About optical fonts. There are some fonts that present more series of glyphs, divided by optical ranges ranges (generally three or four). For example:
Caption = {-8.9}
Text = {8.9-14.9}
Subhead = {15.0-22.9}
Display = {23.0-}.
Is the generation of these families somehow automatic?
Given that obviously it must take into account the intrinsic characteristics of the "starting" font (Text), which features should it possess?
Does the design of the gyphs fit in without altering the x-Height and C-height metrics, etc.?
0
Comments
From what I understand, these four font styles are cuts you need to create individually. I have never designed Caption cuts before, but in my understanding, they are more crude, somewhat stronger, have a higher x-height and sometimes ink-traps (which serve a similar purpose with pixels on a display as they do with ink, although pixels typically don’t bleed like ink).
As for the Display cut, that is usually a more elegant and thinner version of the Text cut (at least with serif fonts). In any case, I believe it takes a lot of extra work to produce these, but depending on the overall design, the type family, the market, etc. it may be well worth that extra effort.
Size-specific adjustments to type designs by Tim Ahrens and Shoko Mugikura (based on Ahrens’ MA thesis, Univ. of Reading 2007) is an entire book which offers considerable info on the subject. Highly recommended! https://justanotherfoundry.com/size-specific-adjustments-to-type-designs
Offering multiple optical sizes was inherent to the way fonts worked back in the foundry metal type days (15th-19th century and somewhat into the 20th). Each size had to be cut independently, and of course adjustments would be made; it would have been more difficult, not less, to keep the design identical and somehow scale it.
That changed in the late 19th century with the Benton pantographic engraver. Now it was possible to scale type without modifications—if you wanted to. However, the idea that one should make changes instead was well-established, and Benton (first with Benton, Waldo & Co. and later with ATF) helped establish the principles of doing so in a systematic way. “Cutting slips” for each typeface had coded notes about exactly what should be changed and how, when cutting a different size—but much was semi-standardized.
Here is a shot of ATF Garamond, from the Ahrens/Mugikura book:
ATF Garamond, from left to right: 6, 8, 10, 12, 14, 16, 18, 24, 72 pt
Phototypesetting made it so that one was literally using a single master across sizes, and that became the norm for some time in the late pre-digital period, and remained so with digital type (particularly early on). In those days, some optical sizes persisted, but they were rare.
From the 1980s onwards, digital technologies have offered ways to make optical sizes more easily than in metal, so some folks have been doing this. Variable fonts (and GX and multiple master before them) offer a means of making it more accessible to end users, as well as easier for type designers. But while a few foundries do them fairly routinely (including Adobe), I think overall they remain more of a high-end / boutique thing.
Anyhow, if you want to know more, go get the book! It is a fabulous resource. As I wrote five years ago: “This is not just the best, but really the only significant work on this intriguing and complex topic. Highly recommended for intermediate and advanced type designers, and anyone else interested!”
(Now to turn this into a blog post....)
Among other things, the installed base of existing fonts out there won’t change overnight. Retrofitting existing typefaces is possible, but not trivial.
And in some cases, when planning a new typeface, the multiplicative interactions of axes are such that one needs to carefully consider which ones to include, depending on the design. While important for serifed text faces, they are not vital for every typeface out there.
I just dropped the optical-size axis in a proposal of mine for a sans-serif typeface, because it would have entailed too many additional masters. And I did that as somebody who is in fact a big fan of optical sizes—they just were not as critical to this design as they are to a typical text case.
I will learn about the concept of axis-based type design, which is completely unknown to me ...
That isn’t using an optical size axis in a continuous way, but referencing discrete styles of standalone optically-sized fonts. Not the same thing.
When using non-variable fonts that are instantiated as separate fonts, needing to research which range each size is intended for with those particular fonts, and then manually coding each of them up front, in a way that is peculiar to the particular typeface... that is a fair bit of work.
Personally, I am looking forward to apps that just use the correct optical size automatically from the optical size axis of a variable font, the way InDesign does for MM fonts, and as defined in CSS (font-optical-sizing, as discussed by @Adam Jagosz above).
For typefaces aimed to print, it becomes even more so.
https://www.typenetwork.com/news/article/inside-the-fonts-optical-sizes
I am wondering if InDesign’s auto-optical size pref can be extended to work with variable fonts (and not just MM fonts)—or perhaps it already has?
https://mobile.twitter.com/Typelady/status/1150489276750209024
Resolution, device, contrast, colour, tone, background and good old paper stock are also factors.
Some of these issues are addressed by grading, rather than optical size.
Ultimately, it’s the typographer’s decision that counts.