The roundest round.

JoelMarsh

Hello!

First post. Experienced designer (UX), but new to type design.

Deceptively simple question: What corrections are needed to make a stressed/contrasted O look as geometrically circular as possible?

The O in most fonts look (to me) a little taller than the width, and when I have been playing with it, there seems to be a fine line between circular and superelliptical. I don't want to cross that line. 

What should I be looking for? What effect does the axis have? Does the amount of contrast change the corrections required?

Johannes Neumeier

Nothing like trying and seeing, and about seeing it all is. One thing to keep in mind is size, since what looks exaggerated or too shy in one size might look just the opposite in smaller or larger.

Are you working on a round type yourself or are you merely interested in what makes round shapes look perfectly round? If the former is the case, you might also post examples of what you have tried yourself

Also a moderator should probably move this topic to a better suited category.

JoelMarsh

Sorry about the category. I noticed that as soon as I had posted it and then I couldn't find anywhere to change/delete the post. Dear moderator, please help!

One of the struggles of the beginner is that I don't know why I am trying things, and that bugs me. It's hard to know what I'm not trying, because I don't really know what I am trying. It's all guess-work.

I attached a very small sample of O's that I have made to see what looks "round".

Generally speaking, the axes are vertical, for no reason other than ease-of-experimentation, because I don't have an opinion about it (hence, this discussion). 

(I think) I understand all the basics like overshoot, making horizontal strokes slightly thinner, pulling out the "corners" a bit, and that a circle looks more even when it is actually a little bit narrower than the height (please correct me if I am wrong!). But typically those lessons are taught with very little contrast as an example.

But is there more? Contrast and axis seem to add more complexity to the mix since the inner counter the outer edge push and pull against each other. Teach me, Johannes, teach me! (and thanks!)

Ori Ben-Dor

The rightmost one looks superelliptical to me.
The other ones look elliptical in the sense that the width looks smaller than the height.
If I had to choose which one gives me the strongest impression of a circular skeleton, it would be the second from the left.

>> The O in most fonts look (to me) a little taller than the width

Just to make sure: you understand that there's no problem with that, right? It's probably intentional.

>> Teach me, Johannes, teach me!

One thing that I've learned from designing and studying typefaces (and it took me years to learn that) is that it's impossible to reduce the question of optical corrections (or more generally, the question of the "correct" shape) into a set of precise, well-defined rules. In the end it's all about subjective impression. The eye test is the only test.

What I can suggest is that you take each model from your jpg and create a gradient of variations that differ from each other by the width-to-height ratio, from a noticeably wider-than-taller variation through a noticeably taller-than-wider one. One of the middle ones will look to you most round, and that's the one you're looking for.

JoelMarsh

Thanks for answering Ori. 

Yes, I understand that there are a wide variety of shapes for an O, and that the best design for a typeface depends on the overall plan of the designer and that there is no "correct" shape or set of well-defined rules for optical corrections.

But, respectfully, this is also a problem with asking questions about type design: you get answers like "you'll know it when you see it"— which actually doesn't help that much. It would have taken me a long time to realize that extending the "corners" of a circle make it look a little rounder to the eye... geometrically that is less round and that is very counter-intuitive. But I read it in a lesson, so now I know that it's something I can experiment with, and I do!

But, is that it? Are there other aspects of an O that I could consider?

It's interesting that you see a circular skeleton in the second-from-the-left (and was the kind of answer that I was hoping for, so thank you). #2 and #3 (from the left) probably deviate the most from a geometric circle, and have the same outer edge shape, but with different counters.

Thanks also for suggesting a method of exploring this. I would be very interested to see other people's gradients and "most circular" choices to get a sense of what more experienced eyes see!

Thanks again.

Ori Ben-Dor

>> you get answers like "you'll know it when you see it"

That's not exactly what I'm saying. There are thumb rules that are worth learning, like the example you've given, that extending a little the "corners" of a circle make it look more round.

Such thumb rules help you start from a better starting point and guide you what to look for.

What you're looking for, I think, is an answer to the question 'how much should we extend the "corners" of a circle in order to make it look most round?'.

What I was trying to say is that there are probably no definite answers to such questions. Or maybe there are, but the model would be so complex, that it wouldn't help us much.

Anyhow, the gradient method I suggested can replace that kind of questions. Instead of trying to come up with a value analytically or something, we can simply try all possible options with our eyes. The one that looks best is the true answer. That's the point I was trying to make.

Chris Lozos

Joel, bear with us. There are rarely direct answers to such questions.  Learning to see is paramount to learning to design.  Most times, you just have to put in the hours and frustration of pulling and pushing until we learn to see, not just look. Ori was trying to help you. Go ahead and fight with the curves on your own for a while so that it will be easier to communicate about it.
Your concept of "round" is also going to vary with what you put next to it. If you put it between 2 squares, you will see a difference than putting it between 2 circles. This is what separates type design from most graphic design. You cannot escape your neighbors.

Ray Larabie

Karl Pilkington's head is the roundest of all possible rounds.

Michael Jarboe

Mirror, mirror, on the wall, who's the roundest of them all?

JoelMarsh

Ok guys, thanks for the help. Maybe I am expecting something that isn't the right thing to expect.

In user experience design we deal with a lot of "perception versus objective truth" types of design issues, and it is very common to hear people say that there are "no rules" even though there actually are. The "rules" are all relative though, rather than "4px will do the trick", which makes them seem elusive and constantly changing to the untrained eye/mind.

I am not saying this is the same, but given that context, maybe you can see why I find it hard to take these things at face value.

Thanks for bearing with ME. I'll follow your advice. 

Also, if anybody releases a typeface based on Karl Pilkington's head, I promise to buy it.

Vasil Stanev

Ray Larabie said:

Karl Pilkington's head is the roundest of all possible rounds.

Ohhh, you've touched a lung there!

Beau Williamson

The 'o' in Futura is often given as an example of a 'perfectly round o'. You could look at how the reality differs from perfectly round to get some hints.