Subpixel typefaces

I am fascinated by the phenomenon of how perception of type is affected by gestalt and physiological stuff. It is really evident with screen typography, and the smaller it gets, the more obvious it becomes. Here is an example: there is no way I could tell what the letters are in the large version, but at a small size (below right), I can make them out. (Open image to see the original).

It’s from this Typophile thread, which is worth reading not only for the discussion of the subject, but also for the vituperation as StoneCypher practically climbs into his computer to attack Miha and his claims.

Some typefaces I have been working on

These faces are works in progress, but I am becoming quite fond of them. They started as a client project, but I continue to develop them. I sometimes find myself working on them when I should really be doing something else.

How I approach thesis

As I am teaching graphic design core project at OCAD next fall, students ask me what my approach to the class (commonly known as thesis) is.

I see it as having three principal qualities

1. an opportunity to become an expert in some area of graphic design and related content

2. an occasion to produce a body of graphic design work representative of the abilities and interests of the student

3. the requirement to record the research, process and chronology of the core project in a well-written, well- designed, well-edited and well-executed significant document.


My interests
I am interested in the typography (I practice as a typography consultant to graphic design companies and book and magazine publishers, among others), in information design (which was the basis of my master’s thesis). So book design, periodical design, public communication (i.e., posters), interaction design and type in motion are principal areas of interest.

I like intelligence, and logical and ingenious solutions. I believe you should research and know the theory that are the related to your thesis. If you can’t explain what you are doing, I will think you don’t know what you are doing, though I realize I may be wrong; I value the ability to articulate. I also respect the ability to work hard.

I am not particularly into (for example) tattoos or graffiti, because these seem media of self expression (art) more than communication (design). If you are interested in a branding project, I will insist that you understand branding theory, how it applies to your project, and that a great deal of research will be what is essentially market research on the branded organization, its competitors, its users and other stakeholders. Visual identity itself is a relatively small part of the overall exercise of branding, and in my view, is seldom substantive enough to be a satisfying year-long project in itself. I think of graphic design as being a part of a successful society, not an avenue for self-expression. I believe that audiences are as complex and interesting as designers are, and that design is a bridge between people. I am interested in idea and concepts. I appreciate visual brilliance, but I insist on good execution.

Legibility typefaces

Sure it's the same old game: look, I made the x-height really high and so was able to give it big counters, so it's as legible at 8.5 point as Times is at 10. The only trouble is that because of those big lowercase letters it now takes up just as much horizontal space at 8.5 point as Times 10 point does.

Looking at Gerard Unger's 1993 legibility typeface Gulliver, which at 8.5 point “looks just as large as a type like Times (1932) in 10-point,” I am struck once again by the futility of making typefaces “more legible” by making the x-height proportionally larger.

Because sure, it is easier to identify letters if you make the x-height and counters larger. And sure, you can set it with tighter leading, and it won't crash, but with that big x-height, you want more leading...about the same proportion of vertical white space to x-height as Times... And you can even, as Unger did, reduce the size of serifs so tightly spaced letters don’t crash…but following that logic, you could set text set in sans serif with hairlines between the characters. And that won't work, because the correct space between letters is at least partly related to the size of the counter space.

Gulliver, like Ionic or Utopia, looks like a reasonably good typeface, though it might be nicer if the ascenders and descenders were a bit longer, and the only way to do that is reduce the x-height, thereby making the counters smaller.....

Oh well.

Here is a face that’s REALLY serious about big counters and legibility, Minuscule Deux, designed for setting at 2 point, by Thomas Huot-Marchand.

I like this carpark signage done by Alex Peemoeller and others in Melbourne. It works because the sightlines and eye height of viewers is controlled by their being in cars, and the signs are optimally legible when you are in the right position to make use of the information they present. The use of colour helps too. Not only is it effective signage, but it makes the usual drab, sense-depriving atmosphere of a carpark into something mentally stimulating. I would park my car here even if I if it were a long walk from my destination. Felice Varini does interesting work of this kind, though not being typographical, isn’t that interesting...

Ecofont

This typeface is ‘‘lightened’’ by putting holes in it with the aim of saving toner (or to ‘‘increase environmental awareness’’). http://www.ecofont.eu/








It’s a nice idea, but the typeface itself could be optimized to save even more toner, both by using less toner because of having a smaller form, and by being more condensed.






OK, not nicely done, but you see what I mean. Stanley Morison was interested in “efficient” typefaces in the 1920s, because it meant that newspapers could print more in a given amount of space, and after almost 1oo years, that interest is being renewed. What we need these days is a reversible printer, where you feed printed paper in and the toner disappears, because paper manufacture is an even bigger ecological problem....

Text type principles

1. Type is (mostly) not graphic design...
a. ... while the choice and setting of display type is more or less graphic design, display type is 90 percent graphic, text type is 90 semantic communication...
b. ... and 90 percent of type is text type...
c. ... but blocks of text type are design elements...
d. ... although text type, even in blocks, is not a design element when you are reading it, which brings us to item 2....

2. Text type is meant to be read. Being looked at is a functionally secondary, if chronologically primary, function. That is, people see and look at text type before they decide to read it or not, so....
a. ... type must look readable. If it doesn’t, people won't even try to read it.
b. ... type must be readable. Or people who have been led to read it will give up. What characters look like is not very important. How they work is. And how they work is dependent on many things, and the typeface that the words are set in is unlikely to be an important element in this.

3. Given that typeface is reasonably conventional in terms of form, weight, and proportion, the choice of typeface is less important than....
a. ... its relationship to the context and other design elements. This includes placement, proximity to other elements, and visual weight of the block of type.
b. ... its line length. That is, it must be neither too short nor too long
c. ... its leading, which will depend on the line length and x-height of the typeface.
All these qualities may be estimated using logic, but reading the text is the only way to test its success.

4. Display type is fundamentally different from text type. It often does not have to be particularly legible, its appearance is central to its success. As such...
a. ... its form and expression should be carefully considered.
b. ... typefaces that make good text faces are unlikely to be the best display faces. In former times, display faces of the same typeface family had different forms, proportions, and interletter relationships than text faces. Few typefaces these days have different cuts for display and text types, but it is worth seeking out those that do.


and some smaller ones...

5. For all most every single typographical purpose, x-height is more important than point size. The x-height not only influences correct leading more than point size (see 3c), it effectively determines how many characters fit in a given space. Proposing to use a face with a large x-height in order to save space by using a small point size is an exercise in futility.

6. Text set in serifs is most successfully set in typefaces classified as Transitional or Old Style. Typefaces which fit the Modern classification are usually to light and contrasting. Sans serif text may well be a better choice, as long as it doesn't stray too far from what a reader is expecting to see.

7. Really looking at the type and reading the type will be a big help. I'll probably come back and add more to this when the mood strikes me.