Nathan Willis Letter fitting is usually described as a qualitative facet of type design, even though the production of digital fonts are boiled down to numeric side with bearings and kerning adjustments. Moreover, the qualitative nature of letter fitting makes it challenging to compare the letter fitting of typefaces or to assess spacing problems in objective terms. This talk will present an approach to empirically measuring successful letter fitting, developed as a part of my PhD research into algorithmic letter fitting at the University of Reading. It will outline the strengths and weaknesses of several general approaches to assessing letter fitting, and demonstrate the test framework that attempts to balance the inherent trade-offs to produce a quantifiable measurement more suitable for real-world testing.
The talk will explore how testing can balance the need for expert opinion with the wants and needs of the general reading public, particularly when the latter group may not possess the trained eye expected within the type-design industry. It will also examine how the design of test specimens and typographic styles impact the conclusions and degree of confidence that we can draw from testing. It will conclude with a re-examination of the fundamental question of assessing letter fitting in type design. How much is measurable, and how much can we, as designers, remove ourselves from the assessment of our work?
This talk does present some aggregate data from my empirical trials into quantifiably testing letter fitting, but my intent is to focus more on the underlying problem of measuring design in an empirical fashion.