20+21 October 2003
Friends of St Bride Printing Library second annual conference | index of proceedings

Jean François Porchez: métro type

Parisine, a Parisian Type

There are two common approaches to typeface design. The first is to design a new typeface to your personal taste, following your own rules or restrictions, and distribute it either through a type distributor or directly. The second is to work for a client on a commission. The latter method offers more financial security and an opportunity to design a typeface following a narrow design brief, suggested by the client or governed by the client’s needs. Technical, historical or design considerations are all difficult to imagine if you design a typeface for your own use.


After designing typefaces for newspapers, it came as an interesting challenge to create a typeface for signage, for a medium other than paper. Unlike typefaces designed for small sizes, for poor quality paper and printing, which factors push the designer to reinforce certain parts of letterforms, typeface characters made for signage need to be cleaner and more minimal in their form. A purity of expression is needed.

Book typefaces from the Renaissance remain our archetype for most fonts created with paper as a final medium in mind. For signage, the purity of the Greek and Roman inscriptions seems historically suitable. Their open counters, proportions, and simplicity in the case of Greek capitals, need to be followed when designing typefaces for monumental inscriptions.

Original Métro entrance shelter'Ordinaire', RATP station lettering before the 1970s

Design brief

The way the Métro started its life strongly influenced signage in the stations. In the early days, a number of commercial companies ran the different Métro lines. This is one of the reasons that the inscriptions varied enormously, from enamel signage to big ceramic station nameplates. Sans serifs were mostly used for big signage, and on the carriages, letters were painted in a style appropriate to the carriage design. Early on, it was Art Nouveau forms. At the time, most of the transportation process was done manually by rail workers, from the sale of individual tickets, to the semi-automatic door closing. Later, the national rail network, the RATP, took over.

It was not before the sixties, however, that the overall signage question was taken into account by the RATP. After the Second World War, at the time of the industrial boom and automisation, the network was extended into the surburbs and signage became a key factor. The situation was similar for buses. Most of the direction signs on both sides of the buses were done by lettering artists, always in caps, in various condensed sans serifs. This method was used on the buses until the end of the seventies, when Helvetica was chosen as to replace these methods.

Adrian Frutiger's lettering for RATP, 1970s1970s métro signs

In the early seventies, the RATP set up a study group, including Swiss type designer Adrian Frutiger. He was asked to design a ‘special variation’ of his Univers typeface. The variant was introduced in 1973 to replace the twenty alphabets previously in use by the network. Later, Frutiger wrote: ‘It is the special charm of the Paris Métro that its applied aesthetics are not stamped with a uniform style. Forms of expression of the past hundred years, such as the beautiful Art Nouveau portals, are in many cases still present. This variety should be preserved as well as possible, as an enrichment of the scene. The joining together of typographical elements into a new harmonious order was a task requiring a certain degree of restraint so far as the creation of new forms was concerned.’

His recommendation was to stick to capitals to fit better with existing signs and with the historical roots of the Métro. The new alphabet was used only when the text needed to be updated or the station renovated. Soon after, around 1973 to 1975, Frutiger's Roissy, a preliminary version of the typeface called Frutiger, was created for the new Aeroport Charles de Gaulle. This time, without historical constraints, he used caps and lowercase instead of the all caps RATP alphabet. Frutiger wrote: ‘We rejected an elongated condensed face because of its loss of legibility. The similarity of the shapes of all letters, due to central vertical lengthening, has an unfavourable effect.’ I remember clearly from my childhood the strikingly contemporary effect of the new airport with its signage on yellow.

Frutiger's alphabet for Roissy airport

It was not until the early nineties that the RATP started to move towards using caps and lowercase signage concepts, which provide better word shapes and contrast. This formula was adopted to improve legibility. For future signage, which was intended to be applied to all the transportation systems, from the Métro to buses in the French capital, a typeface family was needed. The RATP president decided to select from one of the typeface families already in used by the RATP. These included the Adrian Frutiger all-caps face based on Univers, the RER, Albert Boton’s thin, rounded, all-caps face designed specifically for the new fast Métro in the late seventies, Gill Sans, used in recent years for corporate identity and official communication, and Neue Helvetica, chosen by designer Jean Widmer, which was used for bus signage system from 1994.

1990s sign lettered in Helvetica

Neue Helvetica was selected because of its general availability and compatibility with various computer programmes. This seeming advantage actually produced problems. Potential users mistakenly used Helvetica instead of Neue Helvetica. Because of the various widths, weights, and letterforms, the corporate guidelines have never been successfully implemented. Early on in the testing process, the people involved came from very different areas. (The signage, from basic stickers to illuminated boxes or classic enamels metal plates needed a range of production methods.) Non-typographers could not comprehend the consequence of choosing the wrong version.

Why a specific typeface?

The people in charge of the signage system quickly understood that Neue Helvetica did not work well, because of its width and standard narrow spacing. The station name ‘Champs-Elysées Clémenceau’ is obviously longer than the station name ‘Nation.’ Historically, the name plates had been sized according to the length of the name of station or the design style of the station itself. At stations where a lot of information had to be displayed (various connections and ways out), the name plate was smaller than in stations where less information was needed. After some tests in real conditions with strict rules, the final modular system permitted a modest level of adaptation of the name plate size: a real problem in France, which stubbornly resists standardisation.

Due to the problems described, the idea emerged for a specific RATP typeface with some Helvetica characteristics, and with the same legibility, but more economical in width. The rail network asked several of the larger typefonderies to provide quotations for a ‘unique’ Helvetica to belong to the RATP, to avoid multiple versions of their signage applications. They wanted to give away the typeface in one way or another to their suppliers. However, typeface designs are intellectual property. Neue Helvetica is used by the RATP, and many others. RATP could not buy and distribute a font licence, nor could a designer be asked to create a ‘RATP Helvetica’, we call that piracy.

Original lettering presented to RATP in June 1995

Most companies approached responded with quotations based on large corporate multi-site licences, without dealing with the design problems described. Instead of doing this, I asked for a meeting and brought some typeface research. I had made just two station name signs, directly on Illustrator, as we do when we design a logotype or a lettering job. I built some comparisons with Helvetica in reverse (signages are usually white on dark). We went out into a dark corridor, I put two A4 samples on the wall, asked to make the corridor as dark as possible, then started my explanation about the legibility of the open faces, horizontal openings, and various widths. The RATP team was surprised by the result and became very interested. We decided to sell the project in-house as a kind of Helvetica, which it is not at all, but possibly one for non-typographers! This earlier conclusion proved the enthusiasm of the team in charge of the signage. The typeface was adopted, and cut its supposed family links with Helvetica soon after the idea was accepted. Parisine, a name created by the RATP team, was born in the summer of 1995. The development of the new bold and a true italic took a couple of months and the two series were delivered to the RATP in January 1996.

In 1995, was involved in the organisation of the annual August conference of Rencontres Internationales de Lure, in the south of France. One of the speakers that I invited, American digital type designer Sumner Stone, visited me at Malakoff in Paris before the Rencontres. We discussed the Parisine project and he seemed pleased by its humanistic touch. Sumner has strong interests in inscriptions and calligraphy. He has been at the head of the type department of Adobe and managed the Trajan, Lithos, and Myriad projects, among others. We had an impassioned discussion about the importance of the capitals in inscriptions, their spacing, the difficulty of fitting them with lower case lettering. He told me about his experiments on this, which I found enriching, and which allowed me to confirm some of my own views. At that time, he was thinking about an all caps sans serif face for the Cecil H. Green Library, at Stanford University, which he went on to create as Basalt. (The reference to Parisine has been acknowledged by Sumner Stone in an exhibition of his work at the Ditchling Museum in Sussex, United Kingdom in 2000.)


In his Essay on typography, which compared two type displays, one in square narrow heavy letters, the other in purely ‘Gillesque’ caps, Eric Gill wrote: ‘A return to mere legibility seems desirable even if the effect be less striking. To this end it is necessary to study the principles of legibility; the characters which distinguish one letter from another, the proportion of light and dark in letters and spacing.’ Later on, he wrote: ‘Many engineers affect this style of letter, believing it to be devoid of that ‘art-nonsense’ on the absence of which they pride themselves.’ The style of letter to which he refers is close to what Helvetica represents, to my eyes. I have thought for a long time about this idea of contrast that helps legibility. Humanity is the key when you design typefaces, and this is particularly true for public signage, where no special marketing effect is nee, just a service to the public. Type designer Ladislas Mandel perceives typefaces as cultural items, writing: ‘Our first glance of any written work is always cultural. If the perceived forms are contained in our cultural references, we recognise them, we ‘own’ them like the reflection of our own image and we open large the doors to their intelligence.’ With Gill, Mandel and Frutiger share a concern for the act of perception as a key factor.

Parisine type sample


Parisine is intentionally more open to counteract the reduction of the areas of the counters due to condensing. Considering the characteristics of Helvetica and how they change because of the electronic distortion, I tried to make the forms softer and more round, human. Horizontal parts are slightly heavier, optically, to compensate for the verticality accentuated by the slight narrowness of Parisine. It is sensitive in lowercase lettering in particular. The specific form of each character has been carefully optimised to differentiate each from the others, uniformly, to maintain the overall colour of the typeface. The terminals of the characters align more with their ending than the vertical, as in Frutiger, or than the horizontal, as in Helvetica, as it gives more shape to the forms.

Letter shapes of Helvetica and Parisine compared

The counters of b, d, o, p and q are all different from one another. The g is more peculiar, in deference to Edward Johnston’s typeface for the London Underground. The f and t are wide, more than usual to constrain the verticality of the typeface. The R has a strong diagonal tail, to help distinguish it from the B. The capitals are designed to be heavier than today’s norm to affirm the words set in upper and lower case. The proportions of the capitals, based on Roman inscriptions, are large compared to the lower case letters. Figures are also designed to be wide and open, to improve legibility when used alone – they are used a lot for Métro and bus line references, used separately from plain text. This first version appeared only in two series, a bold and its italic, to cater for the needs of the signage system in 1996.

Helvetica slanted compared to Parisine Italic

Translations and tourist information are the main use of the italic. Its forms came from its use, the main characteristics being that it is more condensed, a bit more cursive, and lighter. The simple slanted form of italic was not selected because it lacked visual contrast to the roman. At the same time, the true italic form has a major problem with any a that can be mistaken for an o. Without completely resolving the problem, the ending stroke helps visibility. The e is more round, like in serifed italics. After some tests, the f was designed without is usual italic descender that we find in traditional serifed forms – too cursive for use at large sizes.

Metro sign set in Helvetica Metro sign set in Parisine

A preliminary conclusion

Early on, the new typeface was not widely used (mainly for budgetary reasons) but new Métro lines, such line14, did benefit from new signage set in Parisine. Soon after, the RATP signage team started to use Parisine for other purposes, such as maps. This happened because many kind of information are in between signage and maps, such as local line maps inside train carriages, so the idea of reproducing the Parisine effect in small paper maps came quickly. This proved that the main users had begun to appropriate the new typeface for alternative tasks. The need for a bigger family began to grow, and in 1999 the map team asked for a regular and italic in addition to the bold and bold italic.

Extended family

This new request was a continuation of the earlier 1996 dream: a big family for a broad range of communication and information material from the RATP. I proposed a six weights family with companion italics. I explained that more weights would soon be a necessity, such an extra bold and light, both important for maps. Other wide uses of a family would became an important element of global identity. I prepared two quotations: one more expensive than usual, for a regular and italic, and a second less expensive than usual for a six weights family. My objective was to give RATP a financial incentive to adopt the second solution.

I always considered the light and extra light of open humanistic sans serifs such Frutiger or Thesis too open and black weights too closed. I went back to Gill and thought about how he made his Kayo so different from the lighter weights. This was part of his search for contrast, equilibrium, legibility. I was, however, critical of the desire for rationalism embodied in typefaces like Univers, where twenty-one series are made in the same way for all graphic jobs, from signage to books. Ladislas Mandel expresses strong views on that question: ‘(typographers) confused modular architecture with type design, and they removed all handwriting references which might remind us of our cultural roots... they switched to the nudity of the sans serifs without appeal... and made up to twenty alternatives and more from a unique basic form, to pretend to answer all the needs of typography.’

Alternative e and a shapes

As the endings of the original bold were diagonal, an idea came into my mind. I created an extra light with closer endings, and a black with more open endings; the rest of the weights between these extreme weights and the original bold, came from the computer, thanks to typeface interpolation software. I was also try to prove, to myself at least, that it is not because of some ending, as is sometimes shown in typographic books, that we can recognise the style of a typeface – it is more an overall design question.

When I was creating the intermediate weights, many trials and mistakes helped me adjust the weights of each series. In 1999, I used Fontographer 4.1, for its speed, its ability to clean up early forms easily, its metrics and kerning assistance functions. But it became less useful when moving to tedious things, such as the need for different x and y interpolation percentages, global control of IDs, copyrights, encoding, hinting. Robofog and FontLab soon proved to be more up-to-date for such things at the time: Robofog because of scripting, FontLab because it is the only software able to build professional master fonts for Mac/OS and Windows. This last aspect was essential for me, as RATP uses different systems depending on the use and users, so a project can start in Adobe Illustrator on Mac/OS and finish in Freehand on Windows. The naming issues, because of cross platform compatibility, were resolved by three sub-family naming systems: Parisine Clair, Parisine, Parisine Sombre – each of them including four basic series, regular, italic, bold and bold italic.

Parisine Plus

Parisine Plus started its life as a game when I designed Parisine. The Plus version was a diversion for me, and offered an opportunity to play with extravagant forms that are not feasible in the standard version. In fact, in the early days, I felt very sceptical about the novelty of Parisine, and the Plus helped me to appropriate the forms. The italics of the Plus were very interesting to design because of the various features normally used only in serif fonts.

Parisine Plus characters

Today, the typeface family is not only used by the RATP for signage, maps, and communication, Parisine is available to the public in all of its versions. But, strange as it seems to me, the Parisine ‘standard’ became the most successful of all of my typeface families. I say strange because I always questioned myself about the novelty of it. Why is Parisine so appreciated? Perhaps because it is a synthesis between a Germanic Helvetica and the too Latin style of its creator?

Jean François Porchez teaches typography at Ensad, France and Reading University, England.