bundled by Microsoft (from Monotype Imaging) does not have a bold italic, which is an unfortunate lack if one wants to promote its use for all government documents. (Yes, you can turn on bold and italic in your word processor anyway. You will just get a faked font instead of the actual one, which is ugly and less legible.)

    The question that should be asked is: what font and size combination could be used to maintain or increase legibility while saving money on printing, by reducing page count and/​or ink/​toner usage, with a font that is bundled with common apps (or free), and has all the required font styles?

    But that is a far more complex question, and most folks covering the issue much prefer simple and appealing messages like “high school kids tell gov’t how to save $400 million!”

    I like innovative ideas to save money. Really, I do. But I wish the media and public had consulted some experts on this area before going nuts promoting this idea, because it just doesn’t hold water—or save money—without losing legibility.

    Thomas is currently CEO of ATypI, the international typography society, since 2004. In other relevant background, he was a teaching assistant for a senior level stats course in his second and third years of undergrad, has an MBA from UC Berkeley, and an MS in printing, specializing in typography, from the Rochester (NY) Institute of Technology.

    Updates & notes

    This post has seen some editing for grammar, clarity, adding a few more details, and to be less of a jerk. Also to update my background to be current. Again, I am impressed as heck that a high middle school student is attempting serious research. I would not be analyzing it critically ,like a serious adult study ,if not for the fact that the media initially largely embraced it uncritically as if it were.

    * The student study does not specify which Garamond they used, but it was obvious (to me) in the samples that they were using the Monotype version that is bundled with Microsoft Windows. Because Garamond goes back to the 1500s, and there is no trademark on the name, there are literally dozens of typefaces by that name, with about four or five being fairly common.

    Since I wrote this, there has been some interesting coverage. The Guardian UK was in with the initial pack, with some caveats, but then their Nadja Popovitch wrote about this blog post and interviewed Jackson Cavanaugh of Okay Type for his reaction and analysis.

    Meanwhile, John Brownlee did a nice job of explaining the point-​size part of my analysis in layman’s terms, for Fast Co Design.

    I did more elaborate checking on the study’s original sources and found that their five government test documents each used different body text typefaces: New Century Schoolbook, Minion (with Myriad headlines), Melior with a little Helvetica, Times with Helvetica headlines, and Book Antiqua. The average of these was almost identical to my original estimate using two of them, but I updated my numbers appropriately.

    Given that the five source documents all use different fonts, one could reasonably wonder if they are a representative sample. Generally, as a rough guideline, you need a sample of about 30 to get sufficient statistical reliability for something like this.

    CNN quoted Suvir: “”Ink is two times more expensive than French perfume by volume,” Suvir says with a chuckle.” This may be true, but that stat is not original to him—it dates back ten years, and is specifically about inkjet printer ink. Such printers may still be common in schools (although even there I expect laser printers are taking over), but government agencies are definitely not using inkjet printers for much of their output. Most high-​volume government printing is on laser printers, or even printing presses, whose ink is even cheaper still.

  • Without serifs: a sans by any other name

    Recently, a question by Cynthia Batty on the ATypI mailing list led me to do a quick survey on what we call typefaces without serifs. Click here to take survey.

    Here are the results of my little survey. There have been over 300 responses. It’s certainly not a random sample, mostly people deeply involved with typography in some way. Interestingly, the results didn’t really vary by expertise level. (The order of the possible answers was randomly varied so as not to influence the answers, btw.)

    • “sans serif” 68%
    • “sans-​serif” 27%
    • “sanserif” 2%
    • other 3%

    The most common comments under “other” were that it should be “sans serif” as a noun and “sans-​serif” as an adjective (for example, “a sans-​serif typeface”). Certainly if the noun form is “sans serif” then standard English usage would dictate that the compound adjective would be hyphenated.

    Another common response was that “sans” is an acceptable informal shorthand for “sans serif.”

    Finally, it seems that despite a bit of solid support in the UK for “sanserif,” that spelling is neither particularly widely used nor accepted. The Oxford English Dictionary accepts it and dates it back to 1830, and the Oxford University Press, Robert Bringhurst (The Elements of Typographic Style) and eminent professors James Mosley and Michael Twyman all use it, as does typographer and typography author Robin Kinross. I must confess to not much liking “sanserif” myself.

    [Edited to correct spelling of “Mosley” and add a little more detail on “sanserif,” and again to add Bringhurst the to list of “sanserif” supporters.]

  • ATypI talk submission deadline April 30

    Speaking of end-​of April deadlines, there are only four days left for talk submissions for the ATypI international typography conference in Reykjavik, Iceland in September.

    As always, it will be a great event with tons of fascinating and varied content about fonts, typefaces and typography. Going as a speaker usually nets you free conference admission, which saves you hundreds of dollars. It also gives people you don’t know yet a reason to talk to you about something you are interested in… recommended!

  • Point Size and the Em Square: Not What People Think

    It’s easy enough to determine that a point is 1/​72 of an inch, and used to be about 1/72.27 in the days before digital type. But the challenging question is, when you look at printed type on a page, what part of a 12-​point font is 12 points high? The short answer is “none.” Seriously. For metal type it’s the “body” which is not something you see in print, and for digital type it’s the “em,” which is completely virtual.

    Font Size Measurement Confusion

    The background to this is long and complicated, so I hope you’ll forgive me if I first explain how this is the question that just refuses to die, and the confusion it can cause… in painful detail.

    • In current litigation, Microsoft is contesting Apple’s trademark of the term “App Store.” In the latest salvo, Microsoft says Apple’s most recent brief in the case is too many pages, and in too small a font size. Interestingly, they say how many pages it is over by, but they don’t actually mention the exact font size. I am not convinced this is because of the measurement issues described elsewhere in this blog post, as one can easily check the font size in Acrobat (except for docs that are scanned in to create a PDF). The font size is supposed to be at least 11 point, and at a cursory examination it appears to be… 10.98 pt (according to the Acrobat Touch-​Up text tool, anyway—I didn’t spelunk the PDF the way I would have if I were advising one side or the other). What happened? I can’t be sure, but I will note that there are some workflows that can cause this kind of minute shrinkage inadvertently. For example, when a PDF is printed, and Acrobat “helpfully” tries to make sure the document’s margins fit within the printable area on the currently targeted output device. Perhaps that (or something similar) happened here. It seems unlikely the person making the doc did it on purpose, since they did not fit within the page limit anyway, and many applications do not even allow people to adjust type size in increments that small.
    • The other day I looked at a bunch of lesson slides from a university-​level typography course. One of them claimed that the distance from the baseline (bottom of a letter such as H) to the cap height (top of a letter such as H) was the point size. I wish that were true, as it would be much simpler than the reality. On average, the cap height is about 70% of the point size.
    • When Apple first released the Zapfino script font, they sized it relative to the largest and swashiest capitals in the font. But this made the size of the lowercase letters look very small indeed, relative to most other fonts. Around 2002, they revised it for Mac OS 10.2, so that for any given point size, it was 2.5× as large. This was mostly a marketing/​usability decision; neither version is more “correct” from a technical point of view.
    • In a closely related issue, the “em” is a typographic measurement equal to the current point size (usually an “em square” or “em quad” in two dimensions), but since it relates to point size, it too has no precise measurement relationship to anything one sees in print or on screen. I have sometimes had difficulty convincing non-​typographers of this fact, notably for the relevant article on Wikipedia.
    • Some years ago, I was contacted by the San Diego District Attorney’s office. A snail-​mail spam-​scammer was mass mailing a document that included a legal disclaimer, which the scammer was trying to make as unobtrusive as possible. The legal disclaimer was required by law to be in 12 point type. The font turned out to be a free version of Empire, with its ultra-​narrow and super-​thin caps and small caps. I downloaded it and as best as I could determine from a physical print-​out comparison, it had indeed been printed at 12 points. Of course (for reasons described more below), one could easily have modified in the reverse of the change Apple did with Zapfino, so that 12 point type would be half as big when printed. But the real problem was that Empire is an ultracondensed sans serif, rather like what one often sees in movie poster credits, and is pretty well unreadable at 12 points. If the objective was to get people to not notice the legally-​required disclaimer, the company that wrote the letters did a great job, and seemed to have done so within the law as far as point size was considered. I told the DA’s office I was sorry, but I didn’t think I had anything that could help them.
    • A couple of months ago, I was contacted by New York City lawyer Brad Richter about pretty much the same issue. Recently passed legislation around Power of Attorney in New York state requires that forms granting power of attorney be printed in 12 point type. Brad had done enough reading to strongly suspect the truth: point size doesn’t relate to anything specific in size of printed letters! Yes, given a specific font, the size of 12 point text in print is related to the font data. But 12 point in one font can be bigger or smaller than 12 point in another. If the objective was to provide useful guidance to people using typical fonts, then I’d say the law is just fine. But if we take that the objective is, as Brad described it to me, to legislate the “literal size of text – a minimum physical printed size so that the elderly can easily read the form,” then the law is useless. (Below I propose some wording that might come closer to achieving the desired effect.)

    Historical Background

    Back in the days of metal type, the answer was simple, even if it didn’t relate to anything one saw in the printed output. The point size of the type was simply the height of the metal body the type was cast on. Additional line spacing was added by means of thin strips of lead between the lines, hence the term “leading” (pronounced “ledding”) for line spacing.

    Metal type, showing point size

    Above is shown a piece of traditional metal type (photo courtesy Daniel Ullrich, licensed under Creative Commons Attribution Share-​Alike 3.0). The added red bracket shows the body height, which one would measure to determine the type size.

    In metal type, without leading, the distance from the baseline of one line to the next would be the same as the point size. However as you can see in the example, once the metal type was printed, there was no direct means of knowing what the original point size had been, unless one also knows either the original typeface or the amount of leading used with some certainty.

    Today’s Answer & Implications

    In digital type, the font’s letters are drawn on a grid, where an arbitrary number of units (often 1000 or 2048) are set to equal the “em” which is then scaled to the current point size for output. So to get 12 point type in print, with a 2048-​unit em, that digital space is scaled so that the 2048 units in the design space are equal to 12 points. As Karsten Luëcke put it in a recent discussion on Typophile:

    In digital type, the EM does not refer to a “real” box. You better consider the EM as a yardstick – an abstract letter-​height yardstick which establishes a link between micro and macro level, between font-​internal unit system and font-​external unit system: The font-​internal unit system is defined via UPM, i.e. as the number of units per EM. It is the letter-​design grid or resolution. The font-​external unit system may be typographic point, millimeter, pixel, etc. And this abstract EM serves to project the font-​internal unit system onto the font-​external unit system.

    An example. You have a font with 2048 units per EM, internally, which is to be projected on 12 pt type size, externally. So 12 pts = 2048 M-​units or 1 M-​unit = 12/​2048 pt.

    So to image the font at 12 point, one scales the abstract EM to equal 12 points.

    The catch for purposes of measurement and standardization is that while there are some restrictions on how large one can draw letters in the design space, there is no necessary and required relationship between the size of the letters and the em. On average, with Latin-​based languages such as English, the “cap height” of capital letters is about 70% of the point size, and the “x-​height” of lower-​case letters is about 70% of the cap height, or about half the point size. But (and I cannot stress this enough), those are only averages, and there is no technical requirement whatsoever that one be close to those averages. Indeed, x-​height relative to cap height is one of the ways typographers describe typefaces (“high x-​height” vs “low x-height”).

    [UPDATE : I did some research for a client, and verified that as expected, cap size varies substantially between different fonts. In my sample, cap size was most usually 62%-78% of the em square, averaging right around 67-​70%. Or to put it another way, if you take an “average” font printed at a given point size, other fonts at the same point size will commonly have capitals as much as 10% smaller or 10% larger than the capitals from the average font. At the extreme you can find fonts “in the wild” with caps barely over half the average size! (I expect you could also find fonts with caps close to half again the average size, but I wasn’t looking so hard in that direction.)]

    Moreover, the Zapfino example given earlier shows how a given font could be at a radically different size relative to the point size and still be a legitimate font. Indeed, anyone knowledgeable in modifying fonts could in a matter of minutes, take almost any font and create a modified version, with the only visible difference being that text at a given point size is only a fraction of the size.

    What About the Web?

    The web can use points, but just defines them in terms of pixels. It has inherited the Windows definition of that ratio, so on the web by default 1 pt = 4/​3 pixels, so 12 pt = 16 pixels (but see below).

    It used to be that Mac browsers used the Mac relationship of points to pixels, which was one-​to-​one, but that has been abandoned just a few years ago. so at least points vs screen pixels are now consistent across platforms, though how big a point is on screen (or a nominal browser pixel for that matter) depends on your screen resolution, what zoom level your browser happens to be set to at the moment, and (on Windows) whether you have set something other than the default screen resolution of 96dpi.

    But the relationship between pixels and points is broken in some browsers on Windows (such as Internet Explorer 7 and earlier) when the user has a non-​standard resolution set. For example, if you actively tell Windows your screen resolution is 120 dpi instead of 96 dpi, that means that point sizes get multiplied by 5/​4, but sizes in pixels do not. So at 120 dpi, a font set to 9 pt will instead show up at 15 px, but a font set to 12 px will still be 12 px, and now smaller. Arguably this is a reason never to do font sizes in px. (Bitmapped grapics generally are not scaled by the 5/​4 ratio in browsers, but they are in other apps such as Word or the usual graphics previewing programs.)

    This may get even less standard in the future, as CSS 3 is threatening to make pixels a truly imaginary thing, always equal to 4/​3 the point size. This would cause pixels to scale into virtual pixels when non-​standard resolutions are set.

    Of course, some users (like me) are constantly changing the zoom level in their browsers, which also plays hob with any notion of fixed sizes for points, though at least relative sizes are maintained by browser zoom.

    Things get kinda weird on the web, in another regard. CSS can use “ems” as a measurement unit. Okay, that makes sense, right? I mean, why not set an indent or margin in ems? No problem. Where it gets weird is that you can set the type size in ems. Now, logically based on the “normal” definition of the em, this makes no sense, because the size of an em is always the same as the type size, so the size of the type is always one em. But CSS allows you to break that assumption by setting an em to some specific number of points or pixels, and then setting the type size to some multiple of that. It gets even weirder, actually, because you don’t need to define the em in the first place. If you don’t define it, the standard browser assumption is that one em = 16 pixels (Firefox and possibly Chrome), or 12 points (Internet Explorer). The difference between IE and the rest doesn’t matter with default Windows resolutions, but it gets interesting at non-​standard Windows resolutions because IE then scales the default em, while Firefox does not…. Ouch.

    [Note: edited and expanded this section several times on 21 March 2011 to better reflect system scaling setting issues. Thanks to Beat Stamm for pointing out the omission and helping me with details I hadn’t yet encountered.]

    How to Legislate Type Size Today?

    First, a disclaimer: One can implement reasonable precautions, but it’s not possible to stop determined people with sufficient knowledge of fonts and typography from creating customized fonts, which can in turn be used to create either illegible documents, or disclaimers that most people would never read. To even attempt to cover all possibiities would probably yield many pages of added law, which frankly somebody like me could probably still find a loophole in with a moderate investment of time and thought. What reasonably can be done, however, is to make the laws tight enough that it would take significantly more expertise, creativity and effort to work around them than is currently the case.

    So what variables does the law need to control when it wants to legislate a minimum size and legibility?

    • Instead of (or even in addition to) declaring a minimum point size, one could declare both a minimum cap height (defining that as the height of the smallest of the capital letters A-​Z), and a minimum x-​height (defining that as the height of the smallest of the lowercase letters a-​z), both in physical units. For example, one could require a cap height of at least 7 points and an x-​height of at least 5 points, which would be met by 12 point type in most everyday text typefaces.
    • As evidenced by the case described above from San Diego, adequate width also needs to be legislated. You don’t have to be a font geek to go out and license an ultracondensed font. One way of avoiding this would be to say that the total advance width of the letters a-​z and A-​Z, at the chosen font and size, meet some minimum. Times set at 12 pts clocks in at roughly 208 pts for A-​Z and 143 pts for a-​z. After checking many other fonts, I believe one could go with minimums of perhaps 162 pts (2.25”) wide for A-​Z and 120 pts (1 2/​3”) for a-​z as minimums.
    • I probably ought to add something on stroke thickness as well. However, given that stroke thickness varies within a font, and differs between horizontal and vertical strokes as well, the best way to cover this is not obvious. The desire would be to avoid extra light (or ultra bold) fonts. I wouldn’t feel too sad if it also outlawed typefaces such as Bodoni for legal documents, due to their very thin horizontals. Hmmm.

    Most common system fonts a reasonable person would think of using would mee these requirements, including Times/​Times New Roman, Arial, Helvetica, Courier/​Courier New, Verdana, Trebuchet, Georgia, Calibri, Consolas, Constantia, and Corbel.

    Of course, I’ve only addressed the font size part of the equation. There are many other components to legibility of text in print, such as line spacing, letter and word spacing, line length, and the color of the text and the paper.

    [EDITED various times to clarify minor points and improve wording. Most recently to correct that Zapfino was rescaled by 2.5x, not 4x, and replace a dead link.]

    ADDENDUM 16 August 2012:

    This stuff just doesn’t go away! A recent decision of the Michigan Supreme Court hinged on exactly this issue. The underlying subject matter was the hottest state political issue of recent years, an attempt to put in place a ballot measure that would in effect stop the ongoing removals of collective bargaining rights for folks doing business with cities. Here’s the Detroit Free Press about the case, and the actual court decision (including concurring and dissenting opinions).