Conrad Taylor (http://www.conradiator.com/) recently came to our office to share his experience of publishing tools used before the desktop computer was commonly available. In his presentation he described some old typesetting machines, among them the Berthold Diatype. At this point, I recoiled in terror as dark memories surfaced of a time when I operated this machine, created by what I still regard as some of the most evil minds on the planet. This got me thinking about some of the important advances that have been made in type design technology.
Today, our font developers use Multiple Master (MM) technology to create multi-weight font families. This process allows the designer to draw the extreme weights of a typeface, say Thin and ExtraBold. The designer can then interpolate any point in between, and relatively quickly achieve a large range of weights. The interpolated weights have to be manually optimised to ensure they perform exactly as intended, and the larger the difference between the extreme weights, the more manual work is required.
MM fonts were first pioneered by Adobe around 1992. It was a great idea that ultimately failed as a font format. Users barely understood how to use normal fonts on their Macintoshes at that point, and support by programs and printer drivers was erratic. However, it also provided opportunities for those of us who were tenacious.
In 1994, I was commissioned by Paul Luna at Oxford University Press to design a titling font for the Oxford Dictionaries, the Thesaurus and other publications in the same series. The aim was to have a typeface that could always be set at the same type size, irrespective of the length of title. MM fonts were the answer to the problem, and fortunately, Fontographer 4 had just been released allowing the creation of MM fonts.
As I was working on the project, it became clear very quickly how buggy the implementation was. Over a period of about two weeks, I was in daily contact with Altsys, the developers of Fontographer, to provide test reports. The engineers would release a bug fix overnight, and in the morning I would dial up with my modem to download the latest update, a process that took several hours. Eventually, however, I had a reasonably stable version that allowed me to finish the font successfully.
The resulting typeface was a two-axis MM font, weight and width. It consisted of caps only, sufficient for the setting of the titles. When working with MM, it is important that a glyph in the various extremes – in this case four – has a design and digital compatibility. The number of nodes, their numbering and direction of digitisation must all match. Not doing so will result in some unexpected interpolations. Font developers now have automated tools to help them achieve this compatibility, but I had to manually ensure that all the glyphs across the four designs resulted in the same structure. This manual compatibility check had to be done not only across the glyph design, but also across kerning and other font related data. To say that it was a painful exercise is an understatement.
I think that this was the first custom MM font ever produced in Europe, and I’m proud of the result. It’s a nice typeface that fulfilled the brief. It was used for a number of years by OUP, but I guess once the MM format was no longer supported, the font disappeared back into the drawer. Creating it was a great learning experience and it felt like pioneer work.