This topic was imported from the Typophile platform

If you find yourself with a glyph where the counter doesn't quite match up with the outside form right, is your first impulse to:

A, edit the counter

--or--

B, edit the outside shape

Thinking liminographically, I would say: it depends.

hhp

Step back and try to see the structure of the letter, rather than the two outlines. This should give you an indication of where the problem is, and what needs to be adjusted. Remember, when we perceive letters in text we perceive the structure, not the outlines.

John, your "structure" might be understood as "expanded skeleton" (which I feel is irrelevant). But even if you mean structure in terms of black bodies, I think it's dangerous to think in those terms; to me it's all about notan. So yes, don't look at the two outlines... but that's because it's not two outlines; in a way it's actually one -potentially complex- outline; or maybe no "outline" at all... To me it all boils down to the "total" black/white border (which even exceeds the letter boundary).

hhp

I say, if you don't know it must be an o, for only the first counter encounter should raise such a fundamental inside outside question. If it is the o, then it depends entirely on what you want the rest of the letters to look like.

Actually this occurred to me when I was editing a figure 4, with all straight lines.

Please don't forget to vote guys.

Two bad choices? What a déjà vue... :-)

hhp

Everyone is telling you: your question can't we answered in the abstract, as it depends on the context of a design. In other words it's the wrong question to ask if you want to improve a design.

A good new typeface has: A. a good design idea--something it's trying to do that makes design sense; B. It is carried out consistently; C. It is carried out well artistically. Often you don't know whether your design idea is good—or whether you can make it good—until you try B. and C.

Inside and outside curves and how they relate should flow out of your design idea. And that can only be judged in the context of, say half or more of the alphabet (eg hamburgerfontsiv). Then you can ask: was this a good design idea? Should it be changed, or is some of the execution flawed that I should fix before going on? If some characters are flawed, how?

I don't work impulsively.

I would stare at the letter for a moment, then I'd go away and do something else -- Returning eventually, possibly with a fresh perspective.

p.s. As you are troubled by the 4, I should say: numbers follow different rules than the latin alphabet, but they also have to work together with the alphabet, so they're a special challenge. Perhaps others can be more helpful with general guidelines, but for a start: study a variety of numbers styles and their associated alphabets.

What if you're just making a single glyph and thus have nothing to compare it to? I just think this is an interesting question. Kind of like, if you are at a T intersection somewhere, and are completely lost, and have no landmarks to go by, do you turn right or left? There's no correct answer, which is why I think it is so interesting. I wasn't really looking for help with anything in particular, but I do appreciate it.

You wait for another car and follow it.

hhp

Not a bad solution.

Then one day, the boy was out hunting english script fraction denominators in the woods, when suddenly a huge black wolf with long white descenders leaped from an over-hanging ligature...

Hrant: your "structure" might be understood as "expanded skeleton"

Which would be a mistake. If I'd meant skeleton, I would have said skeleton. What I am talking about might be considered the opposite of an expanded skeleton: a perceptual shape within the black. It is easy to see how this shape is perceptually subject to the relationship of opposing outlines, without falling back on a notion of these as expansions of a skeleton. As I've pointed out in the past, Legato possesses a strong sense of structure that is not reducible to a skeleton through manipulation of the outlines, which would be the case for skeleton expansion (or stroking).

To me it all boils down to the "total" black/white border

It cannot boil down to that in the case of text type at typical sizes, because at those sizes we don't perceive the borders, we perceive the shapes.

Agreed - I meant in terms of definition/design, not perception.

hhp