The rebirth of Ambroise typeface & his tight connection to Didot dynasty

THE revision of Ambroise Pro has taken 15 years. This new version of Ambroise Pro feature fully new italics, new weights and many other new glyphs! We look forward to seeing the first projects with this Didot! It seemed appropriate to explain the process to young generations of type designers and to introduce to type users a little glimpse of Typofonderie everyday world. (Francophones, lisez la version française.)

Ambroise 2016 versus original Ambroise from 2001: from 14 to 28 styles, from 250 glyphs to 900 glyphs by font.

The Ambroise project started in the summer of 2000. My first goal was to complete my initial training. Why? I quickly felt the limits in my range of style. In the years 1980–90 the French tradition of learning type design was based on the use of a calligraphy nib pen and the ensuing humanist typefaces from Jenson to Garamond, etc. It became logically possible to draw a multitude of styles: serif variations, contrasts, from slab serif to sans serif, etc. Nevertheless they remained influenced by the humanist footprint.

Learning to unlearn

When I was at Dragon Rouge as Type Director from 1991 to 1994 (it was during this period that I taught myself to work on a computer, using Ikarus and Fontographer), part of my job was to draw wordmarks, or lettering pieces for various projects for the agency. The richness of these projects quickly pushed me to overcome the strictly “noble” typographical models. Nevertheless I worked my shapes by leaving traces produced by many and varying tools, as the results are faster to achieve: flat or bevelled felt pens, brushes, felt brushes, pencils sharpened in different ways, etc.

Lettering artworks drawn by hand, early 90’s.

In the years that followed — as I finalized Le Monde and other projects such Anisette, Bienvenue, Costa — I quickly felt a missing piece in my shape’s vocabulary: Modern Face and English Roundhand. I looked for a historical model that would allow me to learn a new approach: Gras Vibert quickly became the ideal thing, this ultra bold Didot being explored less than the Didot and Bodoni already on the market in the 90’s (ITC Bodoni, Didot Linotype, HTF Didot). Drawing Ambroise was also a way of addressing a different process: using historical sources (scarce even in 2000–2001) rather than using writing or starting an original design without direct formal reference. It was only the following year that I started a true revival: Sabon Next.

The history of Didots is complex, and research by André Jammes, followed by Michel Wlassikoff, Sébastien Morlighem or Jacques André, whilst fascinating, would be to verbose to discuss fully here. However to better understand the history of Ambroise, at least a few references to the Didot’s history provides some clarity. Where to start when exploring the Didot story? From Romain du Roi or Fournier, as strategic roots? For the sake of brevity, we will start with: Gillé.

From Gillé, then Laurent and Balzac to Deberny and Peignot

Joseph(-Gaspard?) Gillé (17…-1827) was a contemporary of the Didot family who also engraved punches of this style, gathered in a beautiful specimen in 1808. What family would have influenced the other? It’s hard to say. It should nevertheless be understood that the families of French engravers and printers are intertwined for generations. Proof of this is the address of Joseph Gaspard Gillé, identical to that of Estienne and Garamond: Rue Saint-Jean de Beauvais, Paris 5th. Honoré de Balzac, using borrowed funds from his father and his mistress Madame de Berny, went on to purchase the remains of the Gillé legacy in 1827. This same material appeared much later in the Deberny and Peignot collections. Jonathan Hoefler’s HTF Didot designed for Fabien Baron in 1992 is based on the work from Gillé specimens like the one by Commercial Type published in 2016 under the name Le Jeune.

Page from Fonderie Gillé type specimen, 1808. Bibliothèque nationale de France.

Didot dynasty

The Didot’s styles are many. They took root in the Romain du Roi and the Fournier works, Luce, etc. François Didot (1689–1757) was the first of the dynasty of Didot, a native of Lorraine. His son, François-Ambroise Didot dit l’ainé (1730–1804), inheriting the foundry from his father in 1757, introduced several technical innovations: the presse à un coup, or ‘press at a glance’ that increases the efficiency of printers, a typographic point which regulates the typeface sizes (an improvement from the Fournier works), vellum paper in Annonay produced by Johannot (1780). The stereotyping method, improved by Didots, allowed them to refine the print quality of their high contrast typefaces.

Partial Didot family tree.

Vafflard Types

Circa 1757, François-Ambroise Didot will cast (Pierre-)Louis Vafflard (17…–18…) Didot typefaces, a style optimised by his son Firmin Didot around 1782. The beautiful Vafflard typefaces “are a very clever and extremely elegant compromise inherited forms of the Renaissance and the new taste professed by the eighteenth century, at the same time they do anticipate the rigor of the nineteenth” wrote Jeanne Veyrin Forrer Les premiers caractères de François-Ambroise Didot in 1960.

Set in Louis Vafflard types, printed by François-Ambroise Didot, 1786.
Small note set in Louis Vafflard types, visible in Virgil, itself set in Didot Millimetrique, Bibliothèque nationale de France.

Types of Firmin Didot and Pierre Didot

The eldest son of François-Ambroise, Pierre Didot l’ainé (1761–1853) succeeded him in 1789 by installing the presses at the Louvre. The second son of François-Ambroise, Firmin Didot (1764–1836) took over his father’s foundry in 1789 and built various activities connected to the art of printing until 1829 when he started a political career. Firmin Didot engraved a Didot Millimetrique that would be used for printing the Sacre de S.M. L’Empereur Napoléon in 1815 and preserved since in the Imprimerie Nationale. Virgil, the first in-folio by Pierre Didot the elder in 1798 (additionally Horace, 1799 and Racine, 1801 are considered masterpieces of typography) was set in Firmin Didot’s typefaces who inaugurated this characteristic style. Henriade, printed in 1819 by Firmin Didot, is also set in his own typefaces (this book was recommended by Ladislas Mandel to Adrian Frutiger as a basis for his Linotype Didot published in 1989.)

Page from Sacre de S.M. L’Empereur Napoléon set in Didot Millimetrique.
Smoke proofs made from Didot punches along with design comments.
Virgil title page set in Didot Millimetrique, Bibliothèque nationale de France.
A special dedication handwritten by Pierre Didot l’ainé.

Types of Jules Didot and Michel Vibert

Later, Pierre Didot l’ainé described a new style in the introduction of Petit Carême by Jean-Baptiste Massillon, Paris, 1812, the first of a new collection set in typefaces by Michel Vibert (1763–1813): “I was perfectly assisted by the indefatigable patience of the talented M.Vibert”; or perhaps by Jules Didot (1794–1871) too, because this book is dedicated to him:

The principles which I have subjected my new types. All the minuscules being so obviously formed from capital letters, several even have them any difference; these are generally the o, the s, the v, the x, the z, I had to try to bring the best as I could from their true origin. The g however appeared with me so altered that there could suspect no analogy to its primitive type, and its general shape is every where so bizarre it has a full irregularly rounded instead of ending each horizontal hairlines letter and marking the line, I ventured to correct this type which also suffered in all countries and in all the various arbitrary time changes, but never any change feature. I have therefore almost totally varied contours, without for that appearance in different parish. The form I gave him recall at least a few strokes of the capital, and thin strokes marking the line found there is their natural position; ensure that this letter became more regular, seems at the same time agree with all the others.
 I am still determined to make a slight change to y, including the following grounds. It seemed inappropriate to give this letter, which until now has always been reproduced in the same way, the slope and inclination of emphasis when the Roman character everything is perpendicular when its capital above all shows have to fix her to lead. So I thought allowed to draw it perpendicular, admitting only to its base a slight inflection to the contour and the final ear, almost like the head of f.
 The usual default there is mainly noticed when it arrives this letter is placed between two long letters, for example in the word Egypt.

Jules Didot later published a Spécimen de la nouvelle fonderie de Jules Didot l’ainé in 1842 (reissued in 2002 by André Jammes and Marc Kopylov at éditions des Cendres) in which he presents diverse typefaces, some dedicated to display settings, such as all capitals variants featuring rounded endings like on A, N… The Lowercase g and y are significantly different to those of Vibert. Such typefaces are probably the source of the Didot published by Swiss Typefaces as New Paris.

Spécimen de la nouvelle fonderie de Jules Didot l’ainé title page.
Jules Didot typeface featuring unique g, along with rounded A, v, etc.

The Spécimen des nouveaux caractères published by Pierre Didot l’ainé and his son Jules Didot in 1819, a punchcutter (using the Didot point to name different sizes of the same typeface) honored their “new style” cuts of Didot typefaces and those of Michel Vibert (1763–1813). The Didot Edler created by François Rappo is based on some of the typefaces featured in this specimen.

Michel Vibert’s typeface featuring unique g, y in a different style from the ones by Jules Didot.
Michel Vibert smoke proofs show the specific g and y as used in Ambroise.
Jean-Michel Vibert fils’ specimen of “Gras Vibert”, a source for Ambroise.
This specimen of Cyrillic types published in St. Petersburg in 1841 display identical style as “Gras Vibert” above.

Son of the former, Jean-Michel Vibert (1797–1862) who might have been taught by Pierre Didot or another punchcutter, continues the style of Firmin and his father, Michel Vibert. He is probably the author of “Gras Vibert” cut around 1820–1830 (named thusly a century later by Charles Peignot or perhaps Maximilien Vox at Deberny & Peignot foundry). This unique treatment of g and y employed by Vibert in his Didots seems to follow the “new style” described by Pierre Didot in 1812. Vibert added it to another specific form, k. Therefore, the existence of Cyrillic variants of this Gras Vibert produced and visible in a specimen published in St. Petersburg in 1841 (visible at the Butler Library, Columbia University) pushes one to investigate a little more into why this style of K k is characteristic of Cyrillic… Stay tuned.


Sources in the idea of ​​a family built in Multiple Master

In 2000, my limited resources focused on Vibert were limited to the Deberny and Peignot prewar and postwar specimens, and some books, such as Tresory of Alphabets by Jan Tschichold. Starting from scans from these various sources, the first step was to draw the existing forms, letter by letter. Then to find a particular finish, observing different existing digital Didones that I describe as colder and drier than what I had seen printed… My goal was to allow myself to be guided by proportions, accidents to embrace different things from my pre-2000 past experiences: to draw as accurately as possible, while streamlining the better weights distribution, serifs, rounded endings. Once this first step was complete, I quickly planned an extended family, including weights and widths, to play a little more with interpolation that I practiced since 1993 (extension of Angie and Apolline, then Anisette, etc.).

Designing an extended family requires one to draw extrems suited to potential intermediate styles. So I dramatized the drawing of my sources, accentuating the weight of Gras Vibert, in order to create a stronger style. Then I tested simple condensed variants on Fontographer to quickly see how a Condensed sub-family would work. Type design is arduous and takes time, so it was important to imagine early strategies to maximise my efficiency, avoiding repeating things several times due to conceptual changes: Work on a reduced glyph set, potential extensions planned in advance, and then in a second step: producing the glyphs needed for extremes (or masters). From these stabilized masters emerged a first round of tests to verify the initial strategy. In 2001, the three Ambroise widths have been produced in two steps: interpolation between width extremes in black and thin variants (Ambroise and Ambroise François) to obtain the intermediate widths masters (Ambroise Firmin) used to interpolate intermediate weights. Such was the process of 2001, constrained by the technological limitations of the software available at the time. Ambroise is based strictly on the Gras Vibert, a historical model and examples of condensed variations from French foundries catalogs: the light weights of Ambroise are much narrower than usual Didot text faces. In conclusion, the Ambroise family became more dedicated to display usage.

Names of families and subfamilies were chosen in reference to the Didot dynasty, using Firmin for intermediate family to position this subfamily alphabetically before Ambroise François. Pragmatic choices unrelated to the true history of the Didot dynasty!

Technological limits by 2000

The tools in use around 2000 and 2001 were both similar to what we have today and very different, due to how limited they were compared to now. The main principles of Fontographer’s interface remain in force today in FontLab, Robofont, and partially in Glyphs; but everything else is highly evolved.

The process was quite laborious: I used Fontographer 4 for drawing, then Robofog (based on Fontographer 3.5) and finally FontLab for production. The screen quality wasn’t the same as today: no smooth display and limited to 200% zoom in Fontographer. These limitations were challenging for the rounded details of Ambroise: there was no way to zoom in enough to see what I was doing. So, I had to count the movements of my BCPs (Béziers control points) approximately to control my drawings. Otherwise I would enlarge 400% in order to draw a detail, and then reduce. This technique had its limits in the sense that the designs were changed by the grid of 1,000 fixed units. Fontographer also accepting decimals: the good being the enemy of evil… A strange Fontographer process which did not help because it changed the results in context.

Original outlines of Ambroise E capital, 2001. Screenshot of my own Apple Powerbook wallstreet displaying Fontographer running on MacOS 9 under MacOs X, still working in 2016!

Designing a family in three widths and few weights requires four Masters. The tools available to produce a Multiple Master weren’t as flexible as today (thank you Superpolator!), the process nevertheless remained quite magical. It was at that moment that I had the idea of ​​drawing specifically designed vignettes for each master, to multiply the intermediate variations to offer to users a palette of infinite effects. These vignettes — as well as features like variants and ligatures — were not as easily accessible as today via OpenType features. Each style was therefore delivered in 2 fonts: the basic version (maximum 256 glyphs) and the version including vignettes, alternates and ligatures, mapped to ‘incorrect’ locations on the keyboard, such as the ct ligature placed at the c position on the keyboard…

The original family was started in the summer of 2000, presented to my closest colleagues Jeremy Tankard and John Hudson during the ATypI Leipzig in September for their views on these ideas; then revising during the winter in order to publish Ambroise in spring 2001.

First results

Since its launch, the success was immediate and has continued so. During these years, the technological trends allowed me to observe defects impossible to solve at the time, and of course the advent of OpenType, Unicode required a total redesign of the project.

Meanwhile, I continued to learn new things about Didot randomly thanks to ongoing projects and meetings. In 2003, I rediscovered the Gras Vibert punches during a visit to the Cabinet des poinçons at Imprimerie nationale. I also found other examples of this Gras Vibert and other Didot at the Bibliothèque nationale de France, Butler Library in New York and other conservation areas, or during visits to the library Paul Jammes. Putting in perspective the different typefaces designed by Didot dynasty takes time! Every design decision that results should be taken with hindsight, especially considering my Ambroise is not a strict revival: I have not had this historical claim, which seemed too quick and simplistic an idea when approaching a contemporary typeface.

Original punches of “Gras Vibert” discovered in 2003. Cabinet des Poinçons de l’Imprimerie Nationale.

Renaissance

Ten years after the publication of Ambroise, a small project for the TDC 56 pushed me to imagine italics, drawn very quickly, for Ambroise Black. This project called for the design of very few glyphs. It was much later that other projects pushed me to develop the italics. Linked to the manner in which I’d created Ambroise and his narrow style, adapted to display uses, the emphasis of the design had to take into consideration the context of contemporary use of Didots. In fact, rather than seeking historical Pierre Didot, Jules Didot or Vibert italics and trying to fit the narrow style of Ambroise, I preferred to start from scratch, designing the most suitable form: Condensed, reduced slope (in total contradiction to the Didots), but keeping some specific details from Didot such the single g, cursive y, straight f descender.

In parallel, the designs of Ambroise required a complete overhaul thanks to the progression of technology. It had finally become possible to finalise certain details correctly (see above). It was long and hard: what degree of retouching to achieve? To reconstruct or gently clean? Obviously the characteristic shapes of g, k, y marking the Ambroise had to be accompanied by more conventional alternatives. This was achieved with the introduction of variants g, k, more acceptable in certain contexts and uses.

Production of instancess for Ambroise Pro family using Superpolator, here building the Vignette sets.
Production proofs of Ambroise Pro, used to test matching weights through the three widths of Ambroise Pro.

Other extensions of the Ambroise family followed, such as small caps — not so simple to achieve as you might think, due to the small x-height and forms of figures (also revised). How tall should the small caps be? Several tests were necessary. Final small capitals being derived from an extrapolation via Superpolator between the original capitals and the first taller small caps. In 2014, I finally finished the standard width family. A few months later, Joachim Vu took over the project, carrying these optimisations out to condensed versions. While the general direction of the optimisation of the narrow versions seemed it would be easier, being based on the standard widths versions, two more years were needed — with the help of other members of the team Typofonderie — as soon as the projects ZeCraft allowed. Thank you to all the designers who have worked on this project over the years, in addition to Joachim Vu working on since 2014: Roxane Gataud, Sophie Huffschmitt, Malte Bentzen and Benjamin Blaess for the final stage of production.

In the spring of 2016, the family finally came to life. It was the beginning of a new stage, related to the use of Superpolator allowing us to optimise the consistency of weights throughout the three widths. Very quickly, the need to produce extrapolations to affirm extremes required us to redraw a new set of masters. The delicacy of the rounded endings was at every moment subject to unlikely mathematical interpretations. The last step started in the summer of 2016: Using variants and widths imagined in 2001 to increase the possibilities of vignettes was deployed to new vignettes and borders.

Jean François Porchez

We hope you will make good use of this new version of Ambroise Pro featuring fully new italics, new weights and many other new glyphs! We look forward to seeing the first projects with this Didot!

Ambroise: Availability of typeface family

The Ambroise fonts are available in our PRO (exclusive), STD, Epub versions. Download the Ambroise specimen in pdf format for full details of these Advanced typography functions.

→ From € 45 to € 55 for any Ambroise styles
 → From € 119 to € 146 for the Ambroise Family of 4 fonts
 → From € 712 to € 872 for the Ambroise Full Family of 28 fonts


Ambroise in use

As a tradition, at Typofonderie, we test a typeface family on various applications, using existing designs. It’s the final part of any project and its great fun for us. See our Fonts in use section for more.

Ambroise: Download the Try-out version!

We’re very pleased to announce that the Ambroise OpenType Full Family of 28 fonts is available in Try-out format: This license solely grants you rights for preparatory works, evaluation and internal testings and must only be used by the licensed owner. Neither production, nor final sketch, nor final artwork are permitted. The Try-Out version includes only capitals but H replaces R, Z. For the minuscules, n replaces h, u. For the figures, 6 replaces 3, 9. Minimal punctuation includes ,.-

→ Download for free the Ambroise Full Family of 18 fonts in Try-out format

¶ Originally published on Typofonderie Gazette, on 27 September 2016. Une version française est également disponible via Typofonderie Gazette.