Understanding the æsthetic of math typesetting

Ulrik Vieth
Stuttgart, Germany
Play (49min) Download: MP4 | MP3

One of the core strengths of TeX is the ability to typeset math to very high level of æsthetic standards. However, this level of quality not only depends on TeX alone, but relies on close interaction between sophisticated algorithms (built into the TeX engine) and the fine-tuning of metric information (built into math fonts), which is not so well understood.

At a previous conference Bogusław Jackowski presented a paper Appendix G Illuminated, in which he translated the formal description of TeX’s algorithms for math typesetting into a visual representation, illustrating the mathematical and geometric relations between the various font metric parameters. While this helps to improve the understanding, it doesn’t resolve the question how to determine good values of font metric parameters when designing a new font.

In this paper, we analyze the values of these parameters in existing fonts and draw some conclusions about the underlying design principles. In the end, we hope to obtain a recipe how to determine good values of font metric parameters based on simple design parameters such as the x–height or rule thickness.