Erik, in the preface to your biography it says Real was based on a special medium weight of Akzidenz Grotesk®. How did that occur?
Erik Spiekermann | I have been talking about this forever and it has been on my list for many years: Long before I owned Akzidenz Grotesk wood type, I perceived this particularly light weight of ‘AG’ as the ideal typeface. When you receive that so-called ‘medium’ weight today, it appears much bolder. Ferdinand Theinhardt also designed a face in this weight. To me this is the essence of Swiss Typography. They have designed so much with it, you can recognize it in so many good posters.
During my time at Berthold I came across that typeface in photo composition, but later it wasn’t available digitally. It seems they took smaller sizes or AG Old Face for their digitization. If you have a look at the Zurich Airport signage for example, you’ll notice that the digital version of the medium weight turned out too bold.
However, to me AG medium is the pure typeface. It does not have that peculiarly slanted ‘&’, that is well known in other weights, albeit the ‘ß’ is a bit weird – which is why we changed it –, not to mention the stacked ‘8’, which I don’t like, and this weird long ‘t’ … I get the feeling that in all this time several people have worked on it. When comparing the status quo with early versions from the 1890s, you can tell how much it has been refurbished – and yet it is not as boring as Helvetica. Without any dents and differences in stroke contrast it is perfect, but interesting. A truly amazing story.
Whether it’s FF Meta®, FF Info® or FF Unit®, your particular liking of the double-story ‘g’ is well-known. In FF Real you have once more decided to add this and other so-called anglo-american features or additions, that we are familiar with from American typefaces such as Franklin Gothic™ and News Gothic™. Where does this preference come from?
Erik Spiekermann | I wouldn’t call it additions, but echoes. There is a reason why we find roman types more legible than sans serifs; the roman alphabet has more rhythm, it has more contrast and great readability because it’s not uniform. FF Real is quite uniform. We thought round punctuation might neutralize this uniformity. That’s also how we ended up including the anglo-american echoes, for example the ‘8’ with a loop and the double-story ‘g’, which seems a bit old-fashioned and has slight blots, but certainly has much character. The anglo-american model isn’t static, it has more contrast. Another reason why I like the double-story ‘g’ is because it then turns into a text face. And because it is my personal tradition.
We should have kept the regular ‘g’ as an alternative in fact. FF Meta is equipped with both versions. To be honest, I forgot about it.
Ralph du Carrois | I’m pretty sure that it’s trapped in the font editor somewhere, but ultimately wasn’t exported.
Besides those echoes you redrew several other letterforms. Which changes were particularly important to you?
Erik Spiekermann | Two notes on the figures: I like the ‘1’ with a bottom serif very much, because it takes up more space in the tabular figures. ‘f’ and ‘t’ are a little wider as well, to make them appear less awkward. I find that stacked ‘8’ particularly peculiar: I always get the feeling that it might fall down. The looped ‘8’ suits ‘&’ and ‘g’. These are the core differences.
When you take a close look at News Gothic you’ll notice that it is not so very different from AG medium. It is more condensed and features this oval shape, but the ‘s’ is almost identical and also the ‘a’ isn’t too far away. To me these are the pure shapes of the latin lowercase letters, equipped with the right contrast where it’s necessary, or rather where it’s intentional.
Of course we also redrew the ‘ß’. What’s more, I insisted on a ‘y’ with that Spiekermann-like edge – against Ralph’s advice …
Ralph du Carrois | That was quite funny: I had to give up on the ‘y’ and in return I was allowed to install stylistic sets with round as well as square punctuation.
Erik Spiekermann | … because I just think the ‘y’ in AG is weird. Take a look at the British Transport alphabet, that is based on AG, for example; here they’ve changed the ‘y’ as well, because it always looks different in English. Of course they included round punctuation as well – we have both. But I do prefer round dots. It’s quite amazing how different type looks just by these changes.
The font info reads: “While Spiekermann drew the alphabets, he passed on the font data to Ralph du Carrois who cleaned it up and completed it.” How did this design process work out?
Erik Spiekermann | I drew almost all of the characters digitally. I usually did this in between other things, so some curves have bumps here and there. I then asked Ralph not to design things from scratch, but to clean up this data. The book weight is the one I started with. It’s a bit bolder than regular, but lighter than medium. Johannes Erler wanted that book weight, because for him regular was slightly too thin for certain sizes.
Ralph du Carrois | In this project we’re a bit like studio musicians: We listen to or look at the things at first and begin to consider what can be optimized. Mostly, as in this case, we layer the received outlines into a mask in the background and then build everything on top of that. We do so, in order to make full use of the overlap possibilities in Glyphs, which give us an advantage once we reach the multiple master process.
Basically it’s about really taking a close look at the data and asking yourself: What am I aiming at in this process? We don’t want to take on everything that’s in the sketches identically, but we really want to comprehend and yet always overlook – question it and constantly compare it to all other characters than run through this process.
One of my principles is making everything the same. Ultimately this should be the highest goal in designing a type family, in order to receive a working product at the end of all those design phases. This means that one should try to keep up the curve tension in a span between 2em2 and 222em to move and keep formal small matter as even as possible. Whenever several designers are involved in a process it is particularly important to take notes again and again: Who did what and why where changes made? Consequently all persons involved can retrace the steps and design evenly. Tackling such a case with emotional comments such as: “Make it look nice and be sure that it doesn’t gain too much strength in the upper left part.”, does not allow us to advance.
In the meantime FF Real Text and FF Real Head have been extended to 13 weights respectively, ranging between hairline and black. Is the 2em weight that we’ve been introduced to in your collaborative work on Fira, a new standard?
Ralph du Carrois | No. But it’s relatively easy to create a 2em weight, because you don’t have to recognize contrasts. Whether it is too bold or too light, anyone will tell at 2em right away.
While working on Fira in 2011, we took our first steps in Fontlab, however, we quickly came up against flexibility as a limiting factor. Right in the middle of the ongoing process we transferred the entire project to Glyphs. It was a hell of an act. In the end we created the 2em for the first time in a real project, because the client wanted light weights.
We’ve been using Glyphs exclusively for three years now. With the interpolation capability developed by Georg Seifert, so many possibilities have opened up for us to quickly generate stuff and maintain control over all the curves. This pretty much is making everything the same [laughs].
Inga Albers, who designed the layout for Hello I am Erik, has proven in exemplary manner, that one book can be typeset in several layers of story telling without the use of an italic. No italic has been added for the release of FF Real. Is this brave or an intentional concept?
Erik Spiekermann | This is a concept, for there cannot be a proper italic in a sans serif – it would have to be handwritten. That is why there are slanted weights and those are usually formally unsatisfactory. They’re fine for a sans serif that is lively in itself, as with FF Meta for example. But you cannot really do that to a static sans serif type. I’ve tried it: When you slant FF Real at an angle of 12°, it doesn’t look bad.
In fact we have something else in mind: We will design an italic that is going to be a completely different typeface. That is going to be an italic for stressing words, which will be needed in the scientific realm, quotations for example. Surely this is not going to be a roman, but an italic that’s going to suit Real just as much as Joanna® fits with Gill™.
Why not draw a different typeface? It is going to have the same name, it will share the same metrics such as the x-height and will appear the same. You can see this in other roman types such as Jakob Runge’s FF Franziska™, which is quite explicit, however. On the contrary, there are romans that appear rather differently. This is well-known in classic italics, which are much narrower – Janson® is a popular example. The italics look completely different in these transitional types. We’re used to it there, but in a sans serif people like to think the type needs to be slanted and pseudo-like equipped with some thingy on ‘a’ and ‘g’ to look somewhat lively. We are going to create an italic FF Real that is a typeface of its own.
For this conversation you brought along not only Akzidenz Grotesk type specimens, but also some FF Real wood type letters. What is this about?
Erik Spiekermann | For the first time, with the release of a digital typeface, this very typeface is also going to be available as wood type. This is a world première. I had an entire set of FF Real milled in 16 Cicero1 wood from two colleagues in Romania and another set in 20 Cicero Resopal from Fablab here in Berlin. We are going to present the first letterpress results very soon.
The wood type is featured with all changes necessary for the analogue experience: no accents or umlauts on capital letters and a rather small character set. We’re going public with this simultaneously with the release of FF Real. As prototypes for all future type designers we can say: Guys, if you’d like your type made from wood, send us your data or set up the necessary files without side bearings. Three units on the sides – that’s it. I can imagine that there a couple of folks out there, who will find this interesting and are ready to spend 1,200 to 1,500 euros, depending on how much they need. For round about 500 or 600 euros you’ll receive one set of 16cic. That’s almost the cost for the entire digital package.
Your last type release was HWT Artz – your personal interpretation of the historic Block as it were. Can we expect more interpretations of your Berthold heroes in the future?
[link not found]
Erik Spiekermann | I believe all is worked off with this. However, I would very much like to draw my own version of the two families Block and Berliner Grotesk. While “Berliner” is the light weight …
Ralph du Carrois | We could make a 2em from that [laughs].
Erik Spiekermann | … the Block Family is bold by birth. It starts with bold and moves on to extra bold. In the meantime Berthold also created a medium version, but this one has a completely different origin. Thus, designing the Block family from light/regular to extra bold would be a great pleasure. For the time being I am probably going to draw lowercases for HWT Artz as well as a light weight. In the end HWT Artz is Block with clean curves.
Ralph du Carrois | How do we actually bring those 2em on wood?
Erik Spiekermann | It’s possible, but you don’t want to print that.
1. Cicero is a larger unit used to describe type sizes within the Didot system; it equals to 12pt.
2. 1em is the smallest unit in a font editor, used to describe data of letters and other characters.
Real, Franklin Gothic, News Gothic, Franziska are trademarks of Monotype GmbH and may be registered in certain jurisdictions. Akzidenz-Grotesk is a trademark of Berthold Types Limited. Meta, Unit, Info, FF are trademarks of Monotype GmbH registered in the U.S. Patent and Trademark Office and may be registered in certain other jurisdictions. Joanna, Janson are trademarks of The Monotype Corporation registered in the United States Patent and Trademark Office and may be registered in certain jurisdictions. Gill is a trademark of Monotype ITC Inc. and may be registered in certain jurisdictions.