August 11, 2004

Rotis revolution in my head

Rotis Factory

The skies are a desaturated blue-grey. The rain is pouring down in sheets on my windshield and the tail-lights of the cars in front look like they have come through some streaky Photoshop filter. Billie Holiday is competing with the staccato of the fat raindrops and I�m cruising along in the carapace of my car. It is an altogether trippy experience. The only thing that spoilt it was the sight of the green and white road signs, squinting at me with their slitty-eyed e�s.

There are 17,000 road signs in Singapore. Since August 2001, some 10,000 of them have been replaced by the Land Transport Authority (LTA) at a cost of S$200 to S$700 per sign. The result: the age old all-caps black-and-white Grotesque typeface were replaced by Rotis, an altogether more trendy type with a very unevolved ampersand.

The most legible font in the world?
Rotis was the brainchild of Otl Aicher, a famous German designer known for designing the symbols used for the first time at the Munich Olympics and the Traffic typeface for Munich transport services. He was a contemplative designer who proclaimed Rotis to be �the most legible font in the world.�

Those of you with your Rotis radars up would have noticed that this font has, in recent years, become the most overused typeface in these parts. From Semb Corp to the Land Transport Authority (LTA) to Kinokuniya, announcing the arrival of the Circle Line to being body text in annual reports, Rotis appears to be the font du jour for Singapore designers.

I hate the Rotis road signs with a venomous ferocity. For me, likeability takes second place to legibility when it comes to typography for way-finding signage, which makes Rotis Extra Bold road signs a greater travesty than if it had been chosen for purely aesthetic reasons.

Besides being too anemic, the typeface is too delicate for signage purposes. The great variation in stroke weight means that the thin bits of the letters disappear at a typical distance at which the signs are viewed. While its predecessor had a no-nonsense air about it, Rotis just comes across as idiosyncratic and oh-so-pre-millennium.

Perhaps the greater crime here is not the choice of typeface, but the extra wide kerning that the face is subject to. Which makes you wonder about the typographic decision-making behind the road sign facelift.

Who�s behind our road signs?
Hongkong-based Citigate Lloyd Northover was LTA�s branding and signage consultants and they seem the likeliest suspects. After all, they did use the Rotis Serif for LTA�s logotype.

However, Steve Crooks, Citigate�s Design Director for the LTA project, said that they were not responsible for the road signs.

According to an LTA spokesperson, the architectural conceptual sketches for the new street name signs were done �in-house�, while the detailed design and production of the signs were carried out by an aluminium structure specialist. Some engineers speculated that it might have been a �design-and-build project�, meaning the design was proposed by the manufacturer bidding for the road signage project.

It looks like our road signs were not �designed� in any meaningful sense of the word.

An expert opinion on Rotis road signs
Crooks, a wayfinding and signage design expert with clients like London�s Heathrow Express Airport, British Airports Authority, Hong Kong�s MTR Corporation (with Citigate Lloyd Northover) and Singapore�s LTA, wrote in a report to LTA in November 1998 that �� this particular typeface [Rotis Serif] was primarily designed for use at relatively small sizes in printed material and so is not necessarily the most appropriate font to be used at much larger sizes on wayfinding signage for a transport system�. The extra wide letter spacing of the lowercase letters is a fundamental typographic mistake. The effect is that one notices the individual letters and struggles to read the words.�

When asked whether he felt the new road signs were an improvement from the old ones, Crooks replied, �the new sign seems to replace one set of problems simply with another set of problems�.

Directional signs vs road signs
He also provided a primer on the differences between road signs and directional signs. �Road name signs do not have quite the same requirements as directional signs. Most road name signs are viewed from a stationery or very low-speed point-of-view. Road name signs are often positioned to align with the general direction of the road, rather than facing the direction of the traffic travelling along that road, the way direction signs are.

�A direction sign must be clear, unambiguous and concise to enable the drivers of vehicles to rapidly make decisions whilst they are driving at speed. This consideration directly affects the choice of typeface for such signs. Examples include Transport Helvetica in the UK, Interstate in the USA and DIN-Schrift in Germany. These are all functional san serif typefaces.�

Beyond typography
To be fair, more goes into a road sign than just type and kerning. An LTA spokesperson stated that the new signs have uses a prismatic reflective sheeting, which is six times brighter than the materials used for the previous generation of road signs. He added that �surveys involving both motorists and pedestrians using prototype signs of various materials, font types and sizes found that the new design was 20 percent more legible in the day and 100 percent more legible (that is, it can be seen at twice the distance) at night than the old street signs�.

I was skeptical. The statistics seem to raise more questions than they answered. How were the tests carried out? How many people were surveyed? What other font types and sizes were tested? How far ahead was Rotis in this beauty pageant? And finally, how were the percentages derived?

Reading two signs placed in front of you on purpose is quite different from reading it when driving at 60km/h. In the first instance, almost any typeface is legible and perhaps the enamoured tend to gravitate more towards what they like than what is easier to read.

Polling with a twist
I decided to carry out my own straw poll. And for a twist, I polled designers and non-designers separately.

Much to my surprise, my findings were consistent with LTA�s. Ten out of the 15 non-designers polled preferred the new signs to the old ones and 13 also found the new signs easier to read.

By contrast and coincidentally, 10 out of 15 designers preferred the old signs, with reasons ranging from �I hate Rotis� to �hard to read�. Needless to say, many of them took the opportunity to spout Frederic Goudy�s oft-(mis)quoted line, �Anyone who would letterspace lowercase letters would steal sheep�.

Are the results significant? I don�t know, but they�re sobering enough. Do people really find widely spaced words difficult to read? Or is it a case of designers repeating theories that they learnt in design school? Are our road signs a product of a non-designer for non-designers? Was Rotis a truly democratic choice, made without designer bias and snobbery? Or is it a glaring typographic faux pas on a national scale?

Nothing I learnt during this exercise has made the Rotis road signs any more legible. But learning that Rotis was Aicher�s idea of a typographic panacea, the earnest efforts of non-designers who unwittingly found themselves the architects of our typescape, coupled with Rotis� popularity with the masses, the road signs do suddenly seem a lot more lovable.

See also Road Signs.

Posted by Karen at August 11, 2004 05:39 PM in typography

Nice use of “carapace”!

Posted by: Cheshire at August 12, 2004 07:18 AM


Posted by: Karen at August 12, 2004 08:32 AM

I hate the green background. The only green things meant to be seen in public should be from plants.

Ur… did the LTA guy justify his statistics? I do not know about legibility but visibility is quantified be 2 tests. a) The luminance factor test that uses a photo-colorimeter to measure and b) the retroreflectivity test that measures the yeah… retroreflectivity of the material used.

Posted by: Todd at August 12, 2004 03:10 PM

He he, so funny to have an engineer on the blog. Mega big huge words!

Posted by: Karen at August 12, 2004 04:34 PM

maybe the signs looking ‘better’ now or more legible when you put design theories aside has something to do with them being in upper and lower case as opposed to all being in uppercase. i read somewhere that if something is all set in uppercase, it gets read 20% slower than it does when it’s in upper and lower case….

love your site!

Posted by: Janice at August 13, 2004 11:35 AM

Karen, great piece. Flowing, structured, well-written, even some field research! And the topic itself is very interesting.

I think the two things that can be taken away from this are that:
1) Designers need to pay attention to more than their navels. I mean, can a serious person really think that “I hate Rotis” is a good benchmark for gauging the functionality of the new signs?
2) Laymen can’t necessarily express what’s really happening to them when they “experience” something like new road signs. It could very well be that the aesthetic appeal of Rotis among laymen translates to a “yes” response to virtually any question asked of them, like “Is it legible?” or even something like “Would you use that font for your wedding invitation?”

And I agree that the loose spacing is the worst ingredient here. I wonder why that happened. Hey, since it was a �design-and-build� project, maybe the manufacturer was getting paid in proportion to the physical length of the signs?…


Posted by: Hrant at August 13, 2004 02:33 PM

great write up. the most interesting piece i have seen written on singapore in a long while… well done.
hmm…i wish i had bid on some of the old streetsigns though…

Posted by: jo at August 16, 2004 08:43 PM

karen some of pops saddest sounding songs might belongs to the band JOY DIVISION and finally someone in sunny spore digs Concert Blonde one of my fave band back in the 90’s

Posted by: seng at August 19, 2004 03:37 PM

Yew Weng, the Executive Director of Fuji Signcraft, just called to clarify that they are not responsible for the choice of the font � Rotis was specified in the project requirements. Additionally, he stated that they did not make any adjustments to the kerning as well.

He also candidly stated that they are paid in proportion to the length of the signs, although many other structural factors affect the cost as well.

Further, he said that they would never propose a design just to benefit their business but is unsuitable to the client’s requirements, and added that the role of the aluminium structure specialist was in the design of the structure rather than the typography.

I must add too that he was perfectly hospitable and knowledgeable when I visited his factory and even offered to inform me when the road signs are being manufactured again.

Posted by: Karen at August 23, 2004 06:12 PM

> they did not make any adjustments to the kerning

Then who did?
Surely he realizes it doesn’t come that way.

Also, a new thought: if the physical structures are built first and the letters are only applied later, and maybe a bunch of them were manufactured too wide due to some miscalculation*, is it possible that they kicked in the loose tracking subsequently to fill the space? That’s a long-shot, I guess. Anyway, somebody’s to blame for the loose tracking, period.

* Like maybe using all-caps, or a too-big size, or even a different font.


Posted by: Hrant at August 25, 2004 12:49 PM

> Surely he realizes it doesn�t come that way.
I don’t think he does. He claims that he used the font the way it came from LTA. Which to me, looks monospaced. Perhaps whatever drafting software they used, being not of the typographic variety, afforded no kerning control.

It may also be that the kerning was decided by the LTA and that they merely followed.

I have invited Yew Weng to join us in this discussion but he felt it was more appropriate that I make the comments on his behalf.

That forces me to use words such as “apparently” and “claims”, which I dislike probably as much as he does.

Posted by: Karen at August 25, 2004 01:13 PM

> monospaced

Interesting - that might be the key to this.
I think some measurements are in order.


Posted by: Hrant at August 25, 2004 11:22 PM

Happy birthday!

Posted by: srah at August 27, 2004 03:31 AM

Great article, Karen.

If the LTA were in California’s Gold Country, their options would be simpler.

These signs are from Angel’s Camp, CA., home of the Calavera County Fair & Jumping Frog Jubilee.

Posted by: Roderick at August 28, 2004 08:44 AM

Great links Roderick. Btw, has everyone seen rion’s signages via JFP.

Posted by: Karen at August 29, 2004 01:31 AM

I only found three wooden street signs when I was in Sonora, just south of Angel’s Camp. Seems that the rest of them were replaced by the more traditional, white on green signs. Save for the froggies. More signs sometime in the future.

Posted by: Roderick at August 29, 2004 05:07 AM

Well, I don’t think Rotis is the most appropriate choice for beautiful and very hot Singapore (last time I have been there was in 1987). Rotis is definetly to cold for such city. But Type deisgner with experience in signages typeface can’t count.

About legibility, the problem is not the odd forms of the Rotis but the fact that is a really condensed face and they have chosen the contrasted version who make it more narrow in its counters.

But I can agree that is difficult to compare before and after, as the differences are not just the typefaces but various things togethers. The result is enough appealing to be well received (despite Rotis), specially because it give probably some unity to street signs and its the most important.

Posted by: Jean F Porchez at August 31, 2004 11:21 PM

Oh wow! It’s the great JFP! How nice to have you here.

For the uninitiated, Jean is Porchez Typofonderie and has created typefaces for Le Monde newspapers (Le Monde) and the Paris Metro network (Parisine).

Posted by: Karen at September 9, 2004 06:55 PM

Enjoyed the article. The type for the signs appears to have been chosen by someone uneducated in the history and requirements of street signage - result seems to suffer as a result. The recent Eye magazine article alone might have prevented this unfortunate design.

In my opinion Rotis is unsuitable for a variety of reasons mostly mentioned, and the extra letter spacing just pushes it further into illegibility.

It would be nice to see some clearer photographs of the finished signs, front elevations for instance, and examples in situ.

btw I noticed a typo - ‘alumninium’

Posted by: Toby at September 27, 2004 04:18 AM

Are you referring to this article?

Thanks for pointing out the typo, I’ve changed it.

Posted by: Karen at September 27, 2004 09:22 AM

I’ve recently had the chance to see the signs myself and have photgraphed them and commented on them here. Please forgive my stylesheets formatting, just haven’t had the time to.


Posted by: lyrrad at April 16, 2005 11:28 PM

Hey! Thanks for the link! The title of the publication is designer and they use Univers Condensed for bodytext. :-/

Posted by: Karen at April 18, 2005 03:47 PM

Hey I’m a Singaporean myself and i find that the new Rotis road signs a lot better than the old ‘black on white’s. But you’re right the green colour is rather grotesque.

Posted by: Woody at December 20, 2005 11:19 PM

Hey Woody, just out of curiosity, are you a designer or non?

Posted by: Karen at December 21, 2005 09:52 PM

Hihi, stumbled onto your blog while surfing and stayed to read the entries.

Just wanted to add National Library Board adopted Rotis as well in 1998 when they came up with their corporate identity. And when WTO was in Singapore in 1996. What did NLB and WTO have in common? Su Yeang Design.

Posted by: Anne at March 30, 2006 09:27 PM
Post a comment
As an anti-spam measure, you'll have to preview your comment before posting.

Remember personal info?

Warning: include(): open_basedir restriction in effect. File(/home/httpd/vhosts/ is not within the allowed path(s): (/var/www/vhosts/ in /data/vhosts/ on line 796

Warning: include(/home/httpd/vhosts/ failed to open stream: Operation not permitted in /data/vhosts/ on line 796

Warning: include(): Failed opening '/home/httpd/vhosts/' for inclusion (include_path='.:/usr/local/php5619-cgi/pear') in /data/vhosts/ on line 796