Deviant Art Shop Forum More Submit Join Login

# MartinSilvertant

Dutch entropy

## How to design a typeface (Part 3)

Mon Jun 3, 2013, 11:18 AM

### How to design a typeface (Part 3)

##### Serif — Roman

In the last two article of the type design series we made a geometric sans serif and added italics and a bold weight. The geometric sans serif is the easiest style to do, but for me personally things really get interesting with serif typefaces. Of course there is the addition of serifs, but the structure and weight distribution is also much more advanced. In this article I will make an introduction to designing serif typefaces. This article won’t be as detailed as the previous ones because this third article serves as an expansion on the first two. In other words, I won’t get into the basic vector tools or type principles, nor will I show you how to design the letters step by step. I will explain techniques in detail when necessary, but generally I will refer to the techniques described in the first article particularly. So if necessary go back to the previous two articles of the series.

The first letter
With a sans serif typeface I generally start with a lowercase letter like /I to define the x-height. With serif typefaces I first want to establish the style of the serifs before I define the x-height. So, usually I will start with a capital /I. I start with a vertical stroke of the same weight as a Regular typeface. I convert the stroke to outlines and then add two horizontal rectangles and position them at the bottom and top of the vertical stroke. Now merge the three shapes with Pathfinder and remove any unnecessary vector points which may show up at the very top and bottom after you merge the shapes.

##### Figure 1

Now you have a very basic serif capital /I, and this is where the fun starts. You might be aware that there are quite some different serif styles. Before we continue working on our serif typeface, let’s have a look at a few different styles.

Type categories
In figure 2 are the main serif categories exclusive the slab serif. I suppose you could call these the classical type classes. The 4 classes go from old to new: Venetian/Humanist (Jenson), Garalde (Caslon, Garamond), Transitional (Baskerville) and Didone/Modernist (Didot, Bodoni). If you look at the weight distribution and the details of the serifs you will notice there is a gradual transition from calligraphic to mechanical. This transition is parallel to the development of improved printing techniques and refined letter casting and designing processes. If you’re curious about the history of Roman typefaces (meaning typefaces in the Latin alphabet), read my appropriately titled article "History of Roman typefaces". If you want to design a serif typeface it’s tremendously beneficial if not required to know how serif typefaces developed and which aspects define their style. It’s very hard to make the letters consistent in design/style if you don’t really know what style you want to design or what characteristics are common for that style. For example, it wouldn’t be wrong to design a humanist serif typeface with the weight distributed vertically, but since the weight distribution is historically diagonal in humanist serif typefaces and it’s vertical in transitional serif typefaces, you couldn’t consider your hypothetical design historically human. Most contemporary serif typefaces are hybrids of different styles though, so it’s not wrong to mix. I just want to alert you to the fact that some knowledge of the different styles and their historical context will make it much easier to make specific decisions for each letter and to create a consistent whole. That’s where the real challenge is. It’s one thing to design a beautiful letter, but it can be difficult at times to translate the same style to a different letter and keeping the weight distribution and general weight consistent throughout the typeface. And then there is another challenge: to design italics to fit the roman.

##### Figure 2

Serif styles
Before we will go back to the design I want to show you some different types of serifs. I think by now you’ve already decided what kind of serif typeface you want to design, but let me give you a few tips and show you different types of serifs—the serif types and serif classifications don’t necessarily need to match, and it may bring interesting results to mix a few styles. I’ve tried quite a few things out myself, and sometimes you get a lucky accident. Other times I have a good idea in my head but once the idea is executed it’s not quite as great as I thought it would be, or I just haven’t figured a proper way to execute the idea. You tend to go a lot by trial and error, and as you get more experienced things will become more intuitive, but I suppose there’s always the chance to get into a boring routine, so it’s good to mix it up once in a while.

##### Figure 3

In figure 3 are some different serif styles, but probably not even all of them.
• Blackletter typefaces have all kinds of different “serif” shapes, but they always resemble a rhombus because the gothic script is calligraphic.
• Glyphic serifs are usually triangular in shape and very subtle. These serifs resemble the chiseled Roman lettering. There aren’t many glyphic typefaces around, but a good example of one would be Astoria. You will likely find glyphic typefaces under the name ‘semi-serif’.
• You can recognize oldstyle serifs by their triangular shape and a gradual slope.
• Transitional serifs are usually more pronounced.
• Wedge serifs come from the stem at a sharp angle at the bottom rather than a gradual curve.
• The hairline serif is of course unmistakable. This serif can be seen in the Didone types (Didot and Bodoni), but the Scotch types feature the same kind of serif but with reduced contrast.
• Bracketed serifs feature sharp serifs with a gradual curve towards the stem. There is some common ground between bracketed serifs and glyphic serifs as well.
• The bracketed slab serif is not seen that often among slab serif typefaces, but the ones around are often quite beautiful; they combine the elegant look of the classical styles (Garalde and Transitional mostly) with the strength of a slab serif. The result is often refreshing and what we would consider “modern”. Clarendon typefaces also belong to the bracketed slab serif typefaces.
• The unbracketed slab serif is the more conventional style. Also called Egyptian, these typefaces emerged in the early 19th century and combine the legibility of the classical serif with a more rigid construction to create a crisp type with a lot of impact.
• The Tuscan serif can probably be divided into several sub-classes. There don’t seem to be any clear rules regarding the design of the Tuscan serif, but they all have extravagance in common. The most common type of Tuscan type is the typical Western lettering with the overly big slab serifs, but the serifs can also be highly decorative. It’s common for Tuscan serif typefaces to feature decorations at both sides of the center of the stems.
• Today  we combine roman and italic (which has a cursive construction) like they’re the same typeface, but the cursive type started out as a distinct style. So initially it wasn’t only used for emphasis but whole texts were set in a cursive type. Although we don’t recognize it as a distinct type category any longer, we have to acknowledge that the cursive serifs are very distinct and deserve their own class. The cursive serif can take on all kinds of shapes, partially depending on the shape of the serifs of the roman, but cursive serifs are always either curved or diagonal to resemble cursive writing, which is the root of this kind of type style.
• The sans serif is without a doubt the easiest style to recognize, simply because it’s the absence of serifs.

In figure 4 you can see a few samples of a wide spectrum of possibilities for the serifs. For those who are interested about the designers:
• Mercury was designed by Jonathan Hoefler and Tobias Frere-Jones / 1997.
• Collis was designed by Christoph Noordzij / 1993.
• Jannon was designed by Jean Jannon in 1621 and digitized by František Štorm / 2001.
• Alegreya was designed by Juan Pablo del Peral / 2012.
• Aghari is my own design. It’s still in development, so the design may still change slightly. On that note, any feedback is welcome.

##### Figure 4

Serif letters
So I think we have some background information and inspiration for our own serif typeface now. Of course one should still consider the application of the typeface. If you will use it for newspapers, one of the requirements will be that the typeface should be relatively condensed to allow for more text in a given space, especially for the typeface used for headings. A book typeface needs to be legible and shouldn’t tire the eye so quickly (which geometrical typefaces, high contrast typefaces and typefaces with a focus on the vertical strokes end to do). So these are things you will have to consider, but during the design process a lot of changes can still be made. As I explained in the first article though, you will save yourself a lot of time if you do establish the x-height so you won’t have to change it later. Other than that, I think you know what kind of typeface you want to create, so let’s first focus on the design of the serifs of the first letters, and then worry more about the general construction of the typeface. Mind you, I do recommend you sketch out some ideas on paper. Most of my process is digital and I’m embarrassingly bad at drawing letters on paper (for a type designer that is), but more than once I did get some great ideas out of just doodling letters—often pages full of the same letter, just to get an idea of all the variations and possibilities.

Thus far we have a very minimalist capital /I. Now it’s time to mold this shape into something more exciting. I won’t do this step by step; the basic principles are described in the first article and showing you where to place what vector point at this point of the type design article series will limit your potential to create something unique. The reason I have more fun designing serif typefaces than sans serif typefaces is because there is more variation and more potential to design something unique, because you can bring originality in the serifs and you can play around with the weight distribution. The latter you can also do with sans serif typefaces, but usually a lot more subtly.

What I can tell you however is that usually the two vector points at the base (the corner between the serif and the stem of the letter) are moved up so you can add a bézier curve. Don’t be afraid to add vector points where necessary. For example, my Aghari typeface features vector points at the center of the stems so I could make the middle a bit more thin, which will give the typeface a more organic look. Also, because by thinning the center I essentially focus the weight at the top and bottom. The fact that the center therefore becomes lighter gives me the opportunity to compensate for that weight by making elements in the central region more prominent. For example, by moving weight out of the center, in the center I suddenly have a cleaner canvas as it were to do something special with the way the strokes connect in the letter /k, or do something special with the bowl of /a etc. You have to play around with the black and white space to create something balanced and interesting. I can’t tell you how to do that, other than to practice and look at the details of other typefaces. If you have the typeface on your computer, convert the letter to outlines and inspect the use of vectors and bézier curves.

Symmetry
I hope by looking at other typefaces you will manage to design proper looking serifs for the capital /I. As you can see in a typeface like Didot though, serifs don’t need to be very complicated. Also, most serif typefaces feature symmetrical stems, but depending on the style of typeface and its application, it may be interesting to create something more dynamic. As you can see in figure 4, the stems are symmetrical in Mercury, Collis and Jannon, and this fits the general design perfectly. In Alegreya you see the designer played around with the shapes to create tension and a sense of direction. The result is a vibrant, dynamic typeface. In terms of design this makes things a lot more interesting, but with typefaces you need to consider both design and function. Mercury was designed for newspapers, and this printing environment (the inks, paper and machinery used to produce newspapers) won’t do a highly dynamic typeface much justice. Ink may be absorbed in places where it wasn’t meant to be, and this may distort an already vibrant and perhaps even unruly design. That is not to say all newspapers should look boring (quite the contrary, in fact), but these are limitations which should be considered. For starters a typeface for a newspaper will likely have flat serifs to bring some order to the baselines. In my Aghari typeface you can see I also subtly played around with asymmetry. Not so much to create a lot of tension, but just enough to keep it interesting. Most of my serif typefaces are designed with editorial use in mind, so they’re usually quite legible, feature a medium to relatively high x-height and take a lot of inspiration from Renaissance typography when it comes to the weight distribution and relative proportions between letters.

##### Figure 5

From capital to lowercase
Alright, so I always start with the design of the capital /I to get the stems and serifs the way I want. From there I can easily expand the typeface. Two /I’s with a horizontal bar and you can design /H, cut off the bottom of /I and replace it with a tail and you get /J and use the /I for the stem of B/D/E/F/K/L/M/P/R/T/U/Y. Before you do all that work though, it’s better to first focus on the lowercase letters. As the fact that you can create so many capital letters by using the /I shows, there is less variation in capital letters. I mean, you can still vary a lot, but capitals will always look blocky. It’s with the lowercase letters that you can really define the style. Firstly because the letters vary more, but also because you can play around with the ascender and descender lengths and the x-height. The first step though is to design the top serif (figure 6). To do that, we first take the /I and cut off the top half. You then replace the top with a top serif, as seen in the lowercase /i. I advise to get the shape of this top serif perfect because it’s one of the most prominent elements in a serif typeface. It’s also fairly easy to design something original for the top serif. You can design anything resembling something from the shape of a triangle (Mercury, Garamond) to the shape of a rectangle (Museo Slab, RePublic), but I don’t even see why you couldn’t try a circular top serif. Just play around with it. Also keep in mind though that the shape of this top serif will be one of the elements which creates the texture and it can shift weight. For example, the top serif of Jannon is very sharp at the top, so this will create tension in the x-height and provide a sharper texture. By contrast, Garamond also has triangular top serifs, but they’re a lot softer and subtle. The top serif in Mercury is rather small, which resembles the weight of the regular serifs. By contrast, in my design of Aghari the weight is shifted away from the middle of the stems and it’s largely focused in the base and the top serifs, so while in Mercury the stem is dominant, in Aghari the top serif and base are dominant. So, always keep in mind what each design decision will do for the texture, weight distribution and general appearance.

##### Figure 6

For the tittle (the dot) of /I (and /j), be reminded that it should never be a perfect circle. The tittles in Collis and Jannon are almost circular, but if you superimpose a circle on the tittle and zoom in close enough you will see that the tittles are in fact not perfect circles. I attempted to show this in figure 6. Here I zoomed in on both tittles to show they subtly differ from a circular shape. In case of Collis the tittle is white. In case of Jannon Text I had to color the tittle black and color the perfect circle white, because due to the shape of the tittle if I were to do it the other way around, the circle would completely cover up the tittle of Jannon Text. The difference in presentation between the two tittles aside, you can clearly see that both tittles are in fact not perfect circles.

##### Figure 7

More lowercase letters
Once you have the capital /I and lowercase /i, things become easier for a little while because you can create lowercase letters in a similar way to the process of a sans serif. First you copy /I (excluding the tittle) and extend the length to ascender height to create the /l. Copy the /I again to use as a stem for /n (instead of using rectangles when we designed the sans serif /n in the first article) and follow the process of the first article to create the whole letter, but obviously with a higher contrast between thick and thin. Look at figure 8 to see a quick way to create /n.

##### Figure 8

You can vary more with the point where the arc emerged from the stem, how thin it starts from the stem, and where the weight is distributed along the arc. In figure 9 are a few distinct styles. Click on the picture to see a larger version so you can inspect the details. Note the different ways of distributing the weight, the different use of angles and how the base serifs tend to show some variation within each letter.

##### Figure 9

Once you have the letter /n, you can follow the process described in the first article to create /h (by using /l),  /m, /r and /u. In case of /m, you may want to decrease the middle base serif a bit to increase the aperture (aperture= how open the letter is). In case of /r I would slightly increase the length of the right base serif to compensate for the empty space underneath the arm. Also look at the /r in other serif typefaces for inspiration, because there is quite a lot you can do with the arm of /r.

##### Figure 10

As for /u, be reminded that you have to condense the letter a bit because a /u always looks optically wider than /n so you need to compensate for that. If you just flip the /n upside down though, you might discover the result is not quite as pleasing as it was in case of the sans serif. This is because the base serifs of /n now become the top serifs of /u, and the top serif of /n becomes the spur of /u. So after flipping the letter /n, cut off the top part and replace it with the top serif of /i. Now remove the bottom serif of /u and replace it with a spur. You have some opportunity here though. In some typefaces this spur somewhat resembles the shape of the top serif (the same as the spur of /d), while in others the spur is a curved shape as can often be seen in the lowercase /a. The former is much more common, but this it’s an area where you can make your typeface more distinct—if it fits the design of your typeface that is.

You might also want to increase the length of the right top serif a bit, and it probably looks best if you make the curve at the bottom a bit more squared compared to the /n. I don’t know why, but this ends up looking better every time I do it. I suppose it’s one of those optical illusions again, such as the /n looking optically wider than /n. A very good tip is also to adjust the letter according to the eye and not according to strict rules. Initially you need rules to keep your typeface consistent, but after that comes a process of fine-tuning according to the eye. So don’t be afraid to move the connection point between the stem and the arc up a bit in /u compared to /n, as long as it looks good. The design process of a serif typeface tends to be a lot more like molding. A lot more manual work is involved, and a lot more fun with the weight distribution. In figure 11 you can see the process of how to go from /n to /u. Do note the alignment to the guidelines. So far the top of the lowercase letters were either rounded or sharp so they aligned to the same guideline. Now we turned the /n upside down and suddenly we flat, rounded and sharp objects, and so we need an extra guideline; one for the flat objects, and one for anything which looks optically smaller (rounded and sharp objects) so the objects can extend a bit beyond the baseline and x-height when necessary. This might sound confusing, so just have a look at the guidelines of the typeface provided at the end of this article.

##### Figure 11

Black and white
I’m going on about optical adjustments, molding and weight distribution, but something occurred to me. If you want to create a typeface with an even color (meaning the black weight is optimally distributed), you can think of a white field with a black blob on it. Depending on the way the black blob interacts with the white, you can shape letters. A wider letter (like /m or /w) requires a wider field while a small letter (like /l or /i) requires a smaller field, and the black blob shrinks or grows relative to the area of the field. As such a mathematical system could be constructed to calculate the optimal size of the black blob per letter given a certain weight, because obviously a bold font will have more black space than a light font. If such a mathematical system is constructed, you could also see how much each letter of a given typeface deviates from the optimal amount of black relative to white. This should give interesting statistics, but it should be said that an even color is not necessarily what you want. Didone typefaces like Didot and Bodoni for example don’t have an even color at all; most of the weight is in the stems, and so there is a focus on the vertical strokes while horizontal and many diagonal strokes are very thin or in some typefaces remain completely absent, since the thick strokes alone are enough to recognize most letters. I personally think typefaces with a very even color tend to be dull. If you vary a lot with where you focus the weight though, you will get something wacky and quite hard to read. An excellent example of this is ITC Biblon. Love it or hate it, but it can give some very interesting results. One should just be reminded of the application of the typeface. ITC Biblon works excellent at display size, but at text size the texture resembles that of a blackletter, or something handwritten; so if you want crisp and modern, ITC Biblon is definitely not a good text typeface. If you want something playful and quite dark in color then ITC Biblon might be a good choice. It’s one of those typefaces with such a distinct character it might be difficult to find a project for it though. But then, you don’t select the projects for the typefaces but the typefaces for the project. There are plenty of typefaces I would love to use but just haven’t gotten a project which it would be perfect for yet.

##### Figure 12

The letter ‘a’
In a previous article I talked about the importance of the letter /a; how it’s one of the most prominent letters and one of the most commonly used letters and is therefore a great letter to show all you’ve got and really make your typeface distinct. In the English language the most commonly used letters are /e (12.7%), /t (9.1%) and /a (8.2%). In all other Western languages the /e and /a are also among the most commonly used letters, and in Spanish it’s even the most commonly used letter. So really, spend extra time on these letters to make them perfect. In case you’re interested in the statistics of the frequency in which other letters are used in Western languages, go to this page. Not only is the /a a frequently used letter, but it’s also quite a challenging letter to design. Full of possibilities, but also challenging. All I can say here is to follow the previous two articles and have a look at the /a in other typefaces. Make sure the proportions are consistent with the other letters, meaning the white space inside /a is roughly equal to the white space in /n. The letter /n is probably a good letter to keep as a master for the width, so you compare every other letter to /n. Some variation in proportions may be wanted though. Particularly Venetian and Garalde typefaces tend to feature more variety in the widths of letters; often such a typeface will feature a condensed /B and /L, wide rounded letters (D/C/G/O/Q) and a wide /M and /N. In slab serif typefaces you will probably find the least variety in proportions among the serif typefaces—monospaced typefaces excluded of course. In figure 13 I selected a range of different /a’s. I wish I could add another 2 rows but that’s really pushing it. Besides, you will find plenty of good typefaces on myfonts.com or on one of the major font foundries like Darden Studio, Emigre, exljbris Font Foundry, FontFont, Hoefler & Frere-Jones, HvD Fonts, Process Type Foundry, Storm Type Foundry, Suitcase Type Foundry, The Enschede Font Foundry, Typejockeys, Typotheque, Underwear, Village etc.

##### Figure 13

Bee dee pee quu
The letters b/d/p/q are quite strange in the sense that the letters are reflections from each other at first glance. What makes the letters distinct from each other is the weight distribution. Typefaces have roots in calligraphy, so if you were to visualize writing with a calligraphy pen, I think you can imagine where the stroke will become thick or thin. The weight needs to be distributed according to this principle. If you don’t do this, your letter will look awkward. At the right in figure 14 is the /b reflected. As you can see it obviously resembles a /d, but while the weight distribution and general shape of the bowl worked for /b, once you reflect the letter it becomes very awkward. This is because the weight distribution no longer follows the calligraphic flow typefaces are based on.

##### Figure 14

The letter ‘e’
The letter /e is also a common letter, and a prominent one. There are quite some typefaces where the /e has some quirk to make the typeface more distinct. It’s also fairly common to create some tension in the /e, because its shape lends itself well for that. As you can see in figure 15,  Requiem Text focuses its weight in the lower left and top right corners, but the letter is balanced. By contrast, Relato’s weight is distributed vertically; most on the whole left side, and a bit on the top right. A vertical weight distribution tends to create less tension, but if you look at the shape of the eye, the letter wants to move to the right. The letter isn’t static, but has a sense of movement. In Jannon Text the weight is distributed diagonally, and in fact most of the weight is in the bottom left curve; a characteristic of oldstyle typefaces. Despite the diagonal weight distribution, the letter features hardly any tension, though a bit of tension was added by letting the horizontal stroke slope inwards towards the eye. The top right is quite flat though, which focuses the letter on its own spot rather than creating a forwards tension. In Leitura the weight distribution is roughly the same as in Requiem Text, but other than Requiem Text, Leitura does have a forward tension. This is created with the shape of the eye, the curve on the top right going a bit to the right instead of immediately to the left, and the terminal (the ending of the curve at the bottom) extends slightly beyond the core of the letter, giving the letter not only a sense of moving forward, but a sense of rotation to the left.

So there are several tricks you can apply to the design of your /e to both make it consistent with the rest of your typeface but to also make it distinct. Also be reminded that although you can play around with the weight distribution within a typeface, it’s going to look really awkward if one letter features a vertical weight distribution (Transitional) and the other features a diagonal weight distribution (Venetian or Garalde). Also, the weight distribution of /c needs to be rather similar to /e. In fact, remove the horizontal bar in /e and replace the top right curve with a proper terminal/serif and you’re well on your way to designing /c. You might want to move the weight very slightly towards the bottom. Remember the example of the black blob and how it interacts with the white space? Well, the black of that horizontal bar of /e has to go somewhere now that you removed it. Removing too much black without adding any back in return will result in a lighter letter. One final thing to remember is that the /e and /c can have a fairly diagonal weight distribution anyway. This doesn’t necessarily indicate the weight distribution of the letter /o, and so the weight distribution in these letters alone are not enough to verify whether a typeface is transitional or oldstyle.

##### Figure 15

The letter ‘o’
Alright, so they say the letter /g (the double storey one) is the most difficult letter to design and generally I tend to agree with that, and yet with some typefaces (like my Aghari typeface) it seems impossible to design a letter /o which looks good and which fits the typeface. I’m sure it will become easier as I gain more experience and vision, but that doesn’t take away from the fact that the /o can be a deceivingly difficult letter to design. In the first article I showed you how to design the letter /o, and by increasing the contrast I can easily make the letter /o for Baskerville or Bodoni. Any /o with a vertical weight distribution, basically. Once the weight moves diagonally, it can be very challenging to balance it out. One trick is to move the outer left vector down, the outer right vector up, the inner right vector down and the inner left vector up, the top outer vector to the right, the bottom outer vector to the left, and finally the inner bottom vector to the right and the inner top vector to the left. This will essentially rotate the counter (the white space inside the /o) counter-clockwise. Do note that this is very different from simply selecting the inner vectors and rotate the counter with the Rotate Tool, because in that case the bézier curves rotate as well, which ruins a lot of the flow. Move vector points around and change bézier curves to get the result you want, and I can’t say it often enough, but look at other professional typefaces for reference. Especially have a look at the bézier curves if you have access to the vector design. In figure 16 you can see how I went from a simple ring shape to an /o with diagonal weight distribution. Initially the result was the fourth /o. Two things become obvious to me here though:
1. I’m not great at designing the /o. This explains why it takes me so much time to get it perfect, or why I already redesigned the /o in Aghari numerous times and I’m still not quite content.
2. That the /o of the typeface I’m designing to accompany this article needs a very subtle diagonal weight distribution, if diagonal at all.

So I made a second attempt, which you can see at the far right. I’m still not content, but at least it fits the typeface now.

##### Figure 16

The end
I think that’s it for now. I could make a step-by-step tutorial for each individual letter in each style, but I’m not sure if that’s the proper way to learn. If you’re passionate about type, you will discover most things on your own, and make them your own. I’ve had no official type design lessons, and so what you learn from me is what I learned on my own, by doing a lot of research, constantly looking at other typefaces, and of course practice a lot. With this type design article series I want you to become more and more independent, and by the end of the article series I will likely give information predominantly, and not so much show how something specific is done. Having said that, if you have any questions or would like me to make a tutorial for a specific letter, let me know.

Next time I will probably focus on designing italics for a serif typeface. After that I have an article about small-caps and ligatures to write and an article about adding full language support to your typeface, but other than that I think I’m nearing the end of this series. So if you have any subject relating to type design or typefaces you would also like me to write about, let me know.

##### Figure 17

This time I included a separate file for the typeface I designed to accompany the article. The name of the typeface is “Daser” (dA serif), and it includes guidelines for the baseline, x-height, capital height etc. to give you an idea of the dimensions of a typeface. All is included in the resource package below. Be reminded however that all resources are only meant for educational purposes; you're not allowed to distribute or alter any of the work.

## Why Helvetica is not great

Tue May 7, 2013, 11:42 AM

### Why Helvetica is not great

Type design is a rather obscure profession but even the typefaces themselves don’t get too much attention by the general public. Most people could only name a few typefaces, but among those few are always Times New Roman, Arial and Helvetica. Well-known and popular or not, in this article I will show you why Helvetica is not the great typeface people perceive it to be.

##### Figure 1

The designer
Arial is often frowned upon due to its history, but what most people don’t know is that Helvetica has a very similar history. A bit different, but as far as I’m concerned it’s just as dubious and embarrassing. As most people who are interested in typefaces might know, Max Miedinger is the designer of Helvetica. However, what most people probably don’t know is that Max Miedinger was not a type designer. Miedinger studied typesetting between 1926 and 1930 and later became a typographer for Globus department store’s advertising studio in Zürich. As a typographer he obviously saw a lot of type, and he became a customer counselor and typeface sales representative for the Haas’sche Schriftgießerei (Haas Type Foundry), so without a doubt he had expertise in type. However, this is very different from being a type designer, and this tends to show.

##### Figure 2

In 1954 at around age 44 he designed his first typeface, called Pro Arte (figure 2). It’s a condensed Egyptian (slab serif) like many of its contemporaries. In general Pro Arte is quite nice, but the letters K, Q and the ampersand (&) show he isn’t quite the master type designer people nowadays may perceive him to be when they think of Helvetica. To humor you, this is what I perceive to be wrong with Pro Arte:

K – The slab serif on the arm and leg are positioned too far to the left. Or, if Miedinger would insist on keeping the gaps in between the slab serifs consistent with the gaps in H/U/V/W/X/Y, he would have to change the angle of the arm and leg a bit. It’s an easy fix, but it’s a bad mistake to make in the first place.
Q – I have to say, I find this swash-like tail of Q quite attractive, but then I would expect some elements of the same stroke weight somewhere else in the typeface. The tail is simply not heavy enough.
& – The ampersand looks like it belongs to a completely different typeface; the top isn’t heavy enough, the thin stroke should go more gradually from thin to thick (a bit more weight in the curves), the curves look wobbly and for some reason the ampersand features quite thin and elegant serifs. The glyph is simply more refined than all the other letters. Also, the fact that the glyph is slightly flattened at the left and top left and right sides makes it consistent with the large vertical parts of the slab serifs in other letters, but it makes the ampersand look quite awkward. I don’t know what the solution should be for this, but sufficed to say the current ampersand is seriously flawed.

Plagiarism?
Alright, so Miedinger didn’t do a perfect job with Pro Arte. That shouldn’t deter any type designer from making more typefaces. Eduard Hoffmann—who was director of Haas Type Foundry by the time Miedinger joined—recognized Miedinger’s talent and commissioned a new typeface. Not just any typeface, but a very specific design. At the time the typeface Akzidenz Grotesk (bottom line of figure 3)—released by Berthold Type Foundry in 1896 under the name Accidenz Grotesk—was hugely popular, and Haas Type Foundry became alerted to the fact that they were missing sales because all Swiss designers were specifying Akzidenz Grotesk from Germany. The Swiss wanted some of that market share, and so Haas Type Foundry requested a typeface like Akzidenz Grotesk.

##### Figure 3

Akzidenz Grotesk was based on Scheltersche Grotesk (released by Schelter & Giesecke Foundry in 1880) and both these typefaces served as models for Helvetica. In fact, if you compare Helvetica with Akzidenz Grotesk, you might notice the general proportions are exactly the same. Conscious decisions were made to keep the proportions the same so it was possible to substitute one typeface for another without having to re-set the whole text. This is also why Arial was designed with the same proportions as Helvetica. Arial was based on Grotesque 215 but redrawn to match Helvetica in weight and proportions.

In figure 4 you can see the authentic documents showing the working process of plagiarizing (or at the very least being greatly influenced by) Akzidenz Grotesk. The top lines on the right are Helvetica in progress and each second line is Akzidenz Grotesk. Image provided by Nick Shinn (type designer and founder of Shinntype).

##### Figure 4

Neutrality
While the antique grotesque typefaces were warm, Helvetica was designed to be neutral; compared to Akzidenz Grotesk, Helvetica is rather cold. This neutrality was exactly what the world of design needed at the time. Suddenly a lot of big companies got rid of their hand-lettered, decorative type and replaced it with this new typeface called Neue Haas Grotesk—which would later be called Helvetica so it would be easier to market internationally. While the Bauhaus and De Stijl movements were all about minimalist modernism, the post-war mentality abandoned austerity and embraced midcentury modernism; neutrality and impact were the new modern style—something which Helvetica (and the contemporary Univers) provided. As such Helvetica became a hugely popular typeface, and during the 60’s it had major impact on the advertising industry and on corporate branding. Not only was the typeface supposed to be neutral, but as Helvetica became synonymous with modern design, it also neutralized the whole design industry and impacted modern consumerism in a major way. It set some of the aesthetics we still cherish today, and Apple played a major role in that by using Helvetica as a system default typeface. Look at figure 5 for a comparison between a Coca Cola advertisement from the 60's (left) and the 70's (right) where Helvetica is used. It's still looking clumsy compared to contemporary marketing, but for the time this was a major improvement, and a big step towards commercialism as we know it today.

##### Figure 5

Helvetica may have been neutral in the 60’s, but ironically enough Helvetica is definitely not a neutral typeface. It’s quite cold, but not neutral. In fact Helvetica is highly prolific. In that sense—and only in that sense—Helvetica is like Eurostile; cold in design but highly prolific as a typeface. A typeface like Franklin Gothic is what I would consider more neutral. If you can recognize the typeface straight away, it’s obviously not a neutral design. As such, it's definitely not a typeface to use for any and every project. Every project should be treated individually, and so it seems unlikely that Helvetica would be the optimal choice for each project. In fact, Helvetica isn’t an optimal choice at all, but more about that later.

Modernism
So Helvetica isn’t in fact neutral. Is it modern? Certainly it’s mid-century modernism, but since Helvetica is based on the same model as typefaces from the late 19th century, you can’t really consider that modern as in, contemporary. The design of ‘a’ with the curve going from the top of the belly to the stem is an antique design, stemming from the Egyptians (slab serif typefaces) from the early 19th century. So next time you’re selecting a modern typeface for your project, consider if you want genuine modernism or 19th century modernism. There is nothing at all wrong with 19th century modernism, but it has to fit the context. Although, for me personally I can imagine a typeface working well even if it doesn't fit the context. For example, to create contrast. So I guess fundamentally what concerns me more is the fact that people choose typefaces based on ignorance. Just like all other aspects of a design—such as composition, use of color, symbolism, symbology etc.—the choice of typeface is very important and needs to be carefully considered. I have to admit though, even as a type designer and typography, selecting the right typefaces can be very challenging. Partially because there are so many things to consider, but also because design is subjective anyway. First you need to know the facts though. How you may deviate from certain facts or design principles is your responsibility and possibly your obligation as a designer; to consider all aspects and bend it according to what you think fits the product best. It's partially an intuitive process, but with knowledge and logic at its base.

##### Figure 6

A default typeface
Have you heard the phrase ‘When in doubt, use Helvetica?’. If you have, forget it immediately. As I explained, every project needs to be treated individually, and so when you’re in doubt you should either do more research on typography and branding, or do more market research on your product. Simply selecting Helvetica when in doubt is lazy, and you’re not much of a designer by doing so. In fact, Helvetica has become so prolific that although it’s still considered a designer typeface, it’s actually increasingly becoming the amateur designer’s typeface. It comes with OS X by default and for a Windows user it’s also very easy to get, and the problem with these kind of default fonts is always that everyone has already seen them, and any amateur has access to it and will likely misuse the typeface—which further stigmatizes the typeface just like it did with Comic Sans. Besides, Helvetica is absolutely everywhere on the streets, and I simply tire from seeing it. Many designers still cling to Helvetica, but a truly professional designer won't limit himself to one typeface, or select a typeface for a project because it's his favorite. A type designer or typographer may not even want to use Helvetica at all, both because they recognize there are better typefaces around or because they want to avoid using a system default font to avoid any negative connotations.

##### Figure 7

Better typefaces

Helvetica is by no means a bad design. It could easily be improved (have you noticed the weak design of the belly of ‘a’?), but it was a wonderful typeface in the 60’s and 70’s and it’s still a nice typeface today. But have you seen other sans serif typefaces lately? I strongly suspect that if Helvetica is your favorite typeface, you simply haven’t seen a lot of professional typefaces. Typefaces like Trivia Grotesk, Voice, Bulo, Sixta, Adelle Sans, Argumentum, Supria Sans and Air Soft are certainly not for every project, but they’re very attractive grotesque typefaces. Grotesque isn’t my personal favorite style though (but personal preference shouldn’t matter much when considering a typeface for a project); I’m very fond of humanist sans typefaces and typefaces with some quirks. Typefaces like Winco, Ideal Sans, Sonus, Tabac Sans (Figure 8), Andes, Uniman and Karmina Sans are absolutely amazing to me. Speaking of quirks, I still admire the classic Gill Sans. Interestingly though, I initially hated the typeface for being so awkward and "imperfect", but as I learned more about typefaces and type design, I came to admire the typeface for the reasons I started out hating it. This makes me think sometimes typefaces are a bit like an insiders joke; only few people will truly understand what the typeface is about. This can be rather frustrating for type designers because our main target audience are graphic designers, who generally know quite a bit about typography but not about typefaces. At the same time though, that's the type designer's challenge and tool, to play with such differences in our perceptions of design. This is also why there are never enough typefaces and there will always remain a gap in the market. You would think every style and concept has been done already when it comes to typefaces, but once you really study it, you become aware to how many gaps there are yet to fill, and how many highly unique typefaces are being released each year.

##### Figure 8

Better alternatives to Helvetica
If you’re simply in love with Helvetica’s aesthetic though, there are still plenty of better alternatives and even better renditions of the Helvetica model (like Aktiv Grotesk [bottom line of figure 9], Vaud or Haas Unica for example). If you enjoy the general style of Helvetica [top line of figure 9] but you want something less antique, Univers is a great alternative. It should also be said that Helvetica is nice as a display type, but it’s very bad in long texts; it tires the eye easily. On the web Helvetica is off even worse, as it renders horribly on Windows. For the web I would much rather use Arial than Helvetica because the hinting is better, so it renders better on screen. I believe Helvetica Neue renders better on the web, but most companies seem to use Helvetica rather than Helvetica Neue.

##### Figure 9

Choose objectively
By the way, is it not strange that Arial is frowned upon for being a copy of Helvetica, while Helvetica is a copy of Akzidenz Grotesk and Akzidenz Grotesk is a copy of Scheltersche Grotesk? When people seem to speak from authority, we tend to listen to what they have to say, but in the design field a lot of opinions are often mistaken for facts, and a lot of these opinions are often based on ignorance. Arial is not the horrible typeface people make it out to be, and Helvetica is not the great typeface people make it out to be. Neither one is absolutely amazing to me, but they're both good typefaces. The reason for writing this article is not to mock Helvetica, its history or its designer, but rather to offer perspective from a professional type designer, and so you might reconsider whether to use Helvetica for your next project. Not because it's a bad typeface, but because we've seen it far too often for over 50 years, there are an impressive amount of other professional typefaces to choose from, and no typeface is perfect for each and every project. It’s nice to have a favorite typeface, but it doesn’t speak of a good designer to use a typeface because it's your favorite. Choose your typefaces as objectively as possible.

## How to design a typeface (Part 2)

Wed Apr 24, 2013, 5:32 AM

### How to design a typeface (Part 2)

##### Sans serif — Italics & weights

In the previous article of the typography series I presented some of the basic principles of type design and we started designing a geometric sans typeface such as Futura. In this article we will expand that typeface and add italics and alternate weights. Or rather, I will show you how to design italics and briefly present how you can modify your regular weight typeface to create alternate weights rather than having to re-draw the whole letter.

Before we start
If you haven’t read the last article/tutorial but you want participate in this one, you can download a pack of resources of the last article so you don’t have to design all the Roman letters to follow this tutorial. However, you’re only allowed to use that pack for this tutorial. The letters are of my own design (and they’re not even perfect anyway because the design was done rather quickly only for the purpose of these articles), so to use them as your own is copyright infringement. Besides, you could probably do a better job than I did if you have the patience, and you carefully look at the design of professional typefaces. Looking at other typefaces is probably the best way to learn.

If you already have a typeface designed, perhaps this is a good excuse to expand it with italics and a bold weight. For editorial use this is a must. Most professional typefaces feature a few weights (plus italics for each weight), ligatures, small-caps and different numeral sets. You can expect articles on those features from me in the future. You can find a link to download a pack of resources from part 2 at the end of this article.

##### Figure 1

Oblique
Before we design the italics, we will create an oblique. You might have seen this name before and recognized it as an italic typeface, but they’re quite different. An oblique is simply a slanted roman. Now you might ask yourself what the use of an oblique is, because on the surface it seems to be a more primitive italic. That might be so, but sometimes an oblique simply looks better in regard to the roman. In case of a geometric sans typeface or a grotesque (such as Helvetica and Univers) you can expect to find some obliques because the design in general tends to be quite linear, but in a humanist serif you would expect a true italic because the letter shapes are more humanistic/calligraphic, so a slanted roman looks odd because it slants humanist features which are only native to the roman. The italic design is based on cursive writing, so the structure and weight distribution tends to be too different from the roman to simply slant it and call it a finished design. In case of serif typefaces an oblique is almost unheard of. The italics in a serif typeface are of a quite drastically different design—even more so than in the humanist sans serif—so simply slanting your roman won’t do. However, in all cases it’s very smart to use your roman to create an oblique, and then modify the oblique to turn into a true italic.

So let’s take our roman (see my version from the previous article in figure 1), select it in a vector program (Adobe Illustrator is my choice) and select the Shear Tool. Double click on the Shear Tool icon and select a Shear Angle of around 7°. Make sure the Axis is set to Horizontal so the design will slant horizontally. Before we continue, I should say something about the angle of the slant though. The oblique and italic of a sans serif typeface really don’t have a severe slant, but in case of serif typefaces you can go a bit wilder. Which slant is perfect for your typeface really depends on the design and general proportions. You will probably find italics with a subtle 3° slant up to an extreme slant like 20°. A geometric sans typeface is quite minimalist and so the construction of the roman and italic will almost be the same. For that reason I wouldn’t go below a slant of 5° because it doesn’t make your oblique/italic distinct enough from the roman. For a sans serif I also wouldn’t go higher than 12°. In case of a serif typeface you could even use no slant at all (a so-called upright italic). Upright italics are usually used as a roman rather than an italic though. For reference, have a look at the upright italic Ninfa and compare that with Ninfa Serif (which was designed later after Ninfa’s success).

I’m getting off-topic, but if you applied the slant to your roman typeface, you have a finished oblique (figure 2). It’s really that simple. To change your oblique into a true roman, we actually don’t have to do much more in case of a sans serif, and particularly a geometric sans. In case of serif typefaces the italics are sometimes more work than the roman.

##### Figure 2

Italic
Now things will become more interesting. If you’ve followed the previous article, you probably have two variants of /a: a one-storey /a (figure 1 & 2, first letter) and a two-storey /a (figure 1 & 2, second letter). The two-storey /a is a roman design, while the one-storey /a is used for the italics, but since geometric typefaces are all about minimalism, you will often find a one-storey /a in the roman. I can appreciate that, but I personally like to include a two-storey /a as well. So for the roman you can use both versions, but in case of the italic things are not so clear. There doesn’t seem to be a clear rule which states you can’t use a double-storey /a for the italic, but this is not common and in fact, when I see a slanted two-storey /a I get angry. That says more about me than anything else, but a slanted two-storey /a can mean two things (with only few exceptions):
1. That the font is an oblique rather than a true italic (and often obliques are called ‘Italic’, which frustrates me further).
2. That the typeface only includes roman and you simply used faux italics. In the Character panel of Photoshop you can turn your font into bold, italic or small-caps; when the typeface does include these features, they will automatically be used if you activate bold or small-caps etc. However, if the typeface doesn’t have these fonts, Photoshop will create faux fonts. Faux italic will slant your roman (=oblique, not italic), faux bold will make your font bolder with an algorithm so it usually looks quite ugly (far worse than a real bold, anyway) and faux small-caps will simply make a capital letter smaller, which means the weight will also decrease while true small-caps feature the same weight as the capitals.

In conclusion, seeing a slanted two-storey /a makes me think of an oblique by design or an oblique generated by using faux italic, and I personally perceive this as negative. I said earlier that sometimes an oblique is preferable, but rounded letters should always be adjusted after slanting. If you do adjust it after slanting however it becomes an italic rather than oblique. As such, I perceive an oblique to be undesirable in all cases. If you have clear reasons for including a two-storey /a in your italic, by all means do it, but do adjust the letter. A good example of this would be the popular Gotham. The italics look like an oblique because no letters clearly change and both the roman and italic feature the two-storey /a, but if you slant Gotham Italic -15° and compare that with Gotham Regular, you will notice the italic has clearly been adjusted after slanting. In fact, it’s adjusted to such extent that when you slant the Italic -15° to get back to 0°, the design looks really off. So anyway, Gotham shows that it’s not bad to use an oblique design as long as you make the proper adjustments after slanting, and it also shows an italic two-storey /a is not necessarily bad, but you do need to be able to say why you made that choice; don’t let it be a decision based on ignorance. Anyway, I don’t like an italic two-storey /a for this Futura-like design, so I will remove it. So you now have a roman typeface with two versions of /a and an italic (well, it’s going to be an italic) with one version of /a.

##### Figure 3

Slanting causes undesirable effects
Okay, so the two-storey /a is gone. Since we’re designing a geometric typeface I had a look at Futura and I found out Futura features obliques (with a 10° slant, if you’re curious) rather than italics, which was to be expected considering it’s a geometric design, but I can’t help but feel it’s unfortunate. I will show you why. In figure 4 you will see three different versions of /s. The first /s is Futura Book and the second /s is Futura Book Oblique. Perhaps you think there’s nothing wrong with that /s, but if you haven’t realized this by now, type design is all about nitpicking. Besides, why be pleased with something when it can easily be improved? So anyway, what’s wrong with the oblique /s is that the slant changed the weight distribution. The weight in the bottom curve on the top left (the one going to the spine of the /s) is hanging too low and the weight in the top curve on the bottom right is placed too high. So there’s a weight build-up in the extremes of these curves, and there is also weight build-up in the endings of the terminals (the very top right and bottom left). On the right you see my modified version of the oblique /s (so it’s now an italic).

I’m not going to pretend the letter is perfect now (I will have to modify the whole typeface to get all letters perfect in relation to each other, which is obviously a waste of time considering I just want to illustrate why an oblique is not optimal), but it looks a lot better already. Weight build-up is not always a bad thing (in fact, just look at the italics of a serif typeface and you will see plenty of variety in the weight distribution), but since Futura is supposed to be monolinear (=consistent stroke weight), my /s simply fits Futura better, and Futura’s oblique /s is a bit distorted. It won’t be noticeable at text size, but at display size it’s definitely noticeable. Alright, so I’m a type designer so it’s not strange I would notice these things, so you might tell me most people won’t notice. Well, that’s not entirely true but I do understand that argument. People usually don’t notice what’s wrong with a letter or a typeface, but unconsciously they often do notice something is off. This oblique issue is not a big, noticeable issue, but if your typeface is full of these tiny flaws, unconsciously your typeface will be less comfortable to read than it could have been. That’s why I recommend to always modify your letters after slanting. Slanting predominantly tends to cause a lot of undesirable effects in the curves.

##### Figure 4

Figure 4 actually illustrates most of what needs to be done to change your oblique into a true italic. Adjusting the curves is a big part of that. I will show you in a moment what specifically you need to do to optimize your italic design, but let’s first have a look at why slanting adjusts the weight distribution in the first place.

##### Figure 5

In figure 5 you see the outlines of the letters of figure 4 and you can see how the vector points of the horizontal center of the roman /s are vertically aligned, just as in a vector circle, which has four vector points at the extremes of the shape. Letters work the same way; vector points need to be placed at the extremes. If you look at the oblique /s, you will notice that these vector points still align at an angle. That’s where the problems start. With a simple shape like a circle, you can add a slant and the shape still looks fine. If you do the same to an open circle (like a letter /o) though, the weight distribution will be altered relative to the slant. So, if you slant a letter to the right, then the outermost vector at the top will move to the right and the outermost vector at the bottom will move to the left. This means the weight at the top will move to the right, and the weight at the bottom will move to the left. Since a geometric typeface is supposed to be monolinear, you obviously have to adjust these curves so the slanted design will still be consistent in stroke weight. I think you might now better understand why I don’t appreciate obliques and why I consider them to be flawed, unfinished italics.

You might also have noticed that my version of the Futura /s is a bit narrower at the bottom than the official oblique one. I did this because the oblique /s was simply falling too far to the right. An italic will obviously always do that, but there still needs to be a certain balance in the design. I personally find it useful to see the letters as objects which adhere to the laws of gravitation, so it’s your job to make the letters as stable as possible. Anyway, in case of the oblique /s I thought the bottom terminal was extending too far past the central point of the letter, so I compensated for that. Be reminded though that by moving the bottom terminal to the right and the top terminal to the left, you’re essentially converting the letter back to roman. I didn’t modify the spine, so technically the middle part would still be at a 10° slant, while the outer shape is probably at about 8° now that I moved the bottom terminal to the right. So don’t overdo the adjustments. At the same time though, don’t be afraid to have a bit of variety in the slant in your typeface. In case of a geometric sans you do want all the verticals to slant at the same degree, but when designing italics for serif typefaces (particularly Venetian and Garalde styles) you will probably notice there is a lot of variety in the slants. The letters /a and /e for example are usually less slanted than letters with ascenders (b/d/h etc.) I will go more into detail about this in the future articles about designing a serif typeface.

So what do you need to do to adjust the curves of the slanted letters? Well, as I explained with help of figure 5, by slanting you distort the way the weight is distributed, so you simply need to “normalize” the design without changing the appearance of the slant. So, move the top vector point a bit to the left to move the weight from the right a bit more to the left (but don’t bring the vector point to the center because a bit of weight modulation is needed for the italic), move the bottom vector point a bit to the right, move the bottom right vector a bit down to get a bit of weight away from the spine, and move the top left vector point a bit up. The same principle should be applied to all rounded letters. I should perhaps prominently state that adjusting curves is not only about moving vector points, but also about adjusting the bézier curves. For example, after moving the top vector point a bit to the left, you might want to extend the right bézier curve to compensate a bit for the weight you moved from the right to the left. After all, you want the top to look balanced, not simply move the problem from the right side to the left.

Cursive
Alright, so you now know how to easily slant your roman to become an oblique, and you know how to adjust the curves to make a proper italic. You now know enough to finish the design of the italic, if indeed you’re designing a geometric sans typeface. But just like the one-storey /a, there are letters in the italic which are distinct from the roman. In case of a serif typeface every single italic letter is distinct from the roman, but for sans typefaces it’s a lot less work. First, have a look at figure 6. The first line is Scala Sans Regular, then Scala Sans Regular which I slanted 9° to create an oblique, and then Scala Sans Italic (which also seems to feature a slant of 9–10°). Now, the oblique Scala Sans doesn’t look wrong on its own. In fact, Scala Sans is a humanist sans, and so the distortion from slanting the design is less obvious because the stroke weight is not as consistent as in Futura for example. If you zoom in, you will notice that rounded letters like /s and /c actually do look off, just as the /s in Futura Oblique. But zoomed out, it’s not so bad. Actually, it might be more accurate to say it looks alright as a typeface, but it’s pretty bad as a font. Why? Because an italic in a humanist sans needs to fit the roman (as in any other style of typeface), but unlike a geometric sans typeface, the italic in a humanist sans also needs to be very distinct; in fact it needs to be of a cursive design. I hope you can see why. First off, in my eyes Scala Sans Oblique looks quite dull. Yes, the letter shapes are still attractive, but italics require a certain flow and dynamics. Scala Sans Italic does that, by changing the weight distribution, making the letters more narrow and make the design more distinct. So now, if you set a page of text in Scala Sans and you put a few words or sentences in Scala Sans Oblique, it won’t jump out of the text that much. And let’s face it, nowadays italics are predominantly used to emphasize words in a text. Small-caps and bolder weights are also used to emphasize things. Interestingly enough italics were initially separate typefaces, so a whole page would be set in italic rather than putting a few things in italic for emphasis. While the first roman typeface was released at around 1470 by Nicolas Jenson, the italic typeface was first introduced in 1501 by Aldus Manutius. The italic was probably first used within roman text for emphasis by Robert Estienne in 1532.

##### Figure 6

So which letters in the italic need to be very distinct from the roman? Well, there’s no clear answer to that. It really depends on the typeface. As we’ve seen, a geometric typeface doesn’t necessarily need distinct letters in the italics. By far the most important letter to change is the /a. Or rather, this letter is changed most commonly; this one often changes even in geometric sans typefaces while the other letters do remain the same. /e and /g are also commonly changed, though usually not in geometric sans typefaces. Geometric typefaces tend to use a monocular /g (a /g with one storey, like the italic /a) in both the roman and italic. Actually, many sans typefaces use the monocular /g, while serif typefaces feature a binocular /g (a two-storey /g) for the roman and a monocular /g for the italic. This is because the binocular /g is easier to read because the letter is more distinct from the other letters, but since the /a in italic changes to a one-storey design, a monocular /g is used for consistency.

##### Figure 7

The letters h/m/n/r/u are also often distinct from the roman in sans typefaces. Elements tend to start at a lower point (or in case of /u, a higher point) and gradually emerge from the stem, while in the roman these elements will emerge from the stem more abruptly. This is particularly useful in humanist sans typefaces. In geometric typefaces it’s uncommon because it defeats the minimalist approach.

##### Figure 8

The /f is also quite an important letter. Not only does /f belong to the most common ligatures (fi/ff/ffi), but particularly in humanist sans typefaces the /f tends to get a descender. In case of geometric sans typefaces this might not be a good idea. In case of grotesque typefaces it depends. Use your own judgment. If your typeface is very elegant an /f with a descender might be smart. The letter /e can also be of a cursive design in a humanist sans typeface.

##### Figure 9

These are the main letters which are subject to change in the italics of sans typefaces. But actually, every letter could be distinct. You can design a /k with a cursive curl rather than an arm, design a loop in the middle of /w, design an /x like two half circles, do something wacky with the tail of /y or make the top part curved like the /u rather than at a sharp angle like /v, you can give the /l a little tail or make a bend in the curves of /c (such as in Scala Sans), /e and /o. Really, there are a lot of possibilities; as long as it fits the style and you remain consistent throughout the typeface it’s at the very least worthy to try out. And again, look at other typefaces to see how they go about the italics. The more you look at other typefaces, the more you become aware of the possibilities—and what not to do.

Figure 10

Figure 11

Alternate weights
So that’s the italics for sans typefaces pretty much covered. Let’s create some alternate weights. First you need to define which weights your typeface will include. The range of weights as I know them are:
Hairline, Thin, Extra Light, Light, Book, Regular, Medium, Semibold, Bold, Extra Bold, Black, Ultra Black.

You don’t need to use every weight in the range though. The application of your typeface will affect which weights would be useful and which are unnecessary. For example, Verdana was designed for low resolution screens, so it would be ridiculous to expect the lightest weights to be part of the typeface. Not only is there virtually no need for it, but you’re probably wasting time designing a weight no one will have a use for. For a low-res screen typeface you probably only need Regular, Bold and Italic. If you want to design a so-called workhorse typeface (a typeface with lots of features and styles for advanced editorial use) then obviously you need a wider range of weights. I mostly use this range for my typefaces: Extra Light, Light, Regular, Semibold, Bold, Black (plus italics, which make a total of 12 fonts per typeface). I mostly design book typefaces, meaning they’re legible serif typefaces with lots of features, but the font family is nowhere near as big as a workhorse typeface.

Interpolation
If you intend to have a wider range than 3 weights, it’s best to use the regular weight to create the two extremes in the range. So in my case, I would change Regular into Black and Extra Light. These three weights can be used as master weights to generate the in-between weights with interpolation software. Mac users are probably lucky to be able to use Superpolator, but some font editors also include interpolation and extrapolation (creating a weight or style outside of your master weights, which is a lot messier than interpolation) features. Fontlab Studio 5 does. How this interpolation process works exactly I can’t tell you because I haven’t done it yet, but there are plenty of resources around including video tutorials. So you might want to research this yourself, and also how to turn your type design into an actual usable font. Because frankly, I have plenty of articles to write about the design, so it might take a while before I write an article about the more technical side of typefaces, if at all. But do be reminded that you don’t need to completely design each weight from scratch, nor do you have to modify the regular weight to get each other weight in the range. 3 master weights should suffice. After interpolation you still need to make adjustments, but you save a lot of time and effort anyway. By the way, interpolation is useful for more than just generating weights. If you design Regular and Condensed, you can use that as master styles to generate the in-between width, which would be Narrow. You can interpolate weights, widths and even contrast. Here’s a little taste of what you can do with interpolation: www.youtube.com/watch?v=mfUo6i…

Bold
Forgive my endless talking. We will get back to the typeface now. So you probably have a typeface consisting of Regular and Italic about now. Let’s create Bold. First we need to reference the weight of a bold typeface. In the first article I told you to type something with a professional typeface at 400pt. If you used a different pt size, make sure you use the same now. If you don’t know the pt size of your typeface anymore, just select the regular weight of a random professional typeface and resize it until the /l is as wide as the /l in your own typeface. Make sure you only look at the width of the letter, because the height will always differ between typefaces. Now that you have your pt size, change the font you’re referencing to Bold. Go to Object  > Expand to create outlines of the typeface so it’s easier to handle. Now take your own typeface and make your /l just as wide as the bold /l of the typeface you’re referencing. Make sure you only change the width of your /l and don’t alter the length. You can remove the typeface you’re referencing now because you have the weight in your typeface defined. Now all that is left to do is make all letters the same weight as /l. Look at my Regular, Italic and Bold /l in figure 12.

##### Figure 12

I obviously started with /l because it’s the easiest letter to change the weight of. All verticals are as easy to adjust though. Copy your /l and give it a color. Now place it on a layer below your typeface. You can use this letter as a template for the others. First, put the /l right behind your /i. You might want to lock the layer of /l for a moment so you don’t accidentally change the /l when you adjust the /i. Now select the Direct Selection Tool (the white arrow) and select the two left or right vector points of the stem of /I and drag it to the width of /l. Hold the Shift key while dragging so the vector points are only repositioned horizontally and don’t accidentally move up or down. Repeat this process to the stem of /i, and then resize the tittle (the dot of /i) with the Scale Tool until it’s in proportion with the stem again. Remember, the tittle needs to be thicker than the stem because rounded shapes look optically smaller. Just use your own judgment. I chose 158% and increased the horizontal bézier curves by 0.5pt on both sides (Go to Edit > Preferences > General and select a Keyboard Increment of 0.5pt and now simply select a bézier curve and use your arrow keys to increase the curve) and I moved the left and right vector one increment to the outside. This gives the tittle a subtle horizontal oval shape. Look at figure 13 to see the process to change the regular /I into a bold one.

##### Figure 13

Tittles
You can actually go a bit crazy with the tittles if you want. It’s often nice to have some subtle differences between the Regular and Bold weight. Also, if you give Regular rounded tittles and Bold or Black squared tittles, you can then interpolate the weights in between and the tittles will slowly change from circle to square as the weight of the font increases. Or, you can make your tittle a bit squared and make it rotate throughout the weights. I haven’t done anything like this before, but I’m very impressed with the tittles in the Vesper typeface.

Oil
Now select the inner four vector points of /o, select the Scale Tool and scale the counter down (uniformly) by about 40% so the vertical parts of /o are the same width as /l or a tiny bit thicker. Now select the inner top vector and move it up a bit and do the opposite to the inner bottom vector so the horizontal parts become skinnier. Move the inner and outer left and right vectors outward a bit to regain a bit of the circular shape. Increase the bézier curves on the inside a bit, and perhaps on the outside if necessary. You should now have something similar to figure 14.

##### Figure 14

d is for difficult
Now things will slowly become a bit more difficult as we’re going to add weight to letters which combine circular shapes with rectangular shapes, like b/d/p/q. Essentially you will repeat the same process as with /o and /l, but with a bit of a twist. After you know the method to do this though, it shouldn’t pose much of a problem. First, select the inner vectors of /d and scale it down to 40% just like you did with /o. Now, press Ctrl + R to activate the rulers and drag a horizontal guideline from the top ruler to the inner top of /o and to the inner bottom. Select both guidelines and place them in the Guidelines layer you created in the previous article (or create one now). Lock the layer. Now align the inner vectors of /d with the guidelines. Move the inner left and right vectors outwards and move the left vector to the left, just like you did with /o. Now copy /o, give it a different color and superimpose it on /d to check if the counters are roughly the same shape. If they’re not, adjust the counter of /d. You should now have a /d which looks something like figure 15.1. Now copy /l and give it a different color and position it behind the stem of /d; a bit more to the right rather than centered (15.2). Merge the two shapes with Pathfinder (15.3). Delete the extra vector points in the stem if there are some. Now move the vector points in the indents between the stem and the bowl more inwards (vertically) to increase the indents. Adjust the bézier curves accordingly so the transition is smooth. Move the left bottom vector of the stem and the vector in the bottom indent a bit to the right to increase the negative space in between the stem and the bowl. Don’t overdo this for a geometric sans typeface though. You should now have something like 15.4.

Figure 15

Now, the cool thing about a geometric sans typeface is that you can just reflect the /d horizontally to get something like a /b. Do this, and move the top vector of the bowl of /b a bit to the right. You may also want to move the bottom vector to the left a bit. Don’t overdo this because a geometric typeface requires a certain symmetry. It’s predominantly the humanist sans typefaces which feature weight modulation in the curves.

##### Figure 16

Even for a geometric sans typeface though it’s smart to reference the letters /b and /d of a serif typeface because then you know what shape the bowl of /b and /d should roughly be (very subtly in this case) and how the weight is distributed in the curves. Have a look at figure 17, and apply the same principles to the letter /p and /q. When you rotate /b and /d to create /p and /q, first reduce the length of the descenders a little bit. They should be in harmony with the ascenders, but shortening the descenders a little bit looks better to me. Then make sure the top of the stems is the same width as in m/n/r and finally adjust the curves while referencing figure 17. If you get stuck, look at other (geometric) sans typefaces, or download the resource pack at the end of this article so you can have a look at the outlines of the typeface I did for this tutorial. Mind you, it’s far from perfect so I would actually advice to have a look at the outlines of popular typefaces like Futura, Gotham, Univers or Franklin Gothic. Also have a look at the unusual shapes of Gill Sans.

##### Figure 17

Ahmnur
Now we will do an easy letter again before I will show you how to go about h/m/n/u; we will create the one-storey /a. Simply copy the letter /d and turn the Guidelines layer on. Select the Rectangle Tool and create a rectangle over the ascender of /d and align it to the x-height (the line at the top of the stem of /i). Now subtract the two shapes with the Pathfinder panel. Make the top of the stem of /a a bit smaller and make the top indent a bit deeper. That’s your one-storey /a finished.

Copy the bold /I and the regular /n. Remove the tittle from /I and use the stem to add to the stem of /n. Look at figure 18 to see the process. You can either add the shapes together, or you use the stem as a template (like we did with /l) and simply drag the two left vector points of the stem of /n to the left until it aligns with the width of /i. Do the same on the right side; this works better with the template method so you don’t ruin the curve on the inside of the arc. Once the right side is as thick as the stem, condense the letter until it has a harmonious rhythm with the counters of the other letters. Now lower the curve on the inside of the arc and adjust the bézier curves accordingly. I also moved the top vector in the arc to the right to make sure the curve towards the stem is smooth and the weight is consistent at the right side.

Figure 18

Now copy the bold /n and turn it upside down to get a /u. Align it with the baseline again and condense the letter. This is necessary because the /u always looks optically wider than /n. The curve in /u also appears weaker, so I usually move the bottom vector of the arc to the left and extend the left vertical bézier to make the curve a bit stronger. That’s the /u done.

Copy /n again and copy /l. Superimpose /l on the stem of /n and merge them together with Pathfinder. Remove unnecessary vectors if they appeared after merging, and make sure the verticals are still absolutely vertical. And that’s how we make /h.

The /m is a bit trickier. First, copy /n and condense the letter a bit. Now copy that one again and remove the stem. You know have something like figure 19 (left). Align the /n without the stem with the one with a /stem like in the second step of figure 19. Merge the two together and remove any vector points which might have popped up when merging. Now copy this flawed /m and give one of the two a color and lock the layer. Superimpose the black one on the colored one and remove the vector in the indent where the two arcs connect. Now drag the other vector in that area down so you can see the colored /m underneath. Now reposition the vector in the center and cover the colored /m accurately but close the indent. Now drag a horizontal guideline to the indent between the left arc and the stem. Align the vector where the two arcs meet to the guideline so both indents are at the same height. Now adjust the bézier curves and make the letter a bit more condensed if necessary. The letter /m should never be the size of a double /n; it’s always more condensed.

Figure 19

The letter /r can be made from the left part of /n by cutting away most of the right side of the arc. Just repeat the process of /n and make sure the letter shapes remain consistent with the regular weight.

As
Alright, we’re nearing the end of this article/tutorial. I kept the most difficult letters for last, though this tutorial typeface doesn’t feature a /g and a binocular /g is probably the most difficult letter to design and to adjust. In essence you would be using the exact same principles and methods we’ve already covered.

For the /s I thought it would be easiest if I re-draw it with the Pen Tool, just like we did originally with the regular /s in the previous article. Unfortunately the stroke needs to be 72pt now to match the weight of the other letters, and if I keep the /s the same height as the regular /s the two openings will get completely filled. It’s harder to get that open again, so I will introduce a new method. Copy the regular /s and give the letter a color. Add a 30pt stroke to the letter in the same color.

Figure 20

We will now use this letter as a template. Lock the layer and copy another /s onto the colored one. You might have noticed the colored letter goes beyond the top and bottom guidelines. We will use the whole colored letter as a template except for the area at the top and bottom. So just start dragging vector points out and align them with the colored /s, but keep the top and bottom aligned with the guidelines. Other than that I don’t know what to tell you. It took me a while to get the /s reasonably well looking, so it’s just a matter of moving vector points around and changing the béziers until you’re happy with the result. I’m not happy with mine, but I will leave it for now.

Now on to the last letter of this tutorial: the two-storey /a. First select the vectors on the interior of the bowl and select the Scale Tool. Scale the counter down to around 60%. Copy the bold /I and remove the tittle. Make the stem a bit shorter and superimpose it on the stem of /a. Merge the shapes and remove the vector at the very right. Now re-do a few bézier curves and move vector points around until the design is to your satisfaction and the weight is consistent with the other letters. Again, look at outlines of other typefaces if you get stuck.

Figure 21

More weights and styles
Now we’re able to create a bolder weight, a thin weight shouldn’t be much of a problem. It’s essentially the same as making a bolder weight, only you need to scale counters up instead of down and you need to move vectors inwards instead of outwards. When creating a thin weight based on a regular weight it’s handy to use a specific Keyboard Increment. To save time, you could select all the left vector points of the left stems and press arrow right so all stems become skinnier.

And finally, when you have your bold weight finished, it might be worth it to slant it by the same degree as you slanted the regular weight before and repeat the process of creating italics from obliques and you have a typeface consisting of 4 fonts already: Regular, Italic, Bold and Bold Italic.

Figure 22

Lighter weights
If you want to design lighter weights, it's the same principle as a bolder weight but the other way around. Just reference a professional typeface again for the weight, increase the size of the counters and reduce the general line weight consistently. If you want to design a hairline version you might want to trace your Regular design with the pen tool (same technique as we used for the /s) and select a stroke weight of 1pt or so. You might also want to try interpolation between a hairline font and a bolder weight to see what it will produce. You might get some pleasant and unexpected results. Also remember that the contrast in a typeface doesn't necessarily need to remain the same throughout the weights. I tend to give lighter weights less contrast between thick and thin strokes, and darker weights more contrast. Just like with the rotating tittles, you can vary a bit with your design throughout the weights as long as you don't change the dimensions or the general style.

Future articles
So that’s it for now. Let me know if there are any questions, and suggestions for future articles are always welcome. This is what I have planned:
• Designing a serif typeface
• How to create true small-caps

Let me know if there is a specific letter you want me to make a tutorial of. I didn’t think covering capitals is necessary since they’re based on the same principles as lowercase, but do let me know if I can clear anything up with a short tutorial. In fact, it might be fun to take requests for mini-tutorials on how to design a specific letter in a specific style.

## How to design a typeface (Part 1)

Mon Apr 1, 2013, 7:38 AM

### How to design a typeface (Part 1)

##### Sans serif — Roman

In this article/tutorial I will show you how to design a typeface. I will cover the basics and show you how to design advanced letters as well. There is so much I can talk about though, so if people like this article I might expand on this with future articles and get more into detail. I will only cover the design though, so you will have to research how to do the spacing and programming and turn your typeface into a usable font—or wait until I might do an article on that as well, but I don’t have such an article planned anytime soon.

What will we design?
I will first show you how to design some basic letters for a geometric sans typeface like Futura, because it utilizes geometric shapes, which is always my starting point for every typeface regardless of style. Once you understand the basics, you can use those principles to design the other letters. If you don’t know how a specific letter should look, just look at other typefaces for reference. Just as with any other art form, you learn a lot by looking at other people’s works. Don’t ever modify a typeface though! Always design your letters from scratch, or you might get into legal issues. I can recommend however to select a certain typeface, type some letters and expand the Text Field (Object > Expand) to get the contours of the letters and have a look at how vectors and bézier curves are used. That’s how I learned.

Before I get to the tutorial I have to offer you a bit of background information. If you’re only interested in the actual process of designing the font, just skip to the chapter appropriately called “Let’s get started”.

Before you continue, let me state here that I will provide a link to download a zip file featuring all the figures used in this article as an AI vector file and a PDF at the end of this tutorial. So, if you get stuck somewhere, just open the AI file and inspect the outlines of the letters, or just continue with one of my letters. As long as you don’t use the letters in your own typeface feel free to modify and experiment with it.

Software
Now, if you want to create a typeface from start to finish (meaning design, spacing and programming), Fontlab Studio is probably the best option to do all of that. I’m not comfortable with designing in Fontlab Studio though. Many type designers actually do the design in Illustrator. It’s a great program for designing typefaces, as you can zoom in more than in Fontlab Studio, you can adjust your art board so you can have all letters in one spot, and it’s easy to add guidelines and place them in a separate layer so you can turn the whole group on or off. Any vector program will suffice though.

Wait!/Weight
Before you start designing, you need to establish the weight of your typeface (assuming you’ve already thought about what kind of style you want to design and what the application of that typeface would be). Are you going to design a typeface consisting of several weights, or in a single weight? In case of the latter, which weight should it be? You have to consider the application of your typeface. If you design a really wacky font, it may only be useful to use as a display font (meaning for use at bigger sizes, like in titles and headings), and as such it may not be important to have more than one weight. I mostly design serious typefaces, meaning the intended use is for books, magazines, logos, signage; virtually anything. For such a typeface it’s almost required to include more than one weight, because the end user will want to put some texts in bold or italic, or utilize other features which remain absent in a single weight font. Don’t let this discourage you though! There are still plenty of single weight typefaces with plenty of use. I released the single weight typeface Dion (though I’m currently working on Dion Pro, which offers more features, has improved letter shapes and includes Bold and Italic) and I have a single weight (and capitals only) typeface called Crouton Grotesque in production.

I always start with the Regular weight to establish the style and proportions. I then copy the Regular weight and modify it to get the two extreme weights in the family (assuming the typeface consists of more than one weight). So, I usually design either Thin, Light or Extra Light as the lightest weight (I haven’t done Hairline yet; Thin seems light enough for the typefaces I’ve done so far) and Black as the darkest weight (I’ve never done Ultra Black). These three weights will function as masters for interpolation. Interpolation is when you take two weights and let certain software generate the weights in between these masters. You can also use interpolation for width and contrast. For example, you can design a Regular font and a Regular Condensed font and generate each width in between, and select one of them to become Regular Narrow.

For this article I will show you how to design a typeface in the Regular weight. For the other weights the same principles are involved, though in the darkest weight you will have to pay more attention to keeping the counters (the negative space inside letters) sufficiently big so the counters are still visible if you use the font at text size (9–12pt generally). If you design a typeface with more than one weight, you should make sure the proportions and style remain consistent throughout the whole typeface. If you have no idea how to do this, I might write an article on changing your font from one weight into another if there is sufficient interest.

For reference, let me name all the weights I know from lightest to darkest (but do note that a typeface rarely ever includes ALL these weights; usually 6 or 7):
Hairline, Thin, Extra Light, Light, Book, Regular, Medium, Semibold, Bold, Extra Bold, Black, Ultra Black.

I personally mostly use this range: Extra Light, Light, Regular, Semibold, Bold, Black.

Let’s get started
It’s important to note that not all typefaces have the same weight, not even when it only regards the Regular weight. So there doesn’t seem to be a specific weight assigned to Regular, but obviously a really fat typeface shouldn’t be called ‘Regular’. It might be best to base your Regular weight on the weight of a professional typeface. So, select the Type Tool and type /i (without the slash; I only use that instead of apostrophes when mentioning letters because it’s less obtrusive). Select a pt size of around 400. You want to design your typeface pretty big so you can still zoom in a lot to adjust small details. My most recent typeface Aghari (still under heavy development) is being designed at 800pt because the letterforms are quite advanced so I need to refine the details so much more than a simple geometric sans typeface like Futura for example. By the way, some typefaces don’t have a Regular weight but instead use Book (which is usually a tad lighter than Regular) or Medium (which is slightly bolder than Medium).

Now that you have a reference for the weight, select the Rectangle Tool and create a rectangle of the same width and height of /I (excluding the dot). Now select the Ellipse tool and create a perfect circle (hold down the Shift key as you drag). The circle needs to be a bit bigger than the rectangle, and it should stick out both at the bottom and at the top. This is necessary because one of the basic principles of type design is that rounded shapes appear smaller than rectangles, so you need to compensate for this by making rounded shapes slightly bigger.

Let’s create actual letters. First go to Edit > Preferences > General and change the Keyboard Increment to 2pt. This changes how much a vector point moves if you select it and press one of the arrow keys. You will likely change the Keyboard Increment frequently while designing a typeface, depending on what you use it for. For example, modifying a curve requires a smaller Increment value than when I use it to change the Regular weight into Bold. Keep track of which Increment values you use for what. If for example you make the /I one Increment smaller with a 2pt value and you make the /b one Increment smaller with a 4pt value, your weight between these two letters is obviously not consistent anymore. If at a later stage you feel like you might have ruined some of the consistencies in the proportions (it happens to me too at times), just select one letter as the base letter and compare the other letters to the base letter (by superimposing) and modify accordingly.

Select the Direct Selection Tool (the white arrow) and select the left vector point of the circle and move it a few Increments to the right. Select the right vector point and move it a few Increments to the left. You now have a condensed circle. The reason why you should use this method rather than directly create an oval shape with the Ellipse Tool is because an oval would be thinner at the top and bottom. It works better if the shape is a bit bulkier.

##### Figure 1

Now copy the oval shape and put it in the exact same spot as the first oval. So, two shapes superimposed on the same spot. Change the color of the top oval to white and select the Scale Tool. Turn Preview on and go for a Uniform percentage of around 65%. Select both ovals and select Minus Front in the Pathfinder panel. If you don’t see the panel, go to Window > Pathfinder. You now cut the white shape out of the black shape, and in effect you have a badly designed /o.

##### Figure 2

Horizontal strokes
So why would I want you to create a badly designed /o? Well, it’s not an /o just yet. One of the basic principles in type design is that horizontal strokes appear thicker than vertical strokes, so you need to compensate for that. I would usually tell you to select the top inner vector point and move it around 1 Increment up, and select the bottom inner vector point and move it the same amount down, but if you resized the white oval to 65% in the previous step, your /o is thinner than your /i. So instead, select the left inner vector and move it around 2 Increments to the left, and do the opposite to the inner right vector. The letter is still too thin compared to /I, so move the outer left and right vectors around 1 Increment outwards. I say “around” because this is not an exact science. Just modify until it looks right.

##### Figure 3

I think the /o looks pretty good. Now you have this basic letter you really have to decide which direction you want to take your typeface into. If you move the top and bottom inner vectors outwards, the left and right inner vectors inwards or the left and right outer vectors outwards, you will adjust the contrast of your typeface and it starts to look like a serif typeface. The more you raise the contrast, the more you need to adjust the bézier curves. Do this by selecting the inner top curves with the white arrow and move up. You might want to use a smaller Increment size for this, like 1pt. Do the opposite to the inner bottom curves. You now have something that starts to look like a Transitional typeface like Baskerville. Look at my result in figure 4.

##### Figure 4

Raise the contrast even more to get a Didone typeface like Didot and Bodoni. If you go back to the shape of figure 3 and move the left and right outer and inner vectors inwards, you condense the letter and when you select the inner and outer curves and move the bézier curves outwards, you make the letter more squared. If you do that, you might end up with something like DIN. Look at figure 5 for my DIN-like /o.

##### Figure 5

Ahilno
Alright, so that’s some of the basics covered and we designed our first letter. Let’s make more letters. Right now we can still take our typeface into every direction. We can adjust weight, width, style etc. But now we will start defining some things. In order to define our style and proportions, we will design a/h/i/l/n/o. Ahilno!

So why these letters? Well, the /o I don’t usually design first, but it was handy to design it first for the sake of this tutorial as it covers a few basic principles in type design. A letter like /n is technically more advanced. I first design the /I to get the weight and x-height (the height of the lowercase letters without ascenders or descenders). I then use the /I and make it taller to get the /l. The height of the /l defines the ascender height, which is a bit higher than the capital height. I then design the /n before I combine /l and /n to get /h. A lot of letters can be designed by using elements from other letters. You can design a one-storey /a (like in Futura) by using /I and /o and do some modifications. For a two-storey /a (like in most typefaces except for geometric sans typefaces like Futura), there isn’t really a letter you can reuse except the /I to get the stem of the /a. The /a is a very common letter in languages, and so a two-storey /a helps define the style of your typeface. Because it’s such a common letter you might want to spend some extra time on it and give it character. Don’t do anything wacky though, because this can be very obtrusive. Also, consistency is key, so if you give the /a a sharp serif, I expect the other letters to feature sharp serifs as well.

So, the letters in “ahiln” I design first because they define the style and dimensions. The letter /o might actually be a useful letter to design in the beginning because it defines a secondary width. Remember how rounded letters need to be bigger to look optically the same size as letters with a straight top? Well, that also counts for the width. Your /o will be slightly wider than your /n, and other rounded letters like b/d/p/q will likely resemble the width of /o. Don’t look too much at the general width of the letters though; the negative space is actually more important in this case. If the negative space inside /o equals /n, you know you’ve chosen a proper width for a consistent rhythm.

With the letters of “ahilno” you define everything except for the descender height, capital height and diacritic (éèêë etc.) height. I usually define these at a later stage. In fact, in most cases I won’t start designing the capital letters until I have at least half of the lowercase letters done. It’s much easier to define the style of your typeface with lowercase letters anyway.

The descender height should resemble the ascender height (but it can be slightly shorter than the ascenders), the capital height should be underneath ascender height and you will have to judge for yourself how high to position diacritics. Don’t get too caught up in defining all these heights though. I usually adjust these values at least 10 times during the design process of my typefaces. Same goes for the x-height, width and in some cases even the weight. You can adjust all the ascenders at once, but if you change the weight you will have to modify each letter individually—which is quite some work, so the weight and x-height are better defined at the beginning and not changed later.

Making oli
That’s enough background information for now. Let’s make some other letters. Select the rectangle you made at the beginning of this tutorial. Make a copy, because you will be using this rectangle frequently. Now that you roughly have the width defined with /o it’s easier to guess roughly how high your ascenders will be. As I mentioned before though, this isn’t an exact science and the ascender height will likely be subject to change. But for now, let’s extend the rectangle to make the letter /l. Now copy the first rectangle again. Select the Ellipse Tool and make a perfect circle which is slightly wider than the rectangle. Resize the circle to match the weight of the rectangle (not the width!). Center the tittle (that’s what the dot on /I and /j is called) horizontally on the rectangle, and position the tittle vertically somewhere between the lowercase- and ascender height. Now, it’s important to note that tittles and dots look like perfect circles, but they rarely are. The shape of these characters depend on the style of the typeface. A serif typeface could feature quite an unusual shape. Have a look at figure 6 for two examples (left: Swift / right: Brioso Pro).

##### Figure 6

For a sans serif typeface the shape of the tittles and dots will be a lot more conservative—particularly in case of a geometric typeface. Go to Edit > Preferences > General and define a Keyboard Increment of 0.5pt. Now select the top and bottom left horizontal bézier curves and move one increment to the left. Do the opposite to the right side. Now you should have a slightly more bulky looking circle for the tittle. If you want the stem and tittle to stay together, group them by selecting both shapes and press Ctrl + G (Windows) or Command + G (Mac). If you want to ungroup again, do the same but also press Shift. If you did everything right, you should now have 3 letters which look roughly like my result in figure 7.

##### Figure 7

Making nhmu
Alright, so thus far we made some of the easiest characters. Well, easy for a sans serif. A serif is a bit more difficult, and particularly the /o can be deceivingly difficult if you have a diagonal stress (weight distribution according to a diagonal axis). Have a look at figure 8 to see the three ways of distributing the weight.

##### Figure 8

Now make two copies of the first rectangle you made and place them by a little distance apart to resemble the negative space in /o (roughly like in figure 9, but adjustments can be made later).

##### Figure 9

Select one of the rectangles and make another copy. Rotate it 45° so the rectangle is positioned horizontally. Shorten it a little bit and position it between the two vertical rectangles near the top, like in figure 10. Select all three shapes and click on Unite in the Pathfinder panel.

##### Figure 10

I now came to a point where it’s rather complicated to explain what to do, so I will offer you a figure with multiple steps so you can create your letter /n. But first, let’s add a few guidelines. Create a new layer and call it “guidelines”. Press Ctrl + R (Windows) or Command + R (Mac) to show the rulers on the top and left of the canvas and drag guidelines from the ruler to the top and bottom of your first rectangle, and do the same to the top and bottom of your /o and the top of your /l. Lock the guidelines layer so you don’t accidentally move any of the guidelines. You should now have something like in figure 11.

##### Figure 11

Now let’s go through a few steps to finish the /n. Look at figure 12 while following these steps.

##### Figure 12

1. Select the shape with the two verticals and one horizontal stroke. Counting from the right and the top, remove the second vector point to get a shape like in 12.1.
2. Move the most right vector point down, move the second vector point from the right a bit to the left and align it with the overshoot  guideline (the one which aligns to /o). Add a vector point to the bottom part of the horizontal stroke and move it up. You know have a shape resembling an /n made without curves.
3. Select the Convert Anchor Point Tool (one of the options in the Pen Tool panel) and drag each vector point while holding Shift to create horizontal bézier curves. Do this only to the 4 most right vector points. You now have something similar to 12.3.
4. Now move the two most right vectors down and increase the length of the bézier curves. Move the vector on the inner horizontal up. Move the two vector points touching the stem (the left vertical stroke) down; the upper one more so than the bottom one. These two vector points will define at what position the arc starts, and the top vector of the two defines the contrast. The curved stroke is called a “shoulder”; I will refer to it as such from now on. Increase the length of the left bézier curve of the arc. You now have a letter /n, though it’s anything but refined.
5. Move the top right vector of the stem a bit to the left. Select the Convert Anchor Point Tool and define the angle of the top vector near the stem (at the beginning of the shoulder). Click on the left anchor of the bézier curve to remove that side. Do the same to the bottom vector at the start of the shoulder, but removing the left anchor is optional. If you remove it, you will have an abrupt angle between the stem and the shoulder. If you don’t remove it and align the anchors vertically, you will have a smooth transition between the stem and the shoulder. Refine the other vector points and bézier curves. There is no exact method here, so just have a look at the outlines of other typefaces. Do keep the contrast and style consistent with your other letters though. A final thing worth noting is that the weight in the shoulder is usually distributed more to the right to give some space in the area between the top of the stem and the shoulder. Oh, and this has nothing to do with the letter /n in particular, but remember to NEVER let an anchor extend beyond another anchor. Look at figure 13 to see what I mean.

##### Figure 13

Now that you’ve designed /n, it’s easy to make /h. Simply copy /l, superimpose it on the stem of /n and Unite it with pathfinder. You can also design the /m using /n, though this is a bit trickier. First, copy /n. Now, select the right vertical (all four vector points) and move it about 4pt to the left. Now also include the two vector points in the middle of the letter and move the whole right side about 4pt to the left. Shorten the anchors in the arc by about 1pt on both sides. Adjust the top of the shoulder, and you’ve got a more narrow /n. Copy the narrow /n and remove the stem. Attach the /n without a stem to the right side of the narrow /n and Unite with Pathfinder. Fill the gap at the top center of the proto-m and refine the curves to match /n and /h. You now have a proper /m.

Now let’s make another easy letter. Copy /n and rotate it 90° to get a /u. Move the letter down until the arc aligns to the bottom overshoot, the bottom of the stem aligns with the baseline and the top of the stem aligns with the lowercase line. As you might notice though, /u looks optically wider than /n. Simply compensate for this by narrowing the letter by about 1pt on each side of the vector in the middle (the same way as we did with /n when we were designing /m) and reduce the anchors by about 0.5pt on both sides. You should now have something like in figure 14.

##### Figure 14

Making a…
So we created a few extra characters, but we have our basic “ahilno” except for the /a. This is the most advanced letter yet. Well, the two-storey /a is, anyway. For a geometric sans typeface like Futura it may be preferable to design a one-story /a. I will show you how to design both of them.
1. Let’s design the one-story /a first. Copy the basic rectangle and copy your /o. Select the most right vector and move it inwards about 12pt. Select the right inner vector and move it to the right about 4pt. Increase the vertical anchors on the inner right by 2pt each. Now select the Rectangle Tool and create a rectangle about half the size of your basic rectangle. Superimpose them and select Minus Front in the Pathfinder panel. You will now have something similar to figure 15.1.
2. Superimpose the distorted /o with the rectangle shape. I cut a shape out of the rectangle so that I can bring the rectangle closer to the counter of /o without covering part of the counter with a straight vertical shape. Bring the two shapes close enough so that it looks like the counter takes a little bit out of the stem. Now Unite the two shapes with Pathfinder. You should have something similar to figure 15.2.
3. Move the bottom and top vectors in the middle of the bowl about 2pt to the left. Move the top vector where the bowl connects with the stem about 12pt down. Do the opposite to the bottom vector. Now adjust the bézier curves so the bowl smoothly slopes inwards towards the stem. Check if the indent at the bottom roughly matches the one in /u; the indent shouldn’t be much higher or lower. Move the whole stem about 2pt to the right. After refining and adjusting the weight distribution and width a little bit, you should have something similar to /a in figure 15.3.

##### Figure 15

Now that the one-story /a is done, you can easily design /d. Copy /a and /l and superimpose /l with the stem of /a and Unite the shapes with Pathfinder, just like we did with /n and /l to create /h. Don’t forget to cut out a part of the /l before you merge it with /a though (like we did with the stem of /a), otherwise it will cover part of the counter. After merging the two shapes, don’t forget to remove any unnecessary vector you might discover in the stem. These extra vectors sometimes appear when merging shapes if they didn’t align to the micrometer exactly—which they usually don’t. Now your /d is finished, copy and reflect it to make a /b. It’s very important to note here that just like /n compared to /u, a rotated or reflected letter still needs adjustments before it becomes an optimally designed letter. For example, I might move the top serif of the bowl 2pt to the right and increase the length of the top right anchor of the bowl.

##### Figure 16

Let’s design the two-story /a now. Some typefaces feature both versions of the /a so you can choose. It’s not always the best option to give the end user the choice between different characters, but you might be able to judge this best. Sometimes in type design it’s not so much what the end user would like, but how you intended the typeface to be.
1. Copy your one-story /a. Remove the two vectors which form the indent at the top. Select the Rectangle Tool and create a horizontal stroke roughly the same weight as the letters. Superimpose it on the left side of the bowl and click on Minus Front in the Pathfinder panel. You now have something like figure 17.1.
2. Create another horizontal stroke and superimpose it on the top of the bottom part of the bowl and the stem, so you essentially create a new bowl. Merge the shapes and remove the unnecessary vector point . You should now have something similar to figure 17.2
3. Add an extra vector to the inner top of the bowl and move the other vector points around to achieve something similar to figure 17.3.
4. Use the Convert Anchor Point Tool while holding Shift to add curves and move the vector points around some more. You now have a basic shape for /a which you can start refining. Look at figure 17.4 for my flawed design.
5. For a geometric sans typeface you might want to keep the bowl and stem minimalist. For a grotesque or humanist sans typeface you can play with the contrast of the bowl, the weight distribution, the slope of the curve towards the stem and you can even add a spur to the bottom of the stem. Look at figure 17.5 for the result of my geometric /a, and at figure 17.6 for my humanist /a. 17.6 is definitely not fitting for a Futura-like typeface, but just to show some of the possibilities.

##### Figure 17

A final character
So we have “ahilno” completely covered now, and in the process we went through the basic type principles. You now probably know enough to design the other characters as well. If you get stuck, just reference other typefaces. You learn a lot from inspecting professional designs.

There are many more principles to consider when designing a professional typeface, but for now this should be enough to get you going. Before I end this article, let me show you how to design the letter /s, as it seems this is the character most beginners have trouble with. There are several ways to design this letter, but whichever approach you use, it really doesn’t matter as long as you consider the basic principles of type design and refine the result. You can first draw the letter with the Pencil Tool and Expand the outlines and then refine it, you can connect two C-shapes and remove the kinks from the center where the two shapes connect and refine the rest, or you can use the method I’ve been using in this article. Because of the amount of curves in the letter /s I actually wouldn’t recommend using the method I’ve been using so far.
1. For this letter I will use the Pen Tool with a Stroke. To match the weight of the other letters, select a pt size of 38 for the stroke. Now click and place 6 vector points at each extreme end of the shape of the letter /s. Look at figure 18.1 to see where to place the vector points. Either make every anchor curves while you place the vectors, or first place the vectors and then add curves with the Convert Anchor Point Tool. After adding curves you should have a flawed looking /s such as in figure 18.2.
2. If you want the middle part of the spine of /s to remain straight, add another vector in the center and extend the anchor points. Now move the top and bottom vectors so that the stroke perfectly aligns with the overshoot guidelines. Move the other strokes around until you’re happy with the basic construction of the /s. See my result in figure 18.3.
3. Expand the strokes so they become outlines. Make sure the horizontal strokes are thinner than the vertical strokes, and further refine the character. You should end up with something like figure 18.4.

##### Figure 18

And here are all the characters of our geometric typeface together:

##### Figure 19

Future articles
So that’s it for now. I was going to show how to design serifs as well, but this article would become way too long if I have to cover that now as well—particularly since I would have to introduce a few new principles, and I’m sure from there I would want to expand to the next. So let’s do this one style at a time. Future articles will be about:
• Designing a serif typeface
• How to expand your typeface with bolder or lighter weights

If you want to know how to design capital letters, I will consider writing an article/tutorial covering that as well, though in general they’re easier to do than lowercase letters so if you know the basic type principles you should be able to manage. Also let me know if you have any suggestions for future articles.

## History of Roman typefaces

Tue Dec 18, 2012, 3:47 AM

### History of Roman typefaces

Typefaces are our instruments to construct words and sentences. Of course this very article couldn’t be written without type other than writing by hand and scanning it in, but I wouldn’t know how to save the file or how to access the website to upload it to if I had no access to typefaces. Of course I don’t have to say where type can be found; it’s absolutely everywhere. However, most people don’t consider where typefaces come from. Most of my life and even the first 5 years or so of my design career I was absolutely ignorant of where typefaces came from. I mean, they were just "there" on the computer and I never considered someone actually had to make typefaces for us to use—letter by letter.  In this article I will discuss the history of Roman typefaces; how it progressed during the ages, how each style can be recognized and how to select typefaces consciously and logically rather than by personal taste alone.

The printing press
While the first movable type (the system of printing and typography that uses movable components to reproduce the elements of a document) was developed by Bi Sheng in China around 1040, the German  Johannes Gutenberg was the first to use movable type in Europe in around 1439. In that time Gutenberg was involved in a financial misadventure making polished metal mirrors for sale to pilgrims in Aachen, Germany—which supposedly captured the holy light from religious relics. That year the city was planning  to exhibit its collection of relics from Emperor Charlemagne but due to a flood the event was delayed by one year and the investors ended up losing money. In order to satisfy the investors Gutenberg promised to share a "secret" which is widely speculated to have been the idea of printing with movable type. Legend has it that the idea came to him "like a ray of light". In 1440 Gutenberg perfected and unveiled his secret of print and released an essay entitled ‘Kunst und Aventur’ (art and enterprise). After borrowing money to fund his project and a whopping 10 years later the first printing press was finally operational. In 1455 Gutenberg completed his 42-line Bible, the legendary Gutenberg Bible. The bible was printed in a blackletter typeface, which was a script commonly used during the Middle Ages. The specific style of blackletter for the Gutenberg Bible is a form of Textura called ‘Donatus-Kalender’.

Blackletter
The blackletter, also commonly known as Gothic script or Gothic miniscule, was a script used throughout Western Europe from approximately 1150 to well into the 17th century but continued to be used in Germany until the 20th century. The blackletter—as the name implies—is a particularly dark kind of style and tends to be quite illegible to the untrained eye. Styles of blackletter include Textura/Textualis, Schwabacher, Fraktur, Rotunda, Cursiva and Hybrida. The first typeface of the Latin alphabet to become available was a blackletter but this style would soon meet a rival type design: the Roman type.

Humanist/Venetian
The French engraver, printer and type designer Nicolas Jenson went to Venice in 1468 and opened his own printing workshop. It is hypothesized that Jenson studied printing under Johannes Gutenberg for a while, though there are no clear sources to verify that. Jenson would design all kinds of Gothic type but he also designed a new kind of typeface based on the humanist writing of Italian scholars of the Renaissance. For the first time a typeface was designed based on typographic principles rather than the constructed letters from the old manuscripts. He would use his first humanist typeface in ‘De Evangelica Praeparatione’ in 1470. In 1471 he introduced a Greek typeface which was used for quotations and in 1473 a blackletter typeface which he used for books on history and medicine. Jenson became a wealthy man and was eventually able to run as many as 12 printing presses at the same time and would release around 150 book until he died at 60 in 1480.

The Venetian typeface was very short-lived but it’s quite a prolific style of typeface which even nowadays still has its application. The Venetian typeface can be recognized by the following characteristics:
1. Calligraphic, often almost handwritten appearance.
2. Relatively small x-height. This means the lowercase letters are relatively short but with long ascenders and big capital letters.
3. Low contrast between thick and thin strokes.
4. Because of the low contrast the color of the text is dark. This means there is a lot of black space per line—as was specifically the case with the blackletter typeface.
5. Weight distribution according to a diagonal axis.
6. A sloping crossbar on the lowercase ‘e’. Often the right side of the letter featured a so-called ‘beak’ which is the pointy feature but in case of display typefaces could sometimes be larger and more elegant features.
7. Very wide ‘H’, ‘M’ and ‘N’ and characteristic is also the double top serifs on the ‘M’.
8. Small counters. This means the ‘eye’ of the lowercase ‘e’ and the ‘bowl’ of the lowercase ‘a’ are relatively small.

Notable typefaces of the Venetian/Humanist class are Berkeley OldstyleBrioso Pro, Centaur, JensonHightower and Spira.

Garalde/Old Style
After Jenson’s death his typefaces were employed by Aldine Press, a printing office started by Aldus Manutius in 1494 in Venice. Aldus Press is famous in the history of typography, among other things for being the first to release books in octavo size (comparable in size with a contemporary paperback) and most notably for the introduction of italics. The italic type was first used by Aldus Manutius and the Aldine Press in 1501, in an edition of Virgil dedicated to Italy. The italics were based on the Humanist cursive script first developed in the 1420s by Niccolò de’ Niccoli and first started as a distinct condensed type for simple, compact volumes. The punches for these types were cut by Francesco da Bologna also known as Francesco Griffo. It wasn’t until later when roman and italic were used together—italic being used for emphasis.

In the 1540’s Claude Garamond came to prominence first for a Greek typeface he was commissioned to create for the French king Francis I. Garamond’s typefaces would quickly become popular throughout France and Western Europe. Most italics used in contemporary Garalde typefaces are based on Garamond’s assistant Robert Granjon. In 1621, sixty years after Garamond’s death the French printer Jean Jannon issued a specimen of typefaces that had some characteristics similar to Garamond’s typefaces, though Jannon’s letters were more asymmetrical and irregular in slope and axis. For this reason many typefaces are misattributed to Jannon but it is said that in fact most modern revivals are based on Jannon’s work and not Garamond’s but the name just stuck. Whoever designed the typefaces we now know as Garamond, it remains a popular style and typeface even today.

Although the calligraphic influences were still obvious in the Garalde type, the typeface became a lot more constructed and designed. This refinement was the result of the improving skills and tools of the punchcutters. The Garalde style can be recognized by the following characteristics:
1. Relatively small x-height and very long ascenders.
2. Medium contrast between thick and thin strokes.
3. Weight distribution according to an oblique axis—not quite as severe as in the Venetian type.
4. A horizontal crossbar on the lowercase ‘e’.
5. Wedge shaped serifs.
6. Often very elaborate, elegant tales on the italic ‘Q’.

Important type designers and typefaces in the Garalde style can be categorized in 4 groups:
• 1495 – Italian – Aldus Manutius and Francesco Griffo with the Bembo typeface.
• 1540 – French – Claude Garamond, Robert Granjon and Jean Jannon with the Garamond and Jannon typefaces.
• 1600 – Dutch – Christoffel van Dijck and Miklós Kis with the Ehrhardt typeface.
• 1725 – English – William Caslon with the Caslon typeface. It should be noted that while English, the Caslon typeface was very similar to the Dutch typefaces at the time. Caslon became an immensely popular typeface and is an excellent book typeface even today.
Other notable Garalde typefaces are AthelasMinion and Sabon.

Transitional/Realist
In 1692 king Louis XIV commissions a Jacques Jaugeon to create a typeface to for the Imprimerie Royale. The typeface is called 'Romain de Roi' (‘King’s Roman’). It’s engraved by Louis Simonneau and the punches for the metal type are cut by Phillipe Grandjean in 1698. The Romain du Roi was the result of rational design: the letterforms were mapped on grids before being cut into metal. The Romain du Roi was not the first constructed alphabet, however, this was the first time the letters adhered to the grid so closely that it shows a distinct shift in style, with an increased emphasis on the general composition and an increase in the contrast between thick and thin strokes. This style would later influence the transitional typefaces of Pierre Simon Fournier (known for the Fournier and Narcissus typefaces) and John Baskerville. The full Romain du Roi set consisted of 82 fonts and was finally finished in 1745.

The Transitional style can be recognized by the following characteristics:
1. Medium x-height and relatively short ascenders and descenders.
2. High contrast between thick and thin strokes.
3. The weight is distributed according to a vertical axis.
5. A greater focus on horizontal and vertical lines.

In the 1920’s many revivals started appearing and the Baskerville typeface increased in popularity and sparked new transitional typefaces such as Times New Roman. Initially the Baskerville typeface was seen as disturbing, and according to the people—as was the case with the English gentleman Benjamin Franklin spoke of—took it too far in regard to the contrast and details. This is rather amusing considering the style of typeface the transitional typefaces inspired, which featured an even higher contrast.

Notable transitional typefaces are BaskervilleCapitolium 2FarnhamProforma and Tabac.

Scotch Roman
A sub-class of the transitional which should be mentioned is the Scotch Roman. It’s actually a gradual step towards the modern style which became popular in the early 19th century and you can see the last humanist traces are replaced by a minimal, mechanical appearance. The Scotch Roman typefaces feature ball terminals and are modeled on a design done by Samuel Nelson Dickinson in 1839 (cut by Richard Austin and cast by Alexander Wilson and Son in Glasgow).

Examples of Scotch Roman typefaces are Century SchoolbookGeorgiaHarrietMiller and Scotch Modern.

Didone/Modern
Baskerville’s typefaces featured a contrast which some considered to be “blinding the nation” but that’s not the furthest you could take the contrast as the Didone style shows. The first Didone typeface was Didot, designed by the Frenchman Firmin Didot and was first used in print in 1784. The Italian Giambattista Bodoni designed his Bodoni typeface around the same time and took inspiration from Baskerville’s typefaces; he took the horizontal serifs and high contrast and emphasized these features. Giambattista Bodoni would go on to personally engrave 298 typefaces during his life. In the image below you can see how Baskerville and Bodoni are in fact quite similar; Bodoni almost seems like a more minimalist design with a higher contrast.

Characteristics of the Didone style are:
1. Very high contrast between thick and thin strokes.
2. Weight distribution according to a vertical axis.
3. Due to the weight distribution a great emphasis on vertical lines.
4. Unbracketed hairline serifs.
5. Small aperture. This means that the letter shapes are rather closed.

The Didone typeface features such a high stroke contrast that it’s very unpleasant to read in long texts. There is such great emphasis on vertical lines rather than the natural flow of letters we've grown familiar with due to the chirographic foundation of Roman typefaces that you effectively get what typographers call a "picket fence" effect. This effect tires the eye quickly and disturbs the reading experience. The Didone style is therefore best used in short texts such as titles and headings. As such it did increase the typographical variety in the First French Empire and particularly Didot remains a popular typeface in fashion and remains prolific in French culture.

Notable Didone typefaces are BodoniDidot (Hoefler & Frere-Jones' version is simply the best), RePublicTWT ProsperoTeimer and Walbaum.

Slab serif
The slab serif typeface is a style which pops up at the beginning of the 19th century coinciding with the Industrial Revolution. This era asked for a strong, robust typeface which is easy to produce and use in various applications such as advertising and posters. The first known slab serif typeface used was ‘Antique’ by the British punch-cutter Vincent Figgins in 1815.

The slab serif is often called ‘Egyptian’, which is a name given due to the craze for Egyptian artifacts in Europe and North America in the early nineteenth century, which led typefounders producing Slab Serifs after Figgins' work to call their designs Egyptian even though there is no connection with anything actually Egyptian—except perhaps for the robust structure of the letters which could equate to Egyptian architecture. The term Egyptian had previously been used to describe sans serif typefaces in the United Kingdom, so the term 'Antique' was used by British and American typefounders. The term Egyptian was adopted by French and German foundries, where it became Egyptienne.

Essentially the slab serif style can be further categorized into three groups:

Clarendon/Egyptienne
1. Large x-height and short ascenders and descenders.
2. There is some contrast between thick and thin strokes.
3. The serifs are bracketed, meaning there is a bit of a curve between the serif and the stem.
4. Classical proportions, often with a long, curly spur and a teardrop bowl on the ‘a’.
5. The Clarendon style often features teardrop terminals.

The Clarendon style is the most classical slab serif and in construction relates most to the Scotch Roman and the classical Transitional typefaces. Notable Clarendon typefaces are Belizio, Clarendon and Suomi Slab Serif.

Neo-grotesque
1. Large x-height and short ascenders and descenders.
2. Very low to no contrast between thick and thin strokes.
3. The serifs are generally unbracketed.
4. Minimal design and a construction like a sans serif.

The neo-grotesque typefaces are the most minimal slab serif variant and is often considered the most modern type of serif. It would tire the eye if long texts were set in a typeface of this style but in contemporary typography it’s often seen in short texts and display use.

Prolific neo-grotesque typefaces are AdelleKulturistaMuseo SlabPrelo Slab and Sánchez.

Italienne/Tuscan
1. Large x-height and short ascenders and descenders.
2. The serifs are heavier than the stems.
3. The serifs are often dramatic and decorative.
Italiennes are most easily recognizable as this is the only class of Roman typefaces which features decorative elements. These typefaces are probably mostly seen in the circus and Wild Western themes. Notable typefaces are Buckboard, De Louisville, Playbill and Wood Type.

Grotesque/grotesk
Although sans serif type can be found in Latin, Etruscan and Greek inscriptions as early as 5th century BC, it was used in 1748 as an experiment and it wasn’t until 1805 that the sans serif would make its first appearance in printed media (in European Magazine). In 1832 the first sans serif printing typeface was finished including lowercase by William Thorowgood of Fann Street Foundry. Thorowgood also coined the term ‘grotesque’ based on the Italian ‘grottesco’ (‘belonging to the cave’).

Although ‘grotesque’ is often used interchangeably with ‘sans’ and other such terms as ‘egyptian’, ‘antique’ and ‘gothic’, the grotesque/grotesk style as we know it today is a distinct sans serif style:
1. Large x-height and short ascenders and descenders.
2. Very subtle weight contrast.
3. Most commonly features the double story ‘g’.
Sometimes ‘neo-grotesque’ is considered a distinct style from the grotesque because the neo-grotesque would feature a more consistent stroke weight and a more mechanical appearance. For comparison, Akzidenz Grotesk would be considered a grotesque while Helvetica (which is based on Akzidenz Grotesk) would be a neo-grotesque. There are such few genuinely grotesque typefaces which have been digitized though and the differences are simply not very distinct so we will consider both styles ‘grotesque’.

Some of the best known grotesque typefaces are Akzidenz GroteskDINHelvetica and Univers but I also want to give attention to a few of my favorites: Alright SansBuloEpoca Classic and Marat Sans.

Geometric
With the mechanization and increasingly minimalist approach to architecture and general design at the beginning of the 20th century there was a need for more minimal, modern typefaces. This movement in typography went parallel with the Bauhaus/De Stijl movement and featured the same design elements.
1. Relatively low x-height (for a sans serif).
2. Use of geometric elements and repetition.
3. Often a minimalist approach and stylization of details, such as a spurless ‘G’ and a one story ‘a’ and ‘g’.

Some of the most prolific typefaces in the genre are Brandon Grotesque, Futura, GothamNeutraface, NobelSofiaSoleil and Verlag.

Humanist
The humanist sans serif—as the name suggests—features humanist elements. It doesn’t necessarily have to look that different from the grotesque and in fact grotesque and humanist typefaces tend to have some overlap. However, characteristic for a humanist sans are the chirographic features such as more weight in the curves and a subtle weight distribution according to an oblique axis.
1. Medium x-height.
2. Weight distribution according to an oblique axis.
3. Humanist proportions, meaning the eye of the ‘e’ and bowl of ‘a’ are often small and rounded letters tend to be relatively wide.
4. Calligraphic features such as extra weight in the curves.
5. Sometimes features soft terminals rather than a straight cut.
6. Usually has a rather elegant double story ‘g’ reminiscent of serif typefaces.

The best known humanist sans typefaces are FrutigerGill Sans and Scala Sans by far. The humanist sans class is my personal favorite though so I would like to name a few more typefaces I appreciate: Aragon SansElemental SansGraublau SansKarmina SansNovel SansTabac Sans and Winco.

And that’s the history of Roman type covered. When selecting a typeface for your design always consider both its appearance and its history. This becomes even more important when pairing two typefaces together because not everything works together. For example, you may pair two typefaces from the same time or the same proportions or you might pair two typefaces which contrast in style, color or proportions and compliment each other. There are so many combinations which work but you have to know a bit about a typeface in order to use it consciously and appropriately—particularly when establishing a brand with more than one typeface.

## Mastering Typography

Wed Nov 21, 2012, 12:41 PM

### Mastering Typography

Typography is a rather obscure field, which is quite odd because typography is in everything you read. As such, it’s very important for a graphic designer to know everything there is to know about typography and preferably also the history and anatomy of typefaces. Bad typography is usually only consciously noticeable by type geeks (like me), however even someone who knows absolutely nothing about typography will find text with bad typography harder to read. In this article you will learn some basic principles and features to advance your typography.

Software
Before we have a look at the features to improve your typography we have to select the right software. A few features which will be discussed in this article can be used pretty much regardless of the software but for most things you have to use genuine design software, so a program like Microsoft Word won’t suffice. You can achieve a lot with Adobe Photoshop—which is presumably what most are using—but from my experience the best typography is achieved in Adobe InDesign. Adobe Illustrator is better than Photoshop but not as good as InDesign when it comes to typography. If you’re not using Adobe software I can’t verify whether your program has all the features I use, and where they’re located but the principles discussed in this article will still be beneficial even if you can’t use all the features.

Presets
Before we can start playing with text we need to make sure a few presets are correct.  In InDesign (and Photoshop), go to Edit > Preferences > Type and make sure the Typographer’s Quotes (called Smart Quotes in Photoshop) are turned on.

If you’re using InDesign you have to set a few more presets. Go to Edit > Preferences > Units & Increments and set the ruler units to Points instead of Picas. This ensures the increments will be consistent with the pt size you’re using for the text and in general it’s just a lot easier to work with. Then go to Edit > Preferences > Dictionary and set your dictionaries if you have any installed on your computer and make sure the Quotes look like the ones in figure 1. You can also select any of the other quotation marks from the drop down menu but it’s probably not advisable because they’re not universal. For decorative reasons I may on occasion use guillemets (« ») which are predominantly used in French, but the quotes in figure 1 are definitely recommended. To find out more about which kind of quotation marks are used per language, go to this Wikipedia article.

If you’re using InDesign there’s one more function you should turn on by default and this one is very significant. Go to Type > Story and turn the function on. The pt size should be the same as the pt size you’re using in the text or a few pt sizes bigger depending on how it looks, so every time you change the text size make sure you also adjust the Story pt size accordingly. What this function does is add hanging punctuation.  This means all punctuation like quotation marks, hyphens and commas will be placed outside of the text block which gives the text block a more harmonious and professional appearance. The pt size of Story indicates how far away from the text block the punctuation is. Look at figure 2 to see the difference. If you’re using Photoshop, unfortunately you won’t find this function.

The character panel
That’s the basic presets taken care of. Let’s have a look at the character panel. First of all, make sure you have both the character panel and the paragraph panel open and in case of Illustrator also the OpenType panel. In case of Photoshop and InDesign you will find the OpenType functions in the drop down menu by clicking on the icon on the top right corner of the character panel. Now create your text field and fill it with text. At the bottom of the character panel you will notice a field to select the language. Select the language you’re using and if you type a different language, don’t forget to set the language in the language field accordingly. This is essential for proper hyphenation, which we will get to in a moment.

Spacing and kerning
We’re now going to specify some important things to improve the spacing and kerning. At the left side of the character panel you will see the kerning field which is set to “metric” on default. Metric kerning uses the kerning pairs (letter combinations) as defined in the font. If the type designer added specific kerning pairs (like LA, Ta, WA, VA, vy and many others) then your text will probably have the best flow if you keep metric kerning. Use metric kerning only for long texts. If you’re using a typeface of questionable quality you’re probably better off using optical kerning. Optical kerning derives the spaces between letters from the shapes of the letters. This is preferable for headlines because it brings the letters closer together without causing any disturbances like you would by reducing the tracking. If you’re using a lot of text variables like subscript, superscript but also very frequent use of different styles and weights you should go for optical kerning even if it’s for long texts, because kerning pairs do not work with combinations of styles, weights and other text variables.

Tracking
The design of a professional typeface doesn’t only regard the letters but also the space in between the letters. The counters (the negative space within a closed letter) and aperture (the negative space within an open letter) has a specific width and the distance between each letter is balanced with the negative spaces within the letters. This means that usually the tracking is perfect the way it is. However, you may sometimes alter the tracking to improve readability. For example, if I use a rather small pt size (9pt for instance) then the letters will optically look very close together. In my experience in such case you will improve the readability by increasing the tracking by 10 to 25. Don’t go beyond that for long texts because that will decrease the readability. For headlines you may reduce the tracking by 10 to save more space but don’t over-do it. Very tightly tracked text was popular in mainly logos during the 90’s but it’s absolutely out of fashion now. Modern typography is usually clean and fresh and needs some space to “breath”. If you’re setting very big display text (like a typographical illustration in a magazine) you may even go lower than a tracking of -10 but use it conservatively.

If you’re using a typeface with overly generous spacing (Dederon Serif and particularly Mrs. Eaves are good examples of this) and you think it’s too much, don’t try to fix it by altering the tracking! As I mentioned, both the letters and the negative space is designed by the type designer so by reducing the tracking to fix overly generous spacing you will risk obtrusive elements like letters touching each other. Remember that tracking can subtly be altered to improve readability but it should never be used to fix an issue in the typeface itself. Try optical kerning and if that doesn’t fix your issue, choose another typeface.

The leading is the vertical space between each text line. I usually slightly increase the leading of longer texts because it gives the text a bit “cleaner” appearance and it slightly reduces the overall color of the page; with very tight leading the text lines will be so close together that a larger amount of the background is covered, meaning a darker overall color.  Have a look at figure 3 to notice the difference in color. For text at 11pt the leading is automatically set to 13.2pt. For this text size I might increase the leading to 14pt but no more than that.

The leading should also be adjusted according to the width of the text column. For very long text lines the leading needs to be increased so after reading one line of text you won’t confuse the next line with the line you just read (which happens with very long text lines and reduced leading). Equally, with very short text lines the leading can be reduced, but don’t go below the automatic leading relative to your text size (in this case 13.2pt leading for 11pt text).

And lastly, the leading should also be adjusted according to the text size. If you’re using 24pt text for a heading for instance and the heading continues on a second and maybe even a third line then you need to decrease the leading even below the automatic leading (which for 24pt text corresponds to a leading of 28.8pt but depending on the length of the text lines you may find a leading of 26–27pt more comfortable). The larger the text, the more you can reduce the leading relative to the text size.

I suppose to some this begs the question, with this many factors which need to be taken into account when specifying the leading, how do you know which pt size to use? Well, it’s not an exact science and while for a big part it can improve or deteriorate the readability of the text, for some part it’s just a matter of taste. I would play around with the settings and decide what’s best—that’s how I do it.

The paragraph panel
The paragraph panel I’m sure most Photoshop users are familiar with. You have a couple of options on how to align your text: left, right, center, justify left, justify right, justify center and full justify. As you might know, by justifying your text you force the text to align to both sides of the text column. This creates very nice text blocks as opposed to the inconsistent line lengths of aligned text. While aligned text doesn’t give you nice text blocks some professional typographers and graphic designers actually prefer these inconsistent line lengths because the added variation at the end of the text lines according to some theories improve readability. Book typography might be the pinnacle of readability and legibility (the ease at which individual letters are distinguished from each other) and it’s interesting to see that text in books is usually justified to the left. I like text blocks but both align and justify have their function and place. In many magazines you will notice both are used (with consistency though).

Rivers
Justifying your text is great for text blocks in magazines and newspapers but at the same time this is one of the key features in which the difference between an amateur typographer and a professional typographer becomes apparent. Consider what happens to your text when you justify it. Words and even individual letters are driven apart for the sake of aligning the text to both sides. Especially with full justification you can absolutely disrupt the readability of your text block. This is because the justification tends to create big gaps in your text. These gaps often create patterns and are therefore called rivers. Avoid rivers AT ALL COST. Rivers are what make your typography look amateurish—more so than any other typographic tool. Have a look at figure 4 to see what rivers look like. You will also notice the word "philosophers" is distributed over the full width of the text column. Not only is this bad practice but a text should never end with a single word on one text line. In good typography there are at least two words on a text line.

Justification
So if justification can potentially ruin your typography, why would you even use it? Well, the full justification feature is a particularly aggressive tool and the other justification tools are actually more space-friendly. However, full justification still has its place in typography. For example, it’s often used in album layouts for lyrics or track listings. The justification tools in the paragraph panel can make or break your design but what most people don’t seem to know is that there are more options by clicking on the icon on the right side of the paragraph panel. In the menu which pops up you will see “Justification”. Click on that and set the minimum word spacing to 90%, the maximum word spacing to 110% and the maximum letter spacing to 1%. The word spacing values are brought closer to the desired 100% to limit the gaps in the text and a maximum letter spacing of 1% is picked so the individual letters are horizontally distributed by 1%; it’s not enough to add gaps between letters but it’s just enough to fill a bit more of the gaps between words. You can play with these settings a bit. These values always worked for me but how optimal it looks depends on the width of the column, text size, typeface, tracking and whether the Story function is turned on (which should always be turned on, and remember to adjust the Story pt size to the text pt size).

Hyphenation
In the menu of the paragraph panel you will also notice Hyphenation. In this panel you can change how words are broken up if it doesn’t fit on one line or creates unnecessary gaps. Hyphenation improves the distribution and rhythm of your text further. First, decide whether you want better spacing or fewer hyphens. A lot of hyphens can be obtrusive and ugly but of course those rivers should be avoided as well. Try to find a balance between the two and try to change the value of “After First:/Before Last: 2 letters” from 2 to 3 letters to see if that improves anything. As with justification, just play around a bit. Also, make absolutely sure you select the correct language in the characters panel because this is where it should all come together. If you have English text but the language is set to French for example then the text will be hyphenated as if it’s French while it’s actually English. In other words, if you select the wrong language your text will be hyphenated at incorrect places of the words.

If you did everything correctly you should now have a fairly perfect text block. See figure 5 to see my outcome and compare it to figure 4.

Ligatures
The final feature you should make use of for advanced typography is ligatures. In Illustrator these are accessible through the OpenType panel and in case of Photoshop and InDesign you should click on the right icon of the character panel and then click on “OpenType”. Here you will find the different sets of ligatures, alternate sets and numeral sets (have a look at this article about the different numeral sets). If there are brackets around the name (or in case of Illustrator, if the panel or icons in the OpenType panel are grey instead of black) unfortunately your typeface doesn’t have this feature. If it does, you first need to activate the default ligatures. These are your basic ligatures like fi, ffi, fl, fb, fk etc. and although these ligatures are quite subtle it will take your typography to the next level. Without ligatures you may see some obtrusive elements, such as the tittle (the dot on i and j) of the i touching the terminal of f. A ligature will replace that obtrusive element with a much more harmonious design. Always turn the standard ligatures on by default unless you go beyond a tracking of 25. Higher than that and all letters will get far apart while the ligature pairs stay close together and this disturbs the rhythm of the text.

The other ligature sets are discretionary ligatures (decorative ligatures) and contextual ligatures (which improves the rhythm and color [weight distribution] of the text). The contextual ligatures should also be turned on by default but there are very few typefaces with contextual ligatures. As for discretionary ligatures, you have to be careful with those because decorative ligatures can be rather obtrusive, particularly in longer texts. Typefaces like Caslon and Garamond do have discretionary ligatures and although you will probably find a lot of books printed in these typefaces (particularly Caslon), you usually won’t find a book in which discretionary ligatures are used exactly because they can be obtrusive. If you’re working on a project in a Renaissance or Baroque style though, you will definitely want to activate these ligatures because they’re prominent features of classical typography. However, since OpenType was introduced ligatures have made their way back into digital type setting and these days most professional typefaces do have discretionary ligatures as they're useful for creative typography. If you use ligatures appropriately you can make your text look absolutely perfect. That is, if you know how to select the most suitable typefaces for your project…

To find more about the ligatures and to see what each type of ligature looks like, visit this article I wrote about ligatures.

Hyphen, En-dash & Em-dash
A dash is a dash, right? In advanced typography, definitely not! There are in fact 3 different horizontal lines with each their own function. The hyphen we’ve already seen by automatically hyphenating our text but naturally you will also find the hyphen on your keyboard. Only use hyphens to break up words or to combine two words (when appropriate).

The En-dash is the most obscure one of the tree. It’s called an En-dash because this dash has the same width as the letter n. The En-dash should only be used to indicate a range in values or distances, such as in numbers (1–9) or place names (Amsterdam–New York flight). In bad typography the hyphen is used instead but now you know never to do that again. You can access the En-dash on your keyboard by typing Alt + 0150.

The Em-dash is of course named after the fact that it’s the same width as an m. This dash should be used to set of parenthetical elements such as in this sentence:
The brown fox—who was quick as a fox could be—jumped over the lazy dog.
To type the Em-dash press Alt + 0151.

Beauty in typography
Once you get used to each principle described here I’m sure it will quickly become a habit and you will do typography correctly no matter where you are. I find myself using the En-dash, Em-dash and semicolon even on the Internet. Once you know the shortcuts for the two dashes it seems silly to have always simulated an Em-dash with two hyphens—an artifact from the 90’s when digital typography was still severely limited. Typography is more than just a set of rules though. Once you become familiar with it you realize it’s a form of art and it will probably become a passion. There’s something so beautiful about seeing perfection in something other people never even think about. You read every day, but who really thinks about typesetting or even stranger, type design. Who are these geeks who spend months working on a single typeface? Master typographers!

All bells and whistles

## Numeral sets

Journal Entry: Sat Nov 3, 2012, 6:15 PM

TYPOGRAPHY SERIES:
01 – Anatomy of typography
02 – Ligatures
03 – Numeral sets

A lot of people may think that numbers are just numbers but within typography certainly not all numbers are alike. As the Latin alphabet consists of both uppercase and lowercase letters, we also have numeral sets to go with both and we have a few other sets with different functions. Part 3 in the typography series is about the different numeral sets and their functions within typography.

Click on the image below to go to the deviation page so you can comment and/or add it to your favorites and download a zip file including a 2000px wide .jpg and an .ai vector file:

## Ligatures

Journal Entry: Wed Oct 31, 2012, 6:18 PM

TYPOGRAPHY SERIES:
01 – Anatomy of typography
02 – Ligatures
03 – Numeral sets

The first tutorial from the typography series was about the Anatomy of typography and now I offer you part 2 in the series: ligatures.

In the 90's digital typography was very basic but as the technology progressed the typography became increasingly sophisticated. Nowadays it's even possible to use ligatures through OpenType. In this tutorial you will learn about the three types of ligatures and what they're useful for.

Click on the image below to go to the deviation page so you can comment and/or add it to your favorites and download a zip file including a 2000px wide .jpg and an .ai vector file:

## Anatomy of typography

Journal Entry: Fri Sep 28, 2012, 2:08 PM

TYPOGRAPHY SERIES:
01 – Anatomy of typography
02 – Ligatures
03 – Numeral sets

There have been many images going around the Internet explaining the basic elements of typography but most of these sources are incredibly basic and incomplete. That's why I decided to come up with my own version and be as complete as possible.

I excluded several things like diacritics, the different ligature types, numeral sets etc. to avoid making it look too chaotic but most of the basic elements seem to be there. I will focus on diacritics, ligatures and numeral sets later in this ongoing typography series.

Click on the image below to go to the deviation page so you can comment and/or add it to your favorites and download a zip file including a 2000px wide .jpg and an .ai vector file:

## Dark spots in type design

Journal Entry: Thu Aug 23, 2012, 11:20 AM

Type design is all about balancing the black versus white to create a consistent texture. One of the—perhaps more obscure—principles of creating well balanced type is a consistent stroke weight in certain areas—particularly in the joins. This is true even for simple geometric typefaces of which you probably wouldn't expect it if you're not familiar with dark spots.

#### Ink traps

Dark spots occur in places where two strokes meet and this is particularly true when two strokes meet at a sharp angle. When two strokes meet there is a weight build-up which should be optically adjusted. The letter 'M' for example usually has three dark spots, one for each bend. Adding so-called ink traps [picture below] can get rid of these dark spots but it's quite an obtrusive method. essentially you add an extra vector point in the narrow negative space area and simply widen the area to give the black room to "breathe".

Ink traps are mainly reserved for two things:

To avoid the flow of ink into narrow negative spaces when using fast and cheap printing techniques such as newspaper printing.
Nowadays ink traps are often added as a style; a sort of decorative element.

If you want to add ink traps to your typeface for functional reasons rather than aesthetics you have to consider at what point size your typeface will be used. For display typefaces you would probably add relatively small ink traps compared to text typefaces as the flow of ink in smaller type is much more disruptive than it is for big type.

It should also be noted that not all typefaces require dark spots to be avoided/removed. In case of DIN for example these dark spots are actually part of the typeface's aesthetic. In case of a typeface like Futura however the weight of the strokes need to be consistent so it's good to avoid dark spots in this case. Besides, it's fine if you want your typeface to feature dark spots but it obviously should be a conscious choice and shouldn't occur due to ignorance.

## Removing dark spots

So how should dark spots be avoided without using ink traps? The principle doesn't differ much from ink traps. Instead of widening the narrow negative space you make it longer. You will often have to add a few extra vector points so you can make the negative area go inwards into the letter a tiny bit besides elongating it. Look at the image below to see the difference and particularly note the vector points in regard to the guidelines.

And that's it really! I assure you that by avoiding these dark spots when appropriate will improve your typefaces instantly. It will make the texture of your typeface more consistent and in effect will improve the reading experience.

## The primeval atom et The accelerating universe

Journal Entry: Thu Aug 9, 2012, 10:08 PM

Albert Einstein was caught up with the idea of a static and eternal universe. Gravitation however causes all matter to attract each other with a force in relation to its mass. To prevent the universe from collapsing Einstein added the cosmological constant to his field equations of general relativity.

In 1929 Edwin Hubble formulated the empirical Redshift Distance Law of galaxies which is now known as Hubble's Law. After the observations of the cosmological red shift Einstein referred to his cosmological constant as the "biggest blunder" of his life.

In 1998 Saul Perlmutter, Brian P. Schmidt, and Adam G. Riess discovered through observations of distant supernovae—used as so-called standard candles—that the universe is actually accelerating. This is where the cosmological constant actually becomes relevant again as the equations can be used with a positive value to calculate the acceleration. As it turns out, the universe could never be in perfect equilibrium due to its inherent instability: if the universe expands ever so slightly the expansion releases vacuum energy which causes yet more expansion. Likewise, a universe which contracts slightly will continue contracting. The observations of an accelerating universe were consistent with a cosmological solution which Alexander Friedmann derived from Einstein's original general relativity equations.

This brings me to the reason for writing this article. The way I learned it, Einstein and Hubble were major contributors of the cosmological solution which eventually became the Lambda-CDM model—what is frequently referred to as the standard model of Big Bang cosmology. On a side note, another explanation of the accelerating universe is quintessence, which is a hypothetical form of dark energy and a model which only differs from the cosmological constant in that it's a dynamic equation that changes over time. I would go further into this but I'm afraid it's beyond the scope of this article.

As is often the case in physics, several people are working on the same problems and so it can often be challenging to attribute theories verified by observations to the right people. In fact some of Einstein's ideas were (partly) conceived previously by people completely overlooked.

Although not completely overlooked, Georges Lemaître didn't really get all the credit he deserved until later. This Belgian priest, astronomer and professor of physics at Catholic University of Louvain was the first person to propose the theory of the expansion of the universe, which—as we now know—is widely misattributed to Edwin Hubble. He was also the first to derive what is now known as Hubble's law and made the first estimation of what is now called the Hubble constant. He published this in 1927, two years before Hubble's article. Moreover, Lemaître also proposed the hypothesis of the origin of the universe which became known as the Big Bang theory. This is the notion that the universe came from a single point. He called this the hypothesis of the primeval atom. Incidentally, Lemaître described this theory as "the Cosmic Egg exploding at the moment of the creation" and to his discontent was adapted by Pope John XXIII who took it as proof of the biblical story of Genesis and in complete accordance with the teachings of the Church. After all, in a universe which is infinite and timeless a God as the creator becomes irrelevant.

## R+S - Font Gallery Update!

Journal Entry: Fri Jun 29, 2012, 5:51 PM

The new categorization system for fonts on deviantART is online!

and I have been working hard to come up with a proper font categorization system on DA and with the help of it's finally live! I would also like to mention here for helping me name and define the categories in a manner which would be understandable not only to professional type designers and historians but to the common folk on deviantART as well.

To read all about this major update (and a big step forward for the DA typography community) check the Community Relations Journal update of Jun 25, 2012.

And for those who have been waiting for the articles about each typeface category I announced months ago, I have been quite absent lately and I have my reasons, but I promise the articles are still coming!

## Lucid dreams

Journal Entry: Tue May 22, 2012, 1:50 AM

A few days ago I had two lucid dreams. After the second lucid dream I wrote everything down so I wouldn't forget because this seems interesting to share.

I don't remember the first lucid dream. All I remember is being afraid because I realized I was dreaming and I didn't enjoy the experience. I tried hard to wake up, which I did after around 10 seconds. I then went back to sleep, not expecting to experience a lucid dream again (but really, the highest potential to get a lucid dream is when you wake up in the morning and go back to sleep).

When the second dream starts I'm initially not aware that I'm dreaming, so at this point it's a regular dream. I'm sitting behind the table in the living room of my parents' house and the mother of my sister's old school friend is there for some reason. I don't remember who was in the room besides her but it was probably one or both of my parents. I don't have an idea what everyone was doing, either. Suddenly the woman walks up to me and introduces herself. I want to tell her that I know her as we've met several times before, but for some reason I can't speak normal words. I'm speaking in a non-existing language, or rather just trying hard to pronounce the words I know but they just wouldn't come out. The woman and one of my parents were looking at me strangely. This disturbed me greatly and it was at this point I realized I was dreaming. I tried to wake up again but this time it took more time. I first managed to wake my right side up while my left side was still in the dream. That last bit was perhaps most disturbing to me because I can't even explain how just one side can be sleeping. It's not like one side was paralyzed; I was really still half in the dream.

This also reminds me I had a disturbing experience a few weeks ago. I suddenly realized I was sleeping (not dreaming) and not breathing. I fought to wake up so I could breath again. I panicked in the moment but managed to force my mind awake, open my eyes and take a breath. I suspect this was a half-lucid dream because I can't imagine I really wasn't breathing.

## Dinosaurs and the Bible

Journal Entry: Sun Mar 25, 2012, 5:58 PM

Yes, the title is as ludicrous as this article will be. Really, this is the silliest article I've written so far.

Yes, the title is as ludicrous as this article will be. Really, this is the silliest article I've written so far. I just read an article from the genius Ken Ham, written in 1999. Let me state the best parts of his article (which is pretty much all of it) for both my and possibly your amusement:

"According to evolutionists, the dinosaurs "ruled the Earth" for 140 million years, dying out about 65 million years ago. However, scientists do not dig up anything labeled with those ages. They only uncover dead dinosaurs (i.e., their bones), and their bones do not have labels attached telling how old they are."
I guess paleontology is a hoax profession.

"The idea of millions of years of evolution is just the evolutionists' story about the past."
Good argument. It's just the evolutionists' opinion; it's not at all based on evidence.

"No scientist was there to see the dinosaurs live through this supposed dinosaur age."
Another good argument, because the creationist DID see how God was scattering the bones onto the Earth around 6000 years ago and there's plenty of evidence for that as well.

"Other scientists, called creation scientists, have a different idea about when dinosaurs lived. They believe they can solve any of the supposed dinosaur mysteries and show how the evidence fits wonderfully with their ideas about the past, beliefs that come from the Bible."
Creation scientists! I like the fact that Johannes Kepler and Galilei Galilei practiced science at the end of the 16th century and they seem to have had a much more proper understanding of both reality and science than these "creation scientists".

"The Bible, God's very special book (or collection of books, really), claims that each writer was supernaturally inspired to write exactly what the Creator of all things wanted him to write down for us so that we can know where we (and dinosaurs) came from, why we are here, and what our future will be."
That's solid then.

"Genesis tells us that God created everything—the Earth, stars, sun, moon, plants, animals, and the first two people."
No, Genesis tells us that God created:
Day 0 — Heaven and Earth
Day 1 — light (I always turn on the light after I take a piss; never before), a division of light and darkness (which is a good thing, otherwise light and darkness would blend into one substance), evening and morning,
Day 2 — water, Heaven (even though Heaven and Earth were created "in the beginning" which I designated as day 0),
Day 3 — dry land (imagine that; first creating water and THEN the absence of it), seas, plants,
Day 4 — night and day (even though evening and morning already occurred on day 1), seasons (which wouldn't happen with only the concept of day and night, at which point time would've been introduced), days (I like how days are created on the 4th day), years, two lights (is it getting silly just now?); one greater for the day and one lesser for the night,
Day 5 — sea creatures, every living thing which moves, flying creatures (not a living thing which moves),
Day 6 — beasts of the earth (again, this is different from every living thing which moves introduced on day 5), man, mankind's dominion over the world,
Day 7 — God ends his work and rests, though no mention of these finishing touches

There's no mention of stars. Remember God created two lights; a lesser light (moon) and a greater light (Sun). I think God should've created billions and billions of "greater lights" to account for the universe. I also wonder whether God followed the same process for the other planets. If so, I wonder how the phrase "And he saw it was good" pertains to all those inhabitable planets, or planets which became inhabitable.

"As you add up all of the dates, and accepting that Jesus Christ, the Son of God, came to Earth almost 2000 years ago, we come to the conclusion that the creation of the Earth and animals (including the dinosaurs) occurred only thousands of years ago (perhaps only 6000!), not millions of years. Thus, if the Bible is right (and it is!), dinosaurs must have lived within the past thousands of years."
And God created the dinosaurs in his own image (Tsintaosaurus spinorhinus in particular).

"Evolutionists claim that dinosaurs evolved over millions of years. They imagine that one kind of animal slowly changed over long periods of time to become a different kind of animal. For instance, they believe that amphibians changed into reptiles (including dinosaurs) by this gradual process. This would mean, of course, that there would have been millions of creatures during that time that would be "in between," as amphibians evolved into reptiles. Evidence of these "transitional forms," as they are called, should be abundant. However, many fossil experts admit that not one unquestionable transitional form between any group of creatures and another has been found anywhere. If dinosaurs evolved from amphibians, there should be, for example, fossil evidence of animals that are part dinosaur and part something else."
What about Amniotes (among them Archosaurs) and Synapsids (among them Pelycosaurs and Therapsids which are the ancestors of mammals)? Fossil records indicate that birds evolved from theropod dinosaurs and some species survived the Cretaceous–Paleogene extinction event and brought the dinosaur lineage to the present day.

"However, there is no proof of this anywhere. In fact, if you go into any museum you will see fossils of dinosaurs that are 100% dinosaur, not something in between. There are no 25%, 50%, 75%, or even 99% dinosaurs—they are all 100% dinosaur!"
Dinosaurs are a diverse group of animals of the clade Dinosauria; it's not a species, genus, family, order etc. This means that each species within this clade is 100% dinosaur; that doesn't imply that there's no evolution between generations of species.

"Evolutionists declare that no man ever lived alongside dinosaurs. The Bible, however, makes it plain that dinosaurs and people must have lived together. Actually, as we will soon see, there is a lot of evidence for this."
That's an interpretation. The Bible does not make it plain.

"The Bible teaches (in Genesis 1:29–30) that the original animals (and the first humans) were commanded to be vegetarian. There were no meat eaters in the original creation. Furthermore, there was no death"
If everything living was vegetarian in the original creation and modern carnivours are biologically engineered to process meat, doesn't that imply that evolution exists? This is explained away by Adam's sin, which changed the world into what it is now.

"After Adam's sin, animals and people started to die. It was now a different world, one of death and strife. A world that was once beautiful now suffered under the curse placed upon it by the Creator (Genesis 3:14–19)."
A world suddenly full of motivation and a sense of purpose. It sounds atrocious.

"Some people think that dinosaurs were too big, or there were too many of them, to go on this Ark. However, there were not very many different kinds of dinosaurs. There are certainly hundreds of dinosaur names, but many of these were given to just a bit of bone or skeletons of the same dinosaur found in other countries."
Obviously! Let's assume evolution does not exist, which means no new species could have evolved. This implies that the current 1.4 million described species (and there are more) also existed during Genesis as well as the dinosaurs (no in-between species of course!). So even if only one species of dinosaur ever existed, you would have to fit 2.800.002 animal species plus Noah and his family on that Ark. There's no way that's not going to fit.

"It is also reasonable to assume that different sizes, varieties, and sexes of the same kind of dinosaur have ended up with different names. For example, look at the many different varieties and sizes of dogs, but they are all the same kind—the dog kind! In reality, there may have been fewer than 50 kinds of dinosaurs."
That's very far from reasonable to assume, but let's go with it.

"God sent two of every (seven of some) land animal into the Ark (Genesis 7:2–3; 7:8–9)—there were no exceptions. Therefore, dinosaurs must have been on the Ark. Even though there was ample room in the huge ship for large animals, perhaps God sent young adults into the Ark that still had plenty of room for them to grow."
Obviously a baby Argentinosaurus will fit on the Ark. Its size aside, an adult Argentinosaurus would weigh around 100 tons so that gives some suggestion on how much a young one would weigh.

"Interestingly, the word "dragon" is used a number of times in the Old Testament. In most instances, the word dinosaur could substitute for dragon and it would fit very nicely."
That's interesting indeed. Let's go with Revelation 13.2:
"And the beast which I saw was like unto a leopard, and his feet were as the feet of a bear, and his mouth as the mouth of a lion: and the dragon gave him his power, and his seat, and great authority"
So, a dinosaur gave the leopard/bear/lion hybrid a seat and great authority.

"Unfortunately, this evidence is not considered valid by evolutionists. Why? Only because their belief is that man and dinosaurs did not live at the same time!"
Yes, that's the only reason! It's ludicrous indeed.

"However, the more we research the historical literature, the more we realize there is overwhelming evidence that dragons were real beasts, much like our modern reconstructions of dinosaurs, and that their existence has been recorded by many different people, even just hundreds of years ago."
I guess it's a contemporary phenomenon. People used to see dragons and now they see UFOs. It's consistent with the mysteries of the universe which can be explained without a need for dragons and UFOs.

"Evolutionists use their imagination in a big way in answering this question. Because of their belief that dinosaurs "ruled" the world for millions of years, and then disappeared millions of years before man allegedly evolved, they have had to come up with all sorts of guesses to explain this "mysterious" disappearance."
The mind of an evolutionist sure can run wild!

"When reading evolutionist literature, you will be astonished at the range of ideas concerning their supposed extinction. The following is just a small list of theories:
Dinosaurs starved to death; they died from overeating; they were poisoned; they became blind from cataracts and could not reproduce; mammals ate their eggs. Other causes include volcanic dust, poisonous gases, comets, sunspots, meteorites, mass suicide, constipation, parasites, shrinking brain (and greater stupidity), slipped discs, changes in the composition of air, etc."

I'm astonished indeed. Those crazy evolutionists! Do you remember the Cretaceous–Paleogene extinction event where dinosaurs started overeating and got cataract? Those sunspots were also a big issue (damn you cooler regions on the Sun!) and then there were these suicidal dinosaurs as well.

"If you remove the evolutionary framework, get rid of the millions of years, and then take the Bible seriously, you will find an explanation that fits the facts and makes perfect sense"
Which implies the Bible ONLY makes sense if you "get rid of" the millions of years and the evolutionary framework.

"At the time of the Flood, many of the sea creatures died, but some survived."
I guess the sea creatures drowned by the flood.

"In addition, all of the land creatures outside the Ark died, but the representatives of all the kinds that survived on the Ark lived in the new world after the Flood. Those land animals (including dinosaurs) found the new world to be much different than the one before the Flood. Due to (1) competition for food that was no longer in abundance, (2) other catastrophes, (3) man killing for food (and perhaps for fun), and (4) the destruction of habitats, etc."
So there were few enough species to fit on the Ark, but so many that there was a competition about food. Also, competition of food = natural selection = evolution. Other catastrophes is very vague but I guess it couldn't possible have been a meteor. I also like how all dinosaurs at the K-T boundary died at roughly the same time because man killed for food and fun (and remember, man = Noah and his family at this point).

"Creationists, of course, would not be surprised if someone found a living dinosaur. However, evolutionists would then have to explain why they made dogmatic statements that man and dinosaur never lived at the same time."
Rather than let the evolutionist explain his dogmatic statement WHEN a living dinosaur is found, why not let the creationist explain his dogmatic statement NOW, after so much evidence for evolution and the death of dinosaurs?

"We need to recognize that the wickedness in the world is because of sin, because man rebelled against God."
According to the Bible Eve was too curious for her own good. I don't see any rebellion. The only character who rebelled was Lucifer when he defied God by refusing to bow for Adam; the very reason he went from the highest angel to serve God to the devil who reigns in hell.

"We can also be reminded that God, who made all things, including the dinosaurs, is also a judge of His creation. He judged Adam's rebellion by cursing the world with death."</i>
God created a flawed Adam and Eve and proclaiming that "He saw it was good", after which it wasn't good and his own product had to be punished.

"The Bible teaches us that He will again judge the world, but next time by fire"
I see a pattern of increasing violence. First the punishment of a limited lifetime of around 900 years with the introduction of motivation and a sense of purpose, then punishment with water, and later fire. What's the next step? I would love to skip the fire apocalypse and go straight to the gamma ray burst apocalypse.

"But we are also warned that many will not be allowed into this new Earth but will suffer for eternity"
God balances everything out; living 80 years and suffering for eternity.

"The Lord Jesus Christ died on a cross, but on the third day, rose again, conquering death"
Thus not really having suffered the death penalty for sin.

## The grandest of scales

Journal Entry: Thu Mar 22, 2012, 10:20 PM

Everything is so relative and it only gets bigger and bigger.

While the red hypergiant (called red because red dwarfs and stars at the end of their life cycle are the coolest and therefore emit mostly red or infrared light) VY Canis Majoris is the biggest star observed at 1800–2100 solar radii (1 solar radius = radius of the Sun) it's "only" 30 to 40 solar masses. The most massive observed star is the blue hypergiant (the hottest stars emit mostly blue or ultraviolet light) R136a1 with an estimated 265–300 solar masses and a luminosity almost 9 million times greater than our Sun's luminosity.  The reason VY Canis Majoris is so big is because it's in the final stage of its life cycle. During this stage, the star bloats up into a red giant and pushes out its Hydrogen mantle. Eventually the core collapses and the star goes supernova. Stars like our Sun which are below the Chandrasekhar limit (the maximum mass of a stable white dwarf) at 1,38 solar masses will end up in a white dwarf, a small "corpse" of a star the mass of our Sun but the radius of Earth which is mostly composed of electron-degenerate matter and held together by electron degeneracy pressure instead of fusion against gravitational collapse. Stars beyond this limit will end in either a neutron star (1,38–3,00 solar masses) or a stellar black hole (any star greater than 3 solar masses), so both VY Canis Majoris and R136a1 will definitely end up as stellar black holes.

Stars more massive than 20 solar masses — called Wolf-Rayet stars — often lose their mass rapidly by means of extreme  stellar wind. R136a1 is a Wolf-Rayet star and as the star is middle-aged, it was even more massive in its youth. It's absolutely fascinating to get a sense of scale by comparing stars with each other: Wolf 359, the Sun, Sirius, Arcturus, Aldebaran, Rigel, Antares, Betelgeuse, Mu Cephei, VV Cephei A and the largest, VY Canis Majoris. That's nothing though.

While the most massive star is apparently nearly 300 solar masses, supermassive black holes are the real monsters. Just about every galaxy has a supermassive black hole at its core including the Milky Way (it's called Sagittarius A and weighs 3,7 million solar masses). The most massive supermassive black hole discovered so far is at the center of NGC 4889, the brightest galaxy in a supercluster of thousands of galaxies about 336 million light years away in the Coma constellation. This beast weighs a dazzling 21 BILLION solar masses. It should also be noted that supermassive black holes are proportional to the size of the galaxy they reside in. A typical galaxy contains around 400 billion stars.

So those are massive and gigantic cosmic objects, but perhaps a greater insight in the hugeness of everything is derived from zooming out all the way from Earth so each time we have a greater scale to compare with the previous: from Earth to the Sun, the Kuiper belt, the Solar system, the Hills cloud, the Oort cloud, the Orion–Cygnus Arm, the Perseus Arm, the Milky way, the Local Group (a group of 54 galaxies including the Milky Way), the Virgo Supercluster (a supercluster of more than 100 galaxy groups including the Virgo cluster and the Local Group) and eventually the grandest structures of superclusters.

The observable universe is 29 billion parsecs or 93 billion light years in diameter and that's likely to be a small portion of the total size of the universe. The reason we can't see all of the universe is because the universe is expanding. Nothing can go faster than the speed of light, though the cumulative expansion of space results in galaxies moving away from each other at greater speeds than the speed of light. As the space expands faster than the light can travel through it, light from the more distant objects will never reach us.

It's hard to grasp the tremendous scale of things as it and the universe are so unfathomable; I find it easy to see the universe as an endless fractal. The probability of the existence of our universe is very nearly 0 so statistically speaking I shouldn't be pondering this question; in fact, I shouldn't be. To fix this statistical anomaly you can state that our universe is just one of many, perhaps infinite amount of universes, in which case the existence of our universe is no longer  a cosmological oddity (or the notion of intelligent design) . With so many oddities like quantum mechanics, particle/wave duality, wavefunction collapse or the strong evidence that our universe has a flat geometry, I don't think it's hard to imagine our universe being part of a multiverse or a (looping) fractal. Reality is tremendously bizarre.

## The Silvertant type classification

Journal Entry: Sun Mar 4, 2012, 8:10 PM

Before we move on to the series of articles about the different type classifications (starting with 'Venetian/Humanist') I think it's important to establish a proper classification system.

# Why do we need to establish a new classification system?

To be honest, a system is limiting, and with the modern typefaces taking elements from various type classifications it's hard and at times seems pointless to try to categorize these typefaces. Regardless, it's important to know the history of typefaces and when certain features developed and a type classification system can help you keep track of things and label typefaces according to a wide range of parameters like x-height, stroke weight, contrast, texture, the shape of the serifs, the direction of the weight distribution etc. As I said in the previous article, the Vox-ATypI classification is a good point to start but it does need refinement. When we've established a proper system, we can then base the series of articles on this system. Mind you, I won't be establishing a complete system but rather a simple and compact system by which I can write articles where I will discuss the details. We will also be getting rid of some generic terms and update some old ones.

As for the articles though, I will only be discussing the evolution of typefaces. So the chirographic category (hand-written) won't be discussed. I barely researched script fonts and the history of writing and never learned how to properly work scripts fonts so I will leave someone else to bestow his/her wisdom of chirography upon us. The articles to come really only focus on the Roman typefaces and their roots.

I thought a long time about the name of my type classification and for lack of imagination (or out of vanity; you choose) I decided to call it the Silvertant classification (after my last name).

# The Silvertant classification

## Serif

The serif typefaces are primarily characterized by their serifs of course, but also a medium to high stroke contrast, triangular serifs or wedge serifs and a weight distribution according to a diagonal to vertical axis depending on the style and age from which the features were derived.

### Humanist/Venetian

Examples: Centaur, Roos, Brioso

### Garalde

Examples: Garamond, Caslon, Minion

### Didone

Examples: Didot, Bodoni, Filosofia

### Contemporary

Examples: Biblon, Coranto, Mokka

## Slab serif

The slab serif typefaces are characterized by a simple, functional feel that gained momentum during the industrial period. They're often called mechanistic or mechanized and feature slab serifs which are either squared (Egyptiénne) or bracketed (Ionic).

### Egyptienne

Examples: Glypha, Pragmatica Slab, Salvo Serif

### Clarendon

Examples: Clarendon Text, Belizio, Suomi Slab Serif

### Tuscan

Examples: Buckboard, De Louisville, Wood Type

### Contemporary

Examples: Adelle, Museo Slab, Centro Slab Pro

## Sans serif

The sans serif typefaces are characterized by their absence of serifs and their medium to very low stroke contrast. The first use of a sans serif typeface was around 1720. The sans serif typeface wasn't considered attractive in the beginning, hence the names 'grotesque' and 'gothic' are often used.

### Grotesque

Examples: Helvetica, Univers, DIN

### Geometric

Examples: Futura, Eurostile, Nobel

### Humanist

Examples: Gill Sans, Frutiger, Ideal Sans

## Chirographics

The Chirographics are characterized by—a suggestion of—being hand-written.

### Script

Examples: Reklame Script, Gelato Script, Metroscript

### Hand-writing

Examples: Andrij Script, Just Lefthand, Erik Righthand

### Comic

Examples: Comic Sans, Dion, Zoinks

## Blackletters

The Blackletters are a dark script form which developed around 1150.

### Textura

Examples: Goudy Text, Old English, Textura Quadrata

### Schwabacher

Examples: Alte Schwabacher, SchwarzKopf, Sibyl

### Fraktur

Examples: Breitkopf Fraktur, Fakir, Fette Fraktur

### Rotunda

Examples: 1483 Rotunda Lyon, Bucintoro, San Marco

### Hybrida/Bastarda

Examples: Burgundica, Givry, Lucida Blackletter

## Behind the typefaces

Journal Entry: Tue Feb 14, 2012, 10:02 PM

This is an introductory Journal to a series of Journals to come about the different type categories, the history and how to recognize these type classifications.

# Choosing a typeface

At times it seems very hard to differentiate typefaces and recognize from which time, style and designer certain typefaces or elements from a typeface are reminiscent of. It seems an issue which should only regard type geeks, but that almost couldn't be further from the truth. Understanding type and its evolution is in my opinion important to the average Joe as well, but predominantly it's a must for every graphic designer or artist working with text.

To illustrate the importance of understanding typefaces and their evolution, let me give you a specific example. Let's say you—as a graphic designer—are being commissioned to design the layout for a CD album of compiled works by Johann Sebastian Bach (1685–1750) and it should have an authentic look. In this case it's very relevant from which era the music is. If you go by appearance and personal taste in selecting typefaces to use you will risk making a fool out of yourself if you know nothing about the typefaces. For example, you could select Jenson, which looks quite authentic but is actually a Venetian typeface cut in 1470; 215 years before Bach was even born. Obviously it's not quite as crazy to use a typeface from before Bach's time as it is to use one after his time, but historically speaking Jenson wouldn't have been used during the Baroque. Garamond is also too old but it's becoming more authentic already. William Caslon I (1692–1766) lived in the same era as Bach and cut the Caslon typeface which became hugely popular and virtually every book at the time was set in Caslon or in typefaces based on Caslon. So it would be very appropriate to use Caslon (any of the many digitized versions of Caslon, though Williams Caslon Text by William Berkson is probably the most authentic Caslon for body text size).

It's important to note though that it's not absolutely necessary to choose a typeface which is culturally and historically authentic to the project, but you have to know what there is to know about typefaces in order to make a conscious decision to choose a different kind of typeface. The choice of your typeface shouldn't be a happy accident, and selecting a typeface "because you like it" is never a good argument. Choose consciously.

# Type classifications

Even if you understand the history and evolution of type, it's still quite complicated to categorize typefaces, as modern typefaces are designed based on the cumulative knowledge of many generations of type designers and very different styles; it tends to blur the boundaries between type classifications. Still, you can roughly categorize all typefaces which will simplify a lot of things, like selecting 2 typefaces to work together harmoniously.

The Vox-ATypI classification is in my opinion a good start. It's not a perfect system and it's even a little outdated, but I've been using it for years as it's a simple system which is easy to expand for personal reference. Another system I quite like is the Thibaudeau classification. It's a lot simpler and more general, but perhaps that's just what you need. We're going to look at both systems. Note though that I won't be going into much detail about each category as a series of articles is planned for each category—including pictures.

# The Vox-ATypI classification

## Classicals

The classicals are characterized by triangular serifs, oblique axis and low stroke contrast. In other classification systems, this group is often referred to as "oldstyle".

### Humanist/Venetian

Examples: Centaur, Roos, Brioso

### Garalde

Examples: Garamond, Caslon, Minion

## Moderns

The moderns are characterized by a simple, functional feel that gained momentum during the industrial period.

### Didone

Examples: Didot, Bodoni, Filosofia

### Mechanistic

Examples: Rockwell, Clarendon, Museo Slab

### Lineal: Neo-Grotesque

Examples: Helvetica, Univers, DIN

### Lineal: Geometric

Examples: Futura, Eurostile, Nobel

### Lineal: Humanist

Examples: Gill Sans, Frutiger, Ideal Sans

## Calligraphics

The Calligraphics are characterized by a suggestion of being hand-crafted.

### Glyphic

Examples: Trajan, Goudy Trajan, Aviano

### Script

Examples: Reklame Script, Gelato Script, Metroscript

### Blackletter

Examples: Burgundica, Old English, Givry

# The Thibaudeau classification

### Elzévirs

This family contains typefaces with triangular serifs.

Examples: Garamond, Palatino, Lexicon

### Didots

This family groups typefaces with linear or hairline serifs. It generally corresponds to modern or Didone categories.

### Égyptiennes

This family contains slab serif typefaces, called Mechanistic in the Vox-ATypI classification.

Examples: Rockwell, Memphis

### Antiques

This is the sans serif family. In Vox-ATypI classification, this family corresponds to the Lineals.

Examples: Helvetica, Univers, Futura

## Wolf-Rayet 1O4: Earth peeks into the gun barrel

Journal Entry: Tue Oct 12, 2010, 11:36 AM

The Mayan calendar ends in 2012. Will life on Earth end in 2012? [no]

The Sun has a cycle of a solar maximum and solar minimum of activity. The solar cycle takes an average of about 11 years to go from one solar maximum to the next with an observed variation in duration of 9 to 14 years for any given solar cycle. Some scientists predict the Sun will enter an extraordinarily high solar maximum in 2012 or 2014. Or... it might not happen at all in our lifetime. We really don't have the instruments to accurately predict the solar weather. What a heightened solar activity results in is more powerful solar winds with more frequency. Solar winds are streams of charged particles ejected from the upper atmosphere of the Sun. Earth's atmosphere usually defends us from most of these charged particles. Some solar winds however are more powerful, and Earth's atmosphere can't fully protect us from these massive bursts of solar winds called coronal mass ejections (CME).

The Earth is not going to end in 2012 however, and definitely not by CMEs. CMEs are potentially life-threatening, though not to such extent that it would result in mass extinction, and not directly. CMEs—if they hit Earth directly, which is a rare occurrence—cause aurorae to be seen down to the equator, and sends ultra-high voltage through power lines. This creates a strong magnetic field that induces another current in an on-going process. The current becomes so strong it melts the wires, which causes a release of a lot of energy, resulting the transformers to explode. All electricity disappears, and communication with it as satellites are burned out and drop down to Earth. The infrastructure will collapse. There will be no more tap-water and no more gas. This will stop transportation all together, and the stores will no longer be supplied. It goes without saying that a lot of lives will be in jeopardy. The last super CME struck in 1859. It disrupted telegraph service and created aurorae visible down to the equator.

The Sun is not the only star that is a potential threat to Earth though. In 2009, about 8,000 light years from Earth, in the constellation of Sagittarius a star was discovered that forms a potential threat to Earth. It's a Wolf-Rayet star called Wolf-Rayet 1O4. Wolf–Rayet stars are evolved, massive stars (over 20 solar masses), which are losing mass rapidly by means of a very strong stellar wind. Wolf-Rayet 1O4 however is different. It has a large cloud of dust around it. At the Keck Observatory in Hawai'i the star was inspected using an infra-red camera with multiple exposures to create a composited image. What the images revealed was a large tail that spirals the star. Scientists found out Wolf-Rayet 1O4 is in fact part of a binary system, meaning two stars orbit around their common center of mass. Binary systems are not rare. In fact, most star systems consist of two or even three stars. Both stars create stellar winds, but because they orbit each other, the stellar winds collide and gasses are compressed and dust and soot is created. The colliding winds do not spiral, but because the stars orbit around each other, a circular motion is added to the material which forms this long tail that spirals around both stars.

The fact that we can see this spiral is not at all good. Such massive stars like Wolf-Rayet stars live fast and die young ass they eject their mass so rapidly. They live for tens of millions of years rather than 10 billion years, which is the lifespan of main-sequence stars like the Sun. Wolf-Rayet 1O4 will die sometime in the next few hundred thousand years. At the end of its life it will go supernova, which means the massive star will eject its outer layers in a massive explosion while its core collapses onto itself and form a stellar black hole. As Wolf-Raeyet 1O4 is 8,000 light years away, its supernova is of no threat to us. So why is Wolf-Rayet 1O4 so dangerous, and why is the fact that we can see this spiral around the star a bad thing?

Scientists studied the orbit of Wolf-Rayet 1O4's tail of dust and came to a shocking realization. Its orbit reveals that we are looking directly at its axis. Some large stars—and potentially also Wolf-Rayet 1O4—spin incredibly fast on their own axis. When such a star collapses—provided its speed around its axis is sufficiently high—it produces a massive electric field. The energy from this magnetic field is turned into matter and anti-matter particles. As these collide, the most powerful energy beam in the electro-magnetic spectrum is unleashed: a gamma-ray burst. Gamma-rays bursts are devastating. The fact that we are looking right at a pole of Wolf-Rayet 1O4 means we are essentially looking into the barrel of a shotgun.  If this gamma-ray hits the Earth, it could deplete a large portion of the ozone layer, resulting in a mass extinction, as the five mass extinctions that preceded us.

## Time Travel

Journal Entry: Sun Sep 12, 2010, 11:01 AM

Time travel; is it really possible, or doesn't it go much further than the stories about time machines in comic books? In brief I wrote down a few facts, accompanied by a few of my own theories.

Time travel; it would seem impossible, yet we're constantly traveling through time. This is because gravity has a pull effect on the time dimension. If you're close to a massive object – let's say a pyramid – time will go by slower near you. The time difference however is so imperceptible that you might as well neglect it.

For a relevant time difference, you would have to fly around the most massive object known in the universe (or multiverse); a supermassive black hole. You would have to fly around it for a longer period of time, and with enough speed to ignore the gravitational pull, thus avoiding lapsing over the event horizon and getting sucked into the black hole and become an infinitely small point which we call the singularity.

Another theoretical way to travel through time is by traveling at the speed of light (or rather, 99.9999% of the speed of light). If you get in some sort of transportation that travels near the speed of light, double the Earth time (assuming the experiment takes place on Earth) will have passed relative to the time passed in your hyper-transportation. But why wouldn't this hyper-transportation travel at 100% of the speed of light, or even faster that that? According to Einstein's theory of relativity, no physical object, message or field line can go faster than the speed of light*, because the maximum speed of light is finite, and an absolute value. The speed of light in a vacuum is the same for all observers, regardless of their relative motion or of the motion of the source of the light.  What this means is that no form of transportation can reach or exceed the speed of light; not even by cumulative speeds. Now let's assume our hyper-transportation is a long form of transportation. Let's call it a hyper-train. If this hyper-train travels at 99,99% of the speed of light, and you drive from the back to the front of the train with a small yet fast form of transportation, you would still not reach or exceed the speed of light because the speeds are not cumulative. This is because at such high speeds, time slows down just enough to prevent you from ever reaching 100%. Still, the time difference you experience by going close to the speed of light can already count as time travel, even if the travel is not instant.

Every material – even a seemingly solid and smooth material – is everything but flat on a micro level. The material consists of a robust construction with gaps and holes. The time dimension is no different. There are tiny holes in the very fabric of time. These tiny holes are very, very tiny black holes through which you can travel through time. They're just too tiny for humans to be relevant. But what if that isn't necessarily true? What if it's because of these black holes that we can constantly travel forwards in time (though gradually as opposed to controlling the time's speed and direction [meaning backwards or forwards])? What if this is the very reason that we perceive time as being linear? What if this is the only reason that everything is dynamic as opposed to static and lifeless? I guess you could see it as a link from this moment to the next one.

Also, as you might know we are constructed of matter, which is the residue of the big bang. During the big bang matter and anti-matter collided and formed a violent reaction infinitely more powerful than a nuclear explosion or even a supernova. The matter and anti-matter off-set each other, but there was one in a billion particles more matter than anti-matter, which is why we are made of matter. What if anti-matter has its own time dimension which has an opposite effect of "our" time dimension? Just some thoughts.

## The Celestial Hierarchy

Journal Entry: Sun Jan 25, 2009, 6:39 PM

The following article is a an excerpt from my book "The Occult Labyrinth":