Wednesday, April 18, 2007

Case For an Economical Typeface (space-wise, that is)

Wonderful & illuminating visual aid.

... WELL, sort of "to the wall"... I wrote the following diatribe
for the Eddy-fication of various persons and I hope it had a
little impact.

........... , regarding a probable preference for Tiresias as a typeface on signage, I'd like to interject some notes on why I recommended against it and in favour of FF Transit, . ....

Tiresias and APHont are indeed highly legible typefaces, however, the features that make them so also make them quite wide, due to open, generous counters, wide cross strokes and corresponding open letter spacing. In many circumstances, within the context of a service such as __ _______, available horizontal space is limited more often than vertical space. This means that we will frequently lose any advantages that Tiresias and APhont have to offer by being forced to reduce letter height to fit messages into limited spaces. In the (admittedly limited*) testing I conducted in-house, I found that we'd have to shrink Tiresias and APHont by a little over 15% to occupy the same horizontal space as Transit. With some long place names, limited spaces for sign installations - And the inclusion of a second language which tends occupy more space, this is a fact that can't be ignored.

While all legibility testing is subjective, it was clear to me that when this was taken into consideration, Transit was the most legible typeface I looked at. Also, (as of the last time I looked) Tiresias Signfont is available only in one weight with no italic. When I asked RNIB's John Gill if they were planning to add additional weights and italic variants to the mix he said he'd look into the cost. Shortly after that Sylvie Perera (Human Factors Scientist. RNIB Scientific Research Unit) emailed me to say, "We will be producing the italic and bold versions of Signfont for you as soon as possible." . [Oddly, she made it sound as if she was under the impression I'd placed an order - of course she didn't, but I still had to get her to clarify what she meant.] .... To date I have seen no evidence of italic or other variants When I was making my inquiries about APHont, Elaine Kitchel told me that a full suite was with the fontographer and was "due any day." and it is now available.

... Tiresias and APHont were designed by scientists - Dr. John Gill (RNIB) and Dr. Elaine Kitchel (American Printing House For The Blind) respectively. Or, rather teams lead by them. The characters they designed are individually well constructed so that they read well under poor visibility conditions, and they have been thoroughly tested with subjects who have varying degrees of low vision. But, in my opinion they do not "flow" as well as Transit. An experienced reader scans strings of text rather than individual letters and relies on word-shape recognition and context to aid reading, only pausing to distinguish individual letters when a word-shape is unfamiliar or something irregular interrupts eye movement.

The rhythm of Transit is more consistent than Tiresias or APHont. (rhythm being the balance between positive and negative space - which is important to legibility, as it aids in word shape recognition. There should not be uneven densities of colour to distract the movement of the eye.) I don't think those two fonts stand up as well to that test as Transit. I may appear to be a personal prejudice as a graphic designer speaking, but I do think that an experienced typographic designer will ultimately be better at designing readable typefaces.

Perhaps when comparing typefaces for relative legibility, there may be a tendency to also compare the credentials of the designers… Erik Spiekermann, the designer of Transit isn’t a scientist. He is, however, one of the world’s foremost authorities on the design of type. He is well enough qualified to design a highly legible typeface. He has designed many typefaces and many wayfinding systems including for public transit systems, airports etc. He was a co-founder of the first distributor of electronic type and founder of MetaDesign. He has authored several highly regarded books on design and typography - I believe his understanding of legibility is at the very least, equal to Dr. Gill and includes an understanding of how type works in a signage system. Transit was initially designed for Düsseldorf Airport and is now used on the signage of many German public transit systems. It is also the typeface of choice for some projects in Britain to improver cities’ signage systems. (Example;

We need to be mindful of economy of space with our signage, as I indicated above and I thin we can do this without sacrificing legibility. Condensing letters to achieve economy of space without sacrificing legibility has been a goal of type designers since the invention of movable type and long before then, among calligraphers, mindful of the cost of vellum. The condensation of letter shapes in typefaces such as Transit is differential, with the greatest amount being applied to letters such as f, r, s, t and E, F, H, L, S, T, where it is easy to simply make the letters narrower without affecting distinguishing feature points. Less compression is applied to letters with open counters c, h, m, n, etc. and even less to letters with closed counters (insides of a, d, e, o, O, Q, etc.).

This is not written to specifically defend the use of Transit as our signage typeface - though I think it serves very well – It is an attempt to defend the use of a typeface that uses space economically and has superior typographic characteristics. There are many typefaces that fit the criteria of economical use of space and superior legibility. These would include Arrival by Keith Chi-hang Tam of Vancouver and M.O.L., by Gerhard Unger for the Amsterdam Metro.

In conclusion, the inconclusive results of the current CNIB study may be accounted for in words attributed to Eric Gill, (
1882 to 1940, known for the classic typefaces, Gills Sans and Perpetua, & no relation to Dr. John Gill): “Legibility, in practice, amounts simply to what one is accustomed to”. Although a little flippant, every joke has a kernel of truth and the truth in this one has been confirmed by research and common sense: Familiar shapes are more legible than unfamiliar ones. Differences in performance between any group of purposely designed "legible fonts" can be vanishingly small and more dependant on what each individual observer is used to reading than on subtle variations in geometry. - Except where horizontal space is limited. Then, a more economical typeface is likely to be better

* In the course of collecting data for the typefaces I examined for the ----- Typeface Review, I contacted all of the vendors to see what additional supporting information I might be able to acquire, and in every case I was directed to the designers, who were all very helpful. I did discuss my method of testing typefaces using uniformly blurred images and a questionnaire with both Elaine Kitchel (American Printing House for the Blind) and Dr. John Gill. They were both very helpful and generous with their time, and both replied that while not a rigorous test by scientific standards the principle was sound enough to give a good evaluation of relative legibility for what I was trying t
o determine. Erik Speikermann’s comments were equally helpful, and since he has a stake in the sales of his typefaces, I’d like to point out he was unbiased. In addition to explaining a little of the history behind the development of Transit, he suggested several typefaces by other designers he felt might also work.


Anonymous said...

Hello Russell,

Elaine Kitchel here. I enjoyed reading your comments about Transit,APHont, and Tiresias. You defended Transit well, and for legibility, Transit will be great for an varied audience.

APHont has received some severe criticism from certain fontographers, but in every case, they are looking at the font from the point of view of a person with typical vision. I'd like to point out that APHont was designed specifically for documents which are to be read by persons with low vision. It was designed neither for signage or for use by persons with typical vision.

Research conducted by Dr. Gordon Legge, Dr. Stephen Mansfield, and Dr. Aries Arditi, all showed that persons with low vision read more efficiently when:
1. There is a generous t-height
2. There is a generous x-height
3. There is generous kerning
4. There is 1.25 line leading
5. Letters are rounder with generous space between the horizontal elements of letters.

To get a font that met those design needs, we had to make one ourselves. Our low vision readers love APHont and if the number of downloads is any indication, the audience of APHont is growing daily.

I appreciate the work you are doing and the kind way you spoke of me and my research in your comments here. IF Transit works for the greatest number of people as a signage font, and is a clean font with few serifs, then I'd be one to encourage you to use it.

Best wishes,

Elaine Kitchel

Joe said...

Elaine, please stop referring to type designers as “fontographers.” Are architects “autocadders”?

I have read several of the papers by the authors you cite, all of whom are recapitulating the John Gill Paradox of knowing nothing about type but a lot about vision and assuming the latter supersedes the former.

You probably mean generous tracking, as kerning always brings characters closer and only applies to character pairs. Linespacing needs will vary by measure for all readers, including low-vision ones, meaning that 1.25em linespacing isn’t going to be a viable number except by coincidence.

Russell said...

Joe, we all make mistakes with the jargon. (Only the really good architects are AutoCADders. The rest are just CADders. CADs, for short)