Ben Mitchell's typo blog charting the excitement, activities and challenges of my 12 months' studying the MA in Typeface Design at Reading University.

Now with occasional ramblings about type-related things I find interesting.

Opinions are all my own.

This experimental sans was something I drew in my sketchbook whilst travelling in 2010. The concept behind the design was to see what would happen if black and white shapes became somewhat disconnected. Counterforms are made of perpendicular straight lines while outer edges are wide superellipses.


Despite its heavy weight, Argon maintains its counterspace areas in two special ways. Firstly, junctions of straight lines are drawn apart (as on M or N), which allows the counters to retain their full height. Secondly, joins between curves are given a pronounced notch or ‘bite’ out of the junction (as on B, R or 8). The result is a spurless, retro-futuristic display face.


After drawing the Latin in FontLab from sketches, I decided to try the same design with Thai script.


The family will include a couple of avant-garde outlined inverse cuts (shown below), possibly working as a layered font for «double foreground» special effects.

Posted at 1:08pm and tagged with: Argon, In progress, Thai, Latin, font, type design, typography,.

Our first term is over, though it feels as though we’ve all only just settled in. The ten weeks have passed so quickly, in a flurry of workshops, conferences, seminars, critique sessions and typographic delights. I’m feeling lucky to be able to spend this year doing something I enjoy so much at a department with such a great name.

Progress on my typeface is going well, I think. I had the chance to talk to Fiona about my Burmese letters, and we agreed that there was definitely room for improvement. The thing that puzzled her was the inconsistent stroke modulation (see image) which didn’t seem to follow any pattern.

Referring back to the images of old manuscripts and Burmese folding books, we noticed that the heaviest parts of the stroke were often at the tops of the letters, and followed a pattern consistent with the pen-tooling of other Indian scripts, namely having the pen angled the opposite way from the normal Latin model. Unfortunately in my enthusiasm to create something new and exciting, I’d put the stress all over the place, and had to agree that whilst the letters might look interesting, they wouldn’t do a very coherent job at forming words since the eye would be drawn up-down-up-down, rather than along the reading line. Worse, the overall texture of the paragraph would be rather spotty. I’d been seeing the glyphs primarily as cool shapes, instead of as word components. I guess that’s about the worst thing a type designer can say!

Fiona suggested looking at some more Burmese manuscripts, in fact as many as possible, to try and work out the best model to follow. It’s also helpful to look at other south and southeast Asian scripts. The letters of these are typically drawn hanging from the headline (see image above), rather than sitting on a baseline, which lends further support to the idea of stressing the tops of the letters. In fact I noticed with curiosity on one Burmese folding book that the letters had been drawn hanging from a faint ruled guideline. This came as no surprise to Fiona: “It’s a Brahmic script”. I need to be faithful to that ancestry to make it authentically Burmese. Strangely I didn’t feel discouraged by the prospect of starting again: it’s the first time I’ve tried drawing Burmese and if my first Latin is anything to go by (below), first designs are never really very clever! That it will also look more harmonious and read more smoothly only makes me more excited to revise and redraw.

(My first attempt at drawing letters from 2007. In this rather unassured design — just look at that skinny f! — I was seeing the shapes as discrete entities rather than drawing them to fit well together in words and actually paragraphs. You can see I was drawing them at large scale on the screen as the serifs are tiny.)

Anyway, back to my Latin. We hadn’t had group critique with Gerry for a very long time, so people had generally made a lot of new things to show him.

I was quite certain my serif cut was heading in a good direction; and now having had Gerard’sspacingletterfitting workshop twice, feel much more confident about the whole thing. Also I should mention one of last year’s students, Julián, who’s been working in the department, has been very generous with critique and advice about spacing. One piece of advice was that the space between two lowercase /o/s should be about the same as the lowercase stem width. This provided a foundation from which to overhaul the spacing, and I think it really works.

Gerry seems to know what sort of feedback is useful. For me, it’s mainly been things I’ve overlooked or haven’t noticed, rather than raising stylistic questions about what makes an /a/ an /a/ for example. My odd rectangular serif-terminal features on the /a/, /c/, /s/ and /z/ weren’t all the same shape, sometimes being square and other times trapezoid, so that needed attention. The same shape on the diagonals made letters like /w/ too dark — I’m still not quite sure how to resolve that. My /x/ had a disjointed appearance where I’d offset the thick strokes too much at the centre. Here’s progress on the serif face.

One recurring difficulty I face is determining the different amounts of overshoot on the x-line and baseline. My glyphs seem not to align properly, sometimes floating above the baseline, sometimes dipping below, and sometimes looking too tall or short. Although I have learnt how to see the problems, I haven’t yet discovered how to correct it, and end up overcompensating every time.

I was delighted to hear Gerry quite liked my sans serif design (above, with first showing of capitals). Gerry often explains the trouble with sans faces is there are fewer interesting design decisions that can be made. Along with the lack of serifs, there is much less scope for being creative with contrast and stress. This leaves only the overall width and shape of the strokes and terminals to play with. So it was really encouraging when Gerry said there was something very interesting about my design. Although it’s a monoline design, I’ve kept a bit of emphasis at the tops of the letterforms, in the hope that the Burmese will follow suit, and I’ve also kept the noticeable thinning at stroke junctions. Unfortunately, I’d gone a bit boring in the /h/, /m/ and /n/ and needed to ‘do something a bit more interesting’, so I redrew these with more movement and contrast. There was also a difficulty with the crossbars of /f/ and /t/ because they produced too much density along the x-line, where in the rest of the face, that’s quite a light spot. I tried a couple of different solutions to this but ended up following Gerry’s advice to taper the strokes on the left side. I may come back to this in the future. Initial feedback from the class and other type designers has been very positive. I can’t wait to try making the bold!

Paragraph sample showing problem areas. The /x/ with too much offset, also too wide, and the /m/ and /t/ with too much space on the right.

Posted at 2:37pm and tagged with: Burmese, Latin, design, font, sans serif, script, serif, typeface, writing system,.

For the past two or three weeks I’ve been bashing out ideas for my practical project, the type family designed especially for Burmese dictionaries. As I blogged before (Thoughts on a Brief and Exploring Burmese), I’m aiming for a fresh, lively and cheerful design, in which the Burmese and Latin scripts harmonise well.

My first scribbles with paper and pencil (above) looked like a strange imitation of Charter, Miller or Times New Roman. I adopted round ball terminals on the /a/ and /c/, in a crude attempt to reflect the circular forms of Burmese letters. The problem was it’s boring. The vertical stress wasn’t providing enough dynamism, and the flat serifs seemed too modern and fussy. Overall it looked restrained and dull, not at all the sort of thing I wanted to spend the next eight months working on. However, I wasn’t quite ready to throw away my sketches as something about my /n/ seemed to show some interesting potential.

I’d subconsciously picked up on the idea of designing the whites, the counterspaces, before the blacks, the strokes, and this had meant I was focussing on the white notch above the join. Making it large seemed to have two benefits: reducing clogging at small (dictionary) sizes by lightening the junction, and pushing the upper curve forward creating some nice movement and liveliness, giving a nice ‘legato’ effect to letters. In addition, the shape of the trap seemed to represent the path of a bounce, accidentally perfect to suggest liveliness.

A second attractive feature of the /n/ was that the inner curve of the arch didn’t have so much speed in the curve — all the action was in the outer curve — allowing me to indulge my interest of dissociating the two sides of the stroke, in a Dwiggins or Bloemsma* sort of way, producing something unachievable by a real tool. This led me to further research Bloesma’s logic behind Balance or Legato: his solution was to link the letters by rotating the counters, but what if there’s another way?

I began to sketch again (below), trying to incorporate the active stroke movements and flow between letters. A solution here was to make the serifs flow out from the stems, rather than looking like bars across the bottom of strokes. And they needed to become slightly asymmetrical to emphasise the desired forward movement.

I completely redesigned the /a/, actually by reverting to an unmodulated, sans serif design, to see how the underlying architecture could work, and then adding the contrast. With the /e/ and /s/, I tried to make the inner sides of the strokes less round, directing the eye along the reading line to the next letter, rather than curving all the way back into the centre of the letters (below). And yes, that /s/ isn’t quite there yet.

I’d also been playing with Burmese letterforms (below), thinking about how some of the details of my Latin design could be transferred over. I wanted to add a small amount of stroke modulation, and thought the terminals especially would be an area in which some clever solutions could make my design more unique. After all, I want to make something that isn’t just a set of geometric circles. Adding some humanist or calligraphic touches seemed reasonable.

So far so good. And then I presented my design brief and font sketches to Gerard Unger, thinking I had a solid basis for a type family. Instead, Gerard suggested, where I had been considering my Burmese as a non-Latin script, I could consider my Latin as a non-Burmese script. Oops! But what does this mean exactly? Well, start by looking at the Burmese letterforms, thinking about how the strokes are made, with which kind of tool, and what effect that has on the appearance of letters. Sure, they’re circles, but they’re also monoline, and can be rotated or squoosed to look Latin. Not that he suggested I mimic the Burmese details in the Latin, but that it could be instructive to think about the angles and roundness of Burmese when constructing the Latin. He suggested I study Futura, VAG Round and DIN Rounded. An interesting idea, I thought.

Feeling confused by rounded, geometric sans serifs, which I couldn’t link in a meaningful way to a typeface for small text in dictionary columns, I dutifully went back to my Burmese sketches whilst looking at DIN Round. Hm, just what I expected. Monoline, geometric shapes, a real spanner in the plan to make my Burmese more humanist.

But then it struck me I didn’t want to make the Burmese dictate the style of the Latin, in the same way Gerard was eager for me to avoid simply adapting my Latin into different stroke patterns. What I want is to give both scripts equally strong identities in the family, neither dominating the other, and with influences going both ways. After experimenting with Thai letterforms, I also want to ensure the two scripts are clearly distinguishable on the page, rather than blending into each other and getting swallowed up.

A potential solution occurred to me unexpectedly. Whilst looking back at my Burmese, thinking about which letters should be my starting blocks, I realised the importance of writing them on paper: Burmese words are written left to right, but the strokes are drawn in circle sections that go both clockwise and counter-clockwise. Some letter circles are started at the bottom and go clockwise, and some are drawn anti-clockwise from the top. And some letters are constructed in sections: strokes in opposite directions. Gerard’s suggestions were making me think.

Whilst considering the relevance of this, I started sketching the sans-serif version of the Latin letters, which seemed to be the most promising way forward from Gerard’s suggestions. So what happens if the sans version took a little pinch of Burmese roundness, and a little of its geometric construction, and of course the more monoline approach would suit a sans face better, and what happens if into that mix, we start constructing strokes backwards as well as forwards? What if we start a /c/  or an /s/ at the bottom? Can instrokes and outstrokes be flipped on their heads?

To be continued…

*(Lots) more on Bloemsma’s Legato over on Typophile.

Posted at 6:44pm and tagged with: Burmese, Drawings, Latin, Monoline, Unger, construction, one column,.