Tuesday, November 27, 2007

And then there is this crazymaking type thing

I have not been able to find any documented or demonstrable legitimacy behind Tiresias as a Signage typeface. The "scientific" claims said to support Tiresias Signfont are flimsy at best and not even relevant to a signage typeface. All the evidence that is referred to on-line is from testing of Tiresias Screenfont, which is intended exclusively for use in captioning on television screens. This is a unique application of type because of the low resolution. I can find no reference to any testing by RNIB of actual signs in any form. It appears that assumptions were made from testing of Screenfont and extrapolated to fit someone's idea of a typeface for signage.

A review of the backgrounds of the team members involved in the development this typeface reveals that no one with any substantial experience with signage or type design has worked on the font. (Design of signage is not listed on the Laker-Sharville web site as a one of their capabilities).
The research done by the [----] is already just about equal in scope to published results from what has been done by the RNIB and the Tiresias team, and it doesn't show Tiresias to be any kind of Magic Bullet for improved legibility compared to other comparable typefaces.

A careful reading of the documentation behind Tiresias leads me to conclude that Joe Clark is absolutely correct in his assessment of the typeface. His commentary is entirely about Tiresias Screenfont, however, as it would appear that the Tiresias team has done no more than make assumptions about what a signage typeface should look like, based on conclusions made about Screenfont. It is therefore as relevant as any other commentary available about Tiresias Signfont, either pro or con.

The following is my own summation of the development, support and validity of Tiresias Signfont from reading the relevant reports and documentation of this typeface:
  • The Tiresias project was conceived as a project to improve captioning on TV screens.
  • Making a signage font was an after-thought.
  • A 'professional' typeface designer was brought into the project late in the process - Essentially to clean up what had been already created and to convert the letter glyphs into font files. This actual typographic work on Tiresias was done by Chris Sharville.
  • Tiresias is evidently his first attempt at a typeface. There are no others credited to him on the market, and Tiresias is the only typeface listed in his portfolio.
  • Creating, or even editing an existing font, is a very complex task, and it is a lot to expect of a designer to master all of the subtleties required in a first attempt. A review of Mr. Sharville's on-line portfolio will reveal that he is a competent designer, but not exceptional.
  • The typeface (Screenfont) was tested on 35 visually impaired subjects and 48 hearing impaired subjects, (plus a small number of guides and sign language interpreters) using simulations of TV captions instead of actual captions.
  • As with the ---- tests to date, subjects who participated in Tiresias font trials did not represent an average demographic group.
  • Tiresias Screenfont was only tested against two other fonts; Times New Roman and Standard Alpha-mosaic. Times New Roman, with serifs and a lower x-height, will not reproduce very well as small text on a TV screen and Standard Alpha-mosaic is a dot matrix font (they used 7X5 dots and 9X10 dots). It would be difficult to find a sans-serif typeface that was not better than these two for captions on a television screen. That they would choose to compare these typefaces for any reason seems to indicate a remarkable ignorance of Typography.
  • All typefaces were set at 14 points. Point sizes are nominal and measure the the distance from the highest ascender to the lowest descended plus some empty space. Times New Roman has a lower x-height and is visibly smaller than Tiresias set at the same point size (by about 10%).
  • Tiresias Signfont was designed based on assumptions from the research on Tiresias Screenfont.
  • The report states in it's conclusions; "It is freely admitted that this "testing" is far from ideal and could even be described as anecdotal."
  • All references to scientific research in support of Tiresias Signfont point to material about Tiresias Screenfont and "Access Prohibited?", a British website about guidelines for barrier-free environments, which is as vague as similar North American documents, in that all it says about typefaces is that larger text is more legible.
  • There is absolutely nothing in the literature about the development of Tiresias that describes how the design and the geometry of character shapes and the metrics of letter spacing and kerning has been treated differently than in other typefaces that hasn't been axiomatic to the the world of type design for at least a century.
  • Tiresias is a bold single-font typeface with no light or medium weights and no italic. Having only one weight with no italic removes the option of ever using emphasis to call attention to important information while adhering to the graphic standards of a signage system. ( --- says that Signfont now has italics and a lighter weight, however, we should believe that when we see it. There is no mention of this on-line and they are not for sale anywhere.)
  • For a modern typeface intended for signage that's legible to persons with poor eyesight, it is odd and disappointing that there has been no consideration given to the effects of halation for back lit and retro-reflective signs or the 'dazzle' effect of white type on a black background.
  • Browsing through the specimen books of any major type foundry will turn up numerous typefaces that have all of the characteristics Tiresias infers are unique to their typeface in aiding legibility, with the exception that these other typefaces will have multiple weights and italics - and look better. If the initiators of Tiresias Screenfont project had known that, would they have bothered?
The features that make Tiresias legible are not new or complex. In fact, they are fairly commonplace: Make open shapes open, use large x-height, use open tracking (letter spacing) etc., etc. In the 550 year history of type design, I'm sure the list of people who have made the creation of legible typefaces a big part of their life's work, is a long one. It's just good business in the printing business.

One has to ask, how likely is it that a team of scientists with no knowledge of design or typography, and a novice type designer are going to come up with something that hasn't been thought of already in the last 550 years? In modern times, people have become more aware of legibility of signage and screen typefaces, however, the exact same principles apply, except, due to distances and low resolution, simpler shapes (sans serif) and increased tracking are helpful. The key factor is size and you don't need science to tell us that.


As noted elsewhere, I am not defending FF Transit specifically. It was chosen out of a small, predetermined field and it turned out to be a significant improvement over Helvetica. There are other a typefaces available that meet all the criteria for signage for public transit (legibility, economical use of lateral space, versatility, etc.) FF Info would be a good alternative choice as would Arrival, by Keith Tam or M.O.L. , by Gerard Unger, all specifically designed for wayfinding signage in a public transit environment. Our intention from the start, has been to make our signs more legible and therefore more accessible. I think the differences in legibility between one good quality font and the next will be very small and not worth the effort and expense of measuring. It all comes down to some simple geometry and the metrics of letter spacing, x-heights and etc. that make up the basic information that reaches the eye.