Regular and italic with the same character widths

2»

Comments

  • Christian Thalmann
    Christian Thalmann Posts: 1,983
    I very much prefer the last one of these, Nick.
  • John Savard
    John Savard Posts: 1,126
    edited June 2020
    On the subject of gaining insights into the design of type from a width constraint: for unrelated reasons, I had been investigating various systems for proportionally-spaced printing.
    When I look at text printed with an IBM Executive typewriter, or the basically equivalent Friden Justowriter, what I see is text that looks better than monospace typewritten text, but is still clearly and obviously not as good as actual printer's type. The same is true with text produced on a Vari-typer, with the additional limitation that the widest letters are now not as wide as they should be.
    On the other hand, text produced by an IBM Selectric Composer doesn't really seem to have any visible issues caused by the coarseness of its unit system. The widest letters are slightly narrower than they should be, but the major reason it is possible to distinguish Composer-produced text from the product of conventionaly typesetting is that the alignment of a Selectric typewriter mechanism isn't as accurate as what is achieved in normal typesetting.
    ATF offered special typefaces designed around a coarse unit system for fast cold-metal typesetting. There was Self-Spacing Type and then there was Quick-Set Roman.
    The former used a seven-unit system; the latter, a five-unit system, but unlike the IBM Executive, it did constrain the widths of the largest characters, in exchange for less coarse spacing of the others.
    Both of those seemed to me to be nearly indistinguishable from genuine typesetting.
    In the early phototypesetting era, ATF got in the game with a device called the ATF Typesetter. It looked like, and doubtless was, a Friden Flexowriter with an additional mechanism bolted on.
    The first iteration of the ATF Typesetter used a five-unit system. The third iteration used an eighteen-unit system, just like Monotype.
    The second iteration used a seven-unit system. I've seen examples of the text it produces.
    You can too; on the Internet Archive are preserved the 1960 through 1965 issues of Ad Astra, the yearbook of Sarnia Collegiate High School, which were typeset using a seven-unit ATF Typesetter in its version of Baskerville.
    and, as well, I had also located an issue of the Rhode Island Herald which had a front-page story about the ATF Typesetter they were using for the newspaper, and which is dated before the 18-unit version came out, so this is an example of another one of its typefaces:
    Also, on my own web page, I have some information available at
    As well, I'll include this link which I provided here once before,
    to a page that offers a download of a brochure for the Friden Justowriter which includes a sample of their typestyle Rogers - which aroused my curiosity ever since I saw a mention of it in Alexander Lawson's Anatomy of a Typeface.









  • Sander
    Sander Posts: 12
    Thanks @John Savard. That's a lot of information, but it will clarify how problems concerning type design within predefined widths were solved at that time. Insights that may be very helpfull for my project. 
  • John Savard
    John Savard Posts: 1,126
    In that case, at the risk of subjecting you to information overload, I think I should also mention this link:


    which is to a catalogue of typefaces for the four-unit Varityper, which will show many examples of the design of different styles of type for the same constrained unit system.

    The IBM Selectric Composer is considerably less constrained with its finer unit system, but all typefaces are made to the same widths, and thus Italic and Roman match, and so you may be interested in its catalogue as well:


    Also, have you heard the story of how Microsoft's Arial typeface came to be? In the early days of laser printers, when they had their own fonts inside them, to assist in composing text documents to be printed in Helvetica on an HP LaserJet, Microsoft commissioned Monotype to take Monotype Grotesque, and modify it so that its metrics would match those of Helvetica.



  • Jens Kutilek
    Jens Kutilek Posts: 361
    Just something to keep in mind: When designing Roman and Italic on the same width, it’s not that the Italic always must accommodate the Roman design. You could make a compromise on both the Roman and Italic version of a hypothetical difficult glyph and achieve a better overall result.
  • John Savard
    John Savard Posts: 1,126
    Jens Kutilek said:
    Just something to keep in mind: When designing Roman and Italic on the same width, it’s not that the Italic always must accommodate the Roman design. You could make a compromise on both the Roman and Italic version of a hypothetical difficult glyph and achieve a better overall result.

    That is certainly true. In fact, my immediate reaction to your post was to wonder what prompted you to state something so obvious.
    In the case here, since the original poster has a novel and unique approach to type design, it is reasonable to conclude he is likely to be a newbie, not knowing everything that the seasoned professionals on this forum would regard as obvious. But given his stated goal of learning more about the design of letters in a typeface through constraining their width, it would seem to me that he will get around to trying different widths for every character in both the Roman and the Italic to find their consequences.
    Still, this raises an interesting issue.
    How do we even know that the general tendency of typeface designers, in the circumstance where the widths of Roman and Italic are to match, will be to shoehorn the Italic into the optimum widths for the Roman, and not the other way around?
    Well, for one thing, we would suspect it because in a document, text in the Roman is typically far more common than text in Italic, so optimizing the Roman is likely to be seen as more critical.
    For another thing, we have lots of evidence. When typefaces such as Baskerville, which have an Italic significantly narrower than the Roman, are translated to Linotype or the Selectric Composer, the Roman stays about the same, and the Italic is made wider.
    In addition to the Roman being more common, of course, another factor comes into play: in general, it is better for a letter to be too wide than too narrow, so that its details remain visible.
  • Sander
    Sander Posts: 12
    @John Savard  said:
    'In the case here, since the original poster has a novel and unique approach to type design, it is reasonable to conclude he is likely to be a newbie, not knowing everything that the seasoned professionals on this forum would regard as obvious.'

    This would imply that the seasoned professionals on this forum never have a novel and unique approach to type design. That would be a pity ;-)
  • Jens Kutilek
    Jens Kutilek Posts: 361
    edited June 2020
    There may be different ways of working, but for my biggest font project to date, I started with the Regular Roman style and finish it pretty much, then added a light and a bold Roman master and finish them. At that point I would consider my Regular Roman weight pretty much a given, because the other masters are already based on it. When I start the Italics after that, also beginning with the Regular weight, I would be really hesitant to change anything in the Regular Roman, because it has been there for so long, and underwent many iterations already, so it could be considered perfect ;)

    It really takes some courage and determination to throw away or change long present stuff at that point which will cause a long chain of necessary updates (accented letters, weights, ...).

Categories