Hannes Famira

I seem to have a penchant for dropping articles. So I lightly edited the text to turn my ramblings into actual human speech. Other than that the following is pretty much what was said:

Victor: If you’re designing a text typeface, when do you first begin to think about the italic?

Hannes: In an ideal case or in reality?

Both.

The reality has never been an ideal case. It’s always somebody who needs something quick. In reality, I never get the opportunity to follow the steps that I wish I was afforded. Theoretically I think it would be nice if a person who designs a roman typeface had enough time, at some point in the development of the first weight or first style of the roman typeface, to start thinking about different weights and different construction.

For my students I say when they almost have a complete lowercase character set: “As homework think about what your italic would look like. Just a couple of characters—‘a e g f r s’—the difficult ones—and ‘n’ and ‘o’—control characters. Just make sketches for that and then think about what that means for you. [Work on the] roman and then build the [italic] a little bit farther. And when the roman is almost finished, continue working on [the italic] so that you don’t finish one thing and then start the next thing. If the processes are a little bit integrated then the typefaces can feed off of each other.“ Otherwise you might make decisions in form and shape in the roman that are like promises that are hard to keep in the italic. Or you get great ideas in the italic and then you go “I wish I’d known this when I drew the roman.”

But that is really the ideal situation in my professional work. Ninety percent of my work is really fixing other people’s typefaces. They come with terrible drawings or a missing master—they have a roman and a bold but no light. I’m the fixer. That never gives me the opportunity for a good process.

In my own work, most of the stuff that I have kicking around that I am trying to bring up to standard is so old that it’s like I have a terrible client. So I’m not sure if I have actually had the opportunity to follow my own recommendation for my students. I’ve just not been in that position yet.

If you were given a typeface and had to create an italic for it, where would you start looking for inspiration?

I would look at the roman for inspiration for the italic.

What in particular of the roman would you look at?

The roman would bring a number of criteria that I would have to match. If it’s a really clinical sans serif, or a very romantic, serif book typeface, I would try to match that in the italic. Then there are some factors that could pull the entire family in one direction or the other by adding the italic. You could make it look very dry and upright and Protestant or give it an extreme angle and big descenders and flourishes. [...] What italic goes with this Roman? If I just slant it and call the oblique the italic, that is the most literal interpretation. But where can I take it from there? And how far could I go until it’s not a sibling anymore? Until it’s not a companion for the roman anymore but becomes a new typeface? I think that there’s a certain amount of freedom. I look at the roman to find out what the italic should look like. I don’t look at books and history. I know very little about history.

So you’ve got a roman and could go in a number of directions with that. How do you choose which direction to go in?

I’m taking baby steps. If I have two or three masters of my roman that interpolate I would slant it as a first step, figure out what the slant should be, and then go in there and start fixing the branchings. And at some point it’s looking at me and it looks like an italic. If I have the italic and the roman next to each other and I can see if I’m too far away or whether they match.

How much do you slant it to begin?

The reason I would need two masters that interpolate is that I like to use UFO Stretch to create the basic starting point. So that way I can slant, slide and scale and then correct [weight] through interpolation to get a good, functioning amount of contrast without too much distortion. But how much italic angle do I need? I don’t know. It depends on the design. That choice is an interpretation of the roman. If it’s for a client then I would have to go back and show them and talk about it. If it’s for me, then it’s really just a question of taste.

Do you tend to favor more upright italics or more heavily slanted ones?

It really depends on the typeface.

Does it depend on the particular sort of style?

To me what is important is that you look at a typeface and you imagine what it was drawn for: A heavy metal concert, or a brochure for health insurance? You can kind of tell what the flavor is in a typeface. The italic angle should represent that same flavor. It’s not so much about what category it is in. I really don’t care about that stuff. For me what is important is that when you see the roman and then you see the italic it’s not a surprise. Like a family that has all blonde children and one has black hair. My italic should not do that.

So would you start it digitally or would you do sketches?

Calligraphic tools? Certainly not—because I can’t write. So I start digitally and I use [manual] drawing to figure things out when I’m not sure: “I just can’t figure out what’s wrong with this ‘k’.” Sometimes it just helps to print it out and take some a white-out pen and a marker and draw over it. Sometimes my hands know better with pen and paper than with digital outlines. Then pretty fast I can find just what needs to happen to fix it. But generally I start digitally.

So you’ve got the Roman and you slanted it. Then what sort of process from there? What’s the next thing you would begin to adjust?

The construction. Once I have an oblique, it’s still just a slanted roman. In my mind, the roman has only downstrokes and the upstroke is in the air. The italic has the upstroke and downstroke on the paper. So all the little branchings—everything that happens between the first and the second stem of the lowercase ‘n’. Instead of that little corner on the inside of the arch I would need it smooth. How far does the branching go into the bottom of the first stem? That again is a question of what expression is in the typeface. If it’s this clinical sans serif for a health insurance, it would be barely noticeable that there’s an upstroke. It would be a very high branching but rounded out.

What other adjustments would you do?

I would get rid of the serifs where the pen turns around instead of having an end of a stroke. To me a serif is the end of the stroke no matter what kind of ending that is. I’d just call that a serif. And when the pen just hits the baseline and turns around and goes back up into the upstroke then it doesn’t need a serif there. So I would get rid of those—at the top of the ‘d’, the first leg of the ‘n’, the bottom of the ‘p’. I’d kill those and decide what the stem would look like if it just bluntly ended without a serif.

And for that I think about what the pen would look like that draws my typeface. Of course it is a metaphorical pen, it’s not calligraphy but type. The pen is to me a metaphor that combines a lot of properties of individual shapes. There’s pen angle, there’s thickness, there’s width, there is the individual shape of the sides of the pen. And if you look at the typeface you can imagine what the pen would look like if there was a pen that could draw Helvetica. Once I have that pen in my mind then I can decide what that weird blunt ending on the first leg of the ‘n’ would look like at the bottom—whether it is flat or slightly concave or gets a little bit wider. Can you see how the pen turns around—is there a subtle loop?

Any other adjustments that you would consider?

I just had this really interesting conversation. One of my students said that John Downer told him that the italics should have a shorter x-height. His reasoning is that the slant causes the stems to look like they are longer and so because John Downer said it I thought “OK, this guy knows what he’s talking about”. I never considered that. I never thought that you can see the length of the stem. I actually thought to do the opposite—to slightly increase the x-height if anything, because the slant leads to a more narrow counter. If we have a medium sized roman and a medium sized italic and they’re basically identical—except the italic has smaller counters which makes them look shorter—it’s kind of like the equivalent of a medium and black. If you have a black and a light typeface and they share the exact same x-height, the black always looks a little too short. If you slant the roman it [becomes a] parallelogram and automatically decreases the counter. So if anything I would raise the x-height. I’ve never actually done that. In the end I think that the two effects cancel each other out, because if you have a roman and an italic and they have the exact same x-height it looks good.

What about width?

It’s hard to talk about width without talking about slant. If you slant characters they end up looking like they’re more condensed. So if you want an italic that looks like it has the exact same width as the roman I think you would have to slant it and then scale it horizontally [to be] wider.

But I kind of like that it has a different texture. When your eye scans a page and there’s some emphasis in two or three words, set in italic, I think it’s nice if the color is the same but the texture is different. Part of that is slant and width. So I kind of expect an italic to look like it’s more condensed then the roman. I find that elegant. That’s not a rule. That’s just my taste.

What about weight?

It depends on the design—what the typeface wants. If it’s something that I would set a poem in, I could imagine that [the italic] has a really different appearance than the roman. If it’s a typeface that is very sober then I would expect them to be much more similar.

What aspects connect a roman and an italic?

You can come from two directions there. One is just the general weight and amount of contrast. The construction obviously is different. Width might be different. Then there are the details—the way that curves run and the way that details are dealt with. What does the tip of the serif look like? Or how do you deal with those very, very pointy notches at the bottom—the negative space on the bottom of a branching? If there are corresponding sensibilities between the roman and an italic, that binds them together.

Have you had to work within any technical limitations in any of the italic work that you’ve done?

I don’t think so.

How do you know when you’ve been successful?

I’m sure there is a theory about this but I haven’t put the effort in to read about it. I think that what we perceive as beautiful is informed taste. It’s a shortcut that allows us to evaluate the properties of something on a subconscious level. We actually train this aesthetic muscle. So if my grandma has wallpaper with big flowers on it and my parents have wallpaper with small flowers on it, I go out in the world and pick my wallpaper and think ‘flowers’. I’ve been trained to perceive certain things as beautiful. When designers go to art school and they try to solve problems they rearrange that aesthetic sense and they start favoring functionality over our cultural programming. And slowly, slowly you start having an emotional reaction to things that function, and you think ‘oh this is beautiful’ which is the same stupid shortcut. And then we have to find out why is it beautiful.

I read a really interesting interview many years ago with Steve Jobs in Wired magazine. They asked him what the most exciting technology was that he had seen in the last couple of years. I expected him to talk about torrent or whatever. He said that his wife and him had been in the market for a laundry washing machine, and they had studied all the washing machines on the market and they had settled on a Miele. He said something— that sometimes he asks an engineer for a solution and the engineer has an answer but doesn’t know why that’s the right answer. Really good engineers sometimes have almost an aesthetic sense of what the right solution is and then they have to go back and really think about “why do I think this is the right approach”. Type designers function in the same way. Sometimes you look at something and say “this is beautiful”. There are reasons why we think it’s beautiful. I’m sure one could make a list why a certain solution seems more aesthetically pleasing than another. Grandma’s flower wallpaper plays into it still, because everybody is kind of caught in their own aesthetic reality.

I think the lower case ‘k’ is a really good example of how things can look beautiful when they work. And it’s super hard to get it right. And sometimes the ‘k’ totally works—it doesn’t pop out on a page of text and doesn’t make a blob, it isn’t too wide. And still it’s not beautiful. To really get the angles right and really get it right where those three strokes intersect or connect—there’s some kind of magic in there where I think, the aesthetic sense of when you get it right you know it because it looks at you and there’s a recognition of beauty. That’s a good example how sometimes you can tell what to do but you’re not quite sure why that’s the right thing.

How does that relate to you? For example, if you’ve designed a bunch of italics and you want to test them, how would you go about testing them to see if they work well.

If the type is for a client then I usually know what it is for, whether it is for an LCD display on a dishwasher or an advertising campaign for a fashion label. When I know in detail how it will be used, then that’s exactly how I can test it. I try to mimic the target application.

I drew a logo for a client. It’s not a logo—it’s a font with five letters in it. They are ready to to launch their [seasonal] collection. They have these little labels embroidered that go in the back of the shirt and they came back in a panic and said the cap ‘R’ was leaning to the left. It’s certainly not, but the stitches of the embroidery are at an angle. And while the stem is exactly vertical that 45 degree angle of the stitches makes it look like it’s leaning. I could not convince them that any correction would actually make a difference. Just think about the thickness of the yarn for the stitching. If that is one thickness of the yarn over to the left[...] So what we did in the end was I sent them ten fonts and in every font I leaned that ‘R’ two [more] units. We had them all embroidered. Those are cases where you can really test it.

Of course if you design a typeface and it’s more like a work of art, then you’re not drawing it with an application in mind and then you can’t test it as that. In that case, I invent applications—not really that I design a brochure, but I would print out text in the size that it would appear in a brochure.

Do you have any favorite italics?

I was drawing romans for years before I drew my first italic. I felt that I had an underdeveloped italic sense. The first italic that really struck me was that Fred Smeijers’ Quadraat. I thought “that is gorgeous, gorgeous as hell”. That’s still one of my favorites, but only in text sizes.

You said that you taught about italic in a course. How did you go about it?

Now in the Type@Cooper Condensed Program I have a co-teacher, Ewan Clayton. He’s going to do the calligraphy thing. We’ve been doing that together for like three years. He does, like three weeks of calligraphy. He teaches during the day then I have three hours in the evening when I try to repeat the same thing that he did with writing. I try to repeat the same thing with sketching. In the past, the students have [gone through] the calligraphy part, and it looks really good.

Then when we start with sketching, before we start digitizing, they don’t realize that this is the same thing—that they already know everything there is to know about weight and contrast and the pen and the pen angle—all these things. It is the same thing that they’ve already done [in calligraphy]. So we try to integrate it better. I kind of shadow him in sketching. We slightly modify our technique every year. It’s really the hardest thing to do.

The pen gives you [contrast] automatically, and for free. You move it in this direction and it’s that thick. With the sketching technique, if you train your hands to always make the same length of stroke with this oscillating technique, then it gives you automatically the same suggestion. But it is amazing how difficult it is to make that connection, also in very little time. [...] The way we start with italic is really that they do calligraphy with Ewan. Then I sketch the same shapes with them with that oscillating sketching technique. Up until a certain point the sheets are identical—what the calligraphy and the sketching give you. But then with sketching we can modify the pen. We can say—what if the pen was a little thicker? Then we get lower contrast. Then we can take the basic letterforms that you get through calligraphy or sketching and we can say “let’s make a really, really low contrast” and “let’s make it really really condensed” and “make it really, really wide”. Then we massage the letter shapes and hope that the connection to the calligraphy doesn’t get lost. It’s still the same movement, the same stems, the same serifs, the same arches that just get wider or fatter or less thick-thin contrast. That phase of massaging the letter shapes, exploring the design space, seeing what are the variables, the forces that are impacting the letter shapes, that create the different appearances of the same typeface—that’s what usually leads them to their own design. Somewhere in exercises something happens and they have two or three characters that look really awesome.

So what would be the most useful advice that you could give to a student? They created a nice upright text face, and now they want to go off and create their own italic.

We’re sketching roman and italic at the same time. We would have a session where I tell them “now give me a lowercase ‘e’ with this and this and that variables” and “now give me the italic brother of that”. They can see what the criteria are that are shared and what is different. [...] We approach the italic not as a different chapter but just a different variable in the design space.

I don’t know what good advice I would give them. No cheating? Hydrate? I don’t know.

What do you find is the most difficult aspect of italic design?

The really, really pointed angles in the negative shapes—to get that right, especially with low contrast. To correct the color in the branchings without seemingly introducing a high contrast. Localized high contrast but without changing the overall appearance of contrast. I think that’s tricky.

Another thing that I think is tricky about italic is that if you start from a slanted roman then all the stems have the same angle. But there are two things that can really throw you off there. One is that the smaller shape is, the harder it is to see what angle it has, which means that the bigger features are signaling they are slanted much more clearly. If you have a lower case ‘f’ that goes from ascender down to the descender—that’s a really long slanted line. It’s much easier to see that that has an angle than the lower case ‘i’ next to it. It’s one thing to measure and to say “yes, they have the same angle” and the other thing is to look at the page and say “but do they seem like they have the same angle?”

The other thing is that the italic angle is created through the angle in the actual strokes, but much more importantly in the counter shapes. When you actually have an upstroke, then the counters don’t have two parallel stems on either side, but have an upstroke on one side and the downstroke on the other side. The downstroke has the slant of whatever comes from the roman, and the upstroke has a much steeper angle. This means that many characters look like they are much more slanted than their stems are, because the counter shape has one side that is just much steeper.

So what do you do with that?

I correct it. I trust my eye.

The other thing is round shapes. I have worked for a lot of people and with them and everybody has a trick for how to turn a roman into an italic. What a lot of people do is that if your italic angle is ten degrees, you slant everything ten degrees except for the bowls and the round shapes: You rotate them five and you slant them five.

Lucas de Groot showed me a crazy good trick that works better for me. That is to just slant the ‘o’ like every other character, and then counteract the increased stress in the top right and bottom left. Before you correct the extreme points, when the extreme points on the sides are still slanted. You have the extreme point which is still horizontal and the right extreme point (which is now slanted). Move that point five down and three in, and pull the top extreme point five to the left. It works for a fat typeface just as well as for a light typeface. How many points you move it has something to do with how high and how tall your x-height is. Because you’re always somewhere around 470 to 550, it works pretty much all the time. And it looks better than the combination of rotate and skew. Crazy. Lucas is a genius.

You’ve worked for a lot of different designers and picked up different tips and tricks. Who have been the most influential people in the way you think about italics?

Lucas. I worked for him for three years or so and in those three years I only worked on one typeface. That was a really large family—thirty-six masters. We had a lot of extremes—super super fat and light and different x-heights and different ascenders and descenders... That was really intense. It’s like sit-ups. Good exercise. And he’s a great teacher.

In a big family like that, where you have a wide range of weights, what do you do with the angle?

I think it depends on who you’re drawing it for. Are you drawing the masters for yourself so then you can interpolate anything in between and pinpoint what you get? Or are you drawing a system for a client? Like a variable font or multiple master font—where anything in between the masters will be perfect? If you work for a client who is not a type designer, who wants to pull on levers and just get a typeface they can use, in that case I would say that every master has to be adjusted for weight and slant and everything. If you draw it for yourself, as a system to create instances, then I think the opposite is true—that you have to isolate variables and say “all of these guys have the same slant but they differ in weight”, “all of these guys have different slant but they have the same weight”. That way if somebody asks you “can you create this typeface but ten percent heaver?” or “we need a little bit higher contrast because the ink in our printing presses just changed and it doesn’t look right”. If you have a lever to adjust contrast you don’t want the weight to change right along with it.

I think that a lot of people right now are turning their interpolation masters, their [multi-axes] design space for their typefaces into variable fonts. I haven’t seen it yet, but my suspicion is that at some point users will create instances that just don’t look very good because type designers draw the masters for themselves. Then they interpolate members of a family that they export, and they test, and they’re happy, and they give it away. If you use those masters that have isolated variables, and don’t compensate—tall x-height, does that mean it’s also fatter so it has the same color? If you give [control over] those things away you don’t know what you’ll get.

I want to thank Victor Gaultney for generously allowing me to post his interview here.

Here are some links for things, people and places mentioned in the text
• University of Reading
• Type@Cooper typeface design program
• Victor Gaultney
• Victor’s thesis “Designing Italics”
• Ewan Clayton
• John Downer
• Lucas de Groot
• Gerrit Noordzij
• The complex task of the serif
• Robofont – UFO font editor software
• UFO Stretch – type transformation software