My Obsession with Type Alignment

image link-topic-sf0.jpg

It may seem that I am unnaturally obsessed with minute details of the alignment of type. I confess that this is true. Moreover, I can trace this to a specific event in my education as a typefounder. I wouldn't go so far as to call it a "near death experience," or anything so dramatic. Nonetheless, it was a significantly stressful moment for me. (Actually, it was much longer than a moment...)

The event occurred while I was apprenticing on the Thompson Type-Caster, in the first casting run of fonts that I did entirely by myself. (I didn't prepare the casting log and I didn't do the final fonting, but I did everything else - the responsibility for every physical aspect of the types cast was mine.) The project has a number of difficulties built into it. It was a revival of a 19th century face that had originally been cast to some unknown pre-point body size. We were casting it as if it were 8 point (more specifically, we cast 8 on 10 for the convenience of our customers). The original parameters for this type (line, set, fit) were unknown. (You can't necessarily trust the "official" line and set standards even if they survive, but they give a starting point.) No full specimen survived; we didn't even have a printed sample. The type was being cast from matrices made decades ago from then-existing types. No information from any previous casting survived. Finally, the typeface itself was rather curvy - nothing straight or square on it. All together these add up to only a problem of medium difficulty for an experienced typefounder, but I was just an apprentice.

Things weren't going too badly until I got to 'L'. You'd think 'L' would be easy, right? Nice big baseline. But type has no baseline. "Baseline" is not a physical thing inherent in type - it's something that we just make up. As typefounders, we make it up as we go along. The 'L' draws more attention to the decisions we make about baseline than any other sort. It should take a minute or two, tops, to align a sort. The 'L' took over half an hour. The adjustments were all in the thousands of an inch range.

And when I was finished casting all of these fonts we proofed them and inspected the proof... and the 'L' was still wrong. Toss every 'L' into the pot and recast.

What I learned from this, the hard way, was that in lining and fitting type thousandths of an inch matter.

image broadside-divider-sf0.jpg

This Notebook is a part of Making Matrices . But it is also one of a scattered collection of Notebooks in CircuitousRoot which, taken together, constitute a Heretic's Guide to Type.

Select Resolution: 0 [other resolutions temporarily disabled due to lack of disk space]