You are slashing your way into the Papa New Guinean jungle...

Nick Shinn

Not to disparage the brilliance of Optima, but it was anticipated by Stellar with regards to what I assume you’re attributing its originality.

John Savard

Yes, that's true. Optima is still "obviously" original, though, because Stellar isn't all that well-known.

Was Optima more successful because it was all that much better, or because when it came along, in the 1960s, it caught the spirit of the times, as people were looking for things that were "modern"?

That's likely a question that can never be answered.

Craig Eliason

Nick Shinn said:

Not to disparage the brilliance of Optima, but it was anticipated by Stellar with regards to what I assume you’re attributing its originality.

Not to mention some 15th-century carvers (for the caps anyway). 

John Savard

Of course, though, Hunter Middleton's Stellar was principally designed for use as a display type for advertising. Optima, on the other hand, was able to be used for body copy, and that was true because it was designed to be suitable for that purpose.

Imagining that a flareserif would be suitable for such use was very bold and original.

notdef

https://patents.google.com/patent/USD24683

John Savard

Frode Helland said:

https://patents.google.com/patent/USD24683

But that only anticipated Radiant, not Stellar and Optima.

Craig Eliason

John Savard said:

Of course, though, Hunter Middleton's Stellar was principally designed for use as a display type for advertising. Optima, on the other hand, was able to be used for body copy, and that was true because it was designed to be suitable for that purpose.

Imagining that a flareserif would be suitable for such use was very bold and original.

Interestingly, a 1960 recounting by Zapf indicates that it was not designed as a text face, but that potential was realized after the fact:

"A design for Optima was ready by 1952, but it took until 1954 to complete the models in all essentials. A conversation in October of that year with Monroe Wheeler of The Museum of Modern Art (New York) proved so convincing that I changed my original notion of Optima as a display type more into one of a text type; and now that I see it complete, I believe this type can be profitably used with photos and industrial presentations in books and catalogues. The result was an eminently practical roman with a destined area of utility in books of art or photography, in technical or scientific publications, and in children’s books and periodicals." 

Chris Lozos

When I was a sophomore in Design School at Carnegie Mellon, Zapf came there for a meeting with the Hunt Family about their library face that he was working on then. While there, he gave a presentation to us at the Design School and showed us many drawings of his work, much of what I remember was Optima.  Watching him draw freehand letters 3 feet high on a chalkboard was thrilling. I didn't know a piece of chalk could do such  things in seconds.

Nick Shinn

And continuing further from the OP…

Stellar (1929) was available from the Ludlow foundry in Regular and Bold, from 8 pt. to 72 pt., according to Vincent Steer’s Printing Design and Layout, 1945, London.

Here it is used for head, deck and drop cap, with Vogue body text, in the American satirical magazine Ballyhoo, 1932.

notdef

https://americanhistory.si.edu/collections/search/object/nmah_791422

notdef

http://www.woodtyperesearch.com/revivals-of-revivals/

notdef

But back on track: I like your off kilter caps, but to make them truly wrong you’ll need to know what is perceived as correct – i.e. break the rule consistently. Since a typeface is a tool for presenting any text with the same visual expression, each element needs to follow the same rules. If they did not, you would run the risk of different words having different expressions.

The classical capital proportions originates with a geometric pattern of elemental shapes: triangles, squares, circles, and parts thereof. There are schemes to look up, but they can only get you so far. From there, it’s just practice-practice-practice and a lot of looking at things.

Even experienced type designers make amateur mistakes with unfamiliar characters. It is very hard to know how far you can modify something before it looses its essence unless you’ve been exposed to relevant models before.

Craig Rozynski

Karen Cheng covers proportions in her book Designing Type. 

She categories capital proportions broadly into Classic and Modern. 



She writes "Classic proportions are based on ancient inscriptional models. For both aesthetic and practical reasons, the Romans used a square or geometries of a square (golden rectangle and the root five rectangle) for the widths of capital letters...

Although the Roman capitals are beautiful and graceful... they produce capitals with uneven color – the narrow letters can be darker than the expanded forms. In contrast, modern proportions prioritise the goal of even color."

A detailed letter by letter analysis then follows:

Nick Shinn

Karen Cheng is wrong to play favourites. If anything, Roman capitals have more even colour, and modern proportions less.

In particular, the greater interior space in the wider classically proportioned letters M and W balances the space between letters, whereas in the modern, the cramped M and W are intractable.

In this example: top, no kerning; bottom, foundry kerning.
I’ve adjusted type size for equal line lengths.
Adobe Trajan Pro, Bauer Bodoni.


In this next comparison, URW Futura and Helvetica Neue, note how the RAW in Helvetica bunches up weight. Again: top, no kerning; bottom, foundry kerning.


(Of course, one should no doubt kern to the text at hand, beyond just immediately adjacent letters, and probably letterspace all cap text. Also, I didn’t put much thought into the choice of these sample words—Hemingway and Lawyers are two awkward-to-space words I often use when developing types, and Typedrawers, well, here we are.)

John Hudson

See my previous comments regarding proportion of uppercase letters in mixed case text vs all-caps text. Most typographic uppercase letters are designed primarily for use in mixed case text, where greater commonality of width contributes to the overall page texture. So Karen Cheng’s comments make some sense if talking about ‘even colour’ in text, rather than all-caps.

Futura is a really interesting typeface in part because of the classical proportions of the uppercase letters, which work so well in all-caps settings but less well, I think, in mixed case text. This seems especially evident in German, which uses so many more uppercase letters than other languages: a page of Futura text has an unusual texture. It's not bad, per se, just quite different from the same page in types with more typically consistent uppercase proportions. Well-spaced Futura all-caps can be glorious. I remember walking down a restaurant corridor towards a black door with tracked out, perfectly spaced metal Futura letters spelling WASHROOM and thinking it was one of the most elegant pieces of typography I had ever seen.
_____

PS. The all-caps display face that I showed in my earlier post has now been published:
https://github.com/TiroTypeworks/Castoro

Chris Lozos

Arnold Bank taught us Roman proportions by pen widths, not surprising since he was a calligrapher.  The pen width also brings stroke weight to size ratio, trapped spaces vs open spaces and proximity to other forms.  I am personally less inclined to look for some magic calculation that can oversee all the variables than I am to draw the letters as they seem to fit best together.  To that end, there are many paths and different solutions.

John Savard

Given that, in another thread, we saw that even Gutenberg used a unit system, presumably so that he would not have a problem justifying lines, even if the Golden Ratio was an ideal, I doubt that irrational numbers were actually employed in the proportions of letters for typefaces. Of course, 7/5 is close to the square root of two, and 8/5 is close to the Golden Ratio.

Matthijs Herzberg

Noordzij writes that neither classical or modern proportions in capitals bring an even texture, as capitals are far less modular than lower case, and when they are modular they are stacked vertically rather than horizontally. For instance, a /B is two /Ds stacked on top of eachother, whereas a /b is an /l and an /o next to each other (approximately).
The only way to give capitals an even texture is to significantly alter their original forms, and the result of that would not be so classy either. And what gives? A perfectly even texture becomes far less important in one or two lines of titling compared to a whole page, and caps are ill suited for the latter for many more reasons than their naturally irregular color.
I agree with Chris Lozos’ ambassadorial assessment to draw letters as they seem to fit best together.