Adventures in Typeface Creation
Fonts on the brain
After watching the documentary Helvetica in a university design class a few years ago, I had a thought: while the use of a typeface superficially changes the legibility and tone of whatever you are reading, it can more profoundly be associated, for better or worse, with our past experiences. For example, Comic Sans may evoke multi-coloured primary school learning materials, making its appearance in higher education all the more baffling and undermining.
What proportion of this “je ne sais quoi” reaction to a new font is based on its similarity to a font you like? Is it the same for the inverse? Is the all-powerful Golden Ratio involved somewhere? I couldn’t even attempt to answer those questions, so I ended up just trying to design my own font.
Design guidelines
Before I began, I thought it necessary to choose and adhere to some basic guidelines for consistency:
- Sans-serif (sorry, Medium) and mainly for display (signs, not text)
- As geometric as possible: use fixed-width rectangles and tori, along with various tangents and segments
- Round dimensions: multiples of 50 pixels
- Efficient glyph-making: rotate, flip, shear and merge components to save time and effort
The software
From my tumbleweed-infested iTunes, I downloaded Glyph Mini for £34.99. This is in comparison to other software (including the full version) that cost a few hundred pounds.
Despite this, the experience was not bad at all. A tip is to begin with the letters that make up the word ‘adhesion’ (minus the ‘s’ because it’s tricky). For any engineering students reading, Glyph is not unlike CAD software, though little things did annoy. For one, I missed the automatic snapping and alignment; it’s sporadic at best in Glyph. Subtracting shapes also needed a bizarre workaround, although apparently and unsurprisingly, the full version supports it.
The creation of letters, numbers and an assortment of punctuation essentials took about 15–20 hours spread over 4 weeks. As each letter and symbol is a separate glyph in the software, it’s fairly easy to minimize the window, forget about it for a bit and then resume as procrastination beckons (maybe this could be the next App Store sensation, Font Atsume).
Learning the kerning
Kerning is where late-on, the white space between certain pairs of letters is tuned for a more visually appealing result. For example, look at ‘WAVE’. I’d imagine there is an algorithm out there somewhere that could do all of this automatically, but manually tweaking various combinations can get addictive…
It’s all about perception with kerning, almost like designing an optical illusion. This ties into another tip I learnt: the height of a curved top should be 10–15 pixels higher than that of a flat top in order to seem the same!
Final thoughts
This exercise resulted in a couple of reflections:
- The high prices of fonts outside of Word and Pages are probably an accurate reflection of the manhours and licensing required for professional use
- Once machine learning is combined with fontmaking, this industry could be transformed. Custom fonts presented to you based on your personality and interests perhaps? This article by Erik Bernhardsson is fascinating.
So here’s a parting screenshot. All enquiries welcome, of course.
