Overbold: Ignoring Eric Gill's Advice

Christian Thalmann

Here's an experiment with a slim-stemmed /psi/. It looks way too ephemeral, IMHO. Hard to make anything of that bottom with such limited descender space. The second /psi/ is my new frontrunner — it's less of a three-master and more of a slashed bowl like in handwriting.

I tried a curly /omega/, but it feels rather out of character. The right one is an attempt at mimicking the expected ductus without resorting to an actual loop, and while maintaining weight.

I tried new and more angular versions of /xi/ and /zeta/, but I'm not sure I like them. The old versions had some convincing brush logic to them, which made them click even though they're objectively more ooey-gooey than the other letters. I often find those two letters to look the most scripty in other Greek typefaces, though.

EDIT: Aha — cleaned up connections combined with the original rounded horizontal strokes (rightmost version)? Works for me.

Thomas Phinney

That last rightmost one is definitely better.

The second omega, without the hollow in the implied loop, is working better, I think.

This is quite a challenging style to do Greek for—good going!

T

Christian Thalmann

Yes, I find Greek challenging in general, and its wildly variable stress isn't helping here at all! Then again, by the same token, a certain amount of hippie-ness is expected from Greek lowercase, which comes in very handy.

Speaking of which: I found myself stuck with that stupid horizontal /sigma/, so I decided to try something else entirely. I'm not sure it fits the logic of the typeface, but I find it looks more natural as a /sigma/ and plays well with /omikron/.

EDIT: Huh. That /sigma/ might just have been that one thing that kept me from liking the appearance of Overbold Greek. Now it suddenly clicks for me.

Christian Thalmann

I didn't like the juxtaposition of two identical /lambda/, so I made a ligature.

Jacob Casal

Things are going in nice direction! Some may object to the /sigma not being closed though, but I personally think it fits the color when everything comes together.

Christian Thalmann

Or I could just close it...

Craig Eliason

I’d have it go up into the roof, not over to the wall. 

Christian Thalmann

I tried that; looks much less compelling, since I'd have to taper the heavy stem without changing direction. I've seen tightly-curled sigmas in writing before, so I expect the form to be legitimate. I also think the horizontal counter helps with legibility.

Any native readers around here?

Jacob Casal

That too . Have you read any of Gerry Leonidas’s stuff on Greek type design (he may be one man and it’s possible other Greek type designers may differ from his opinion, but he is one of my go-to guys in terms of extensive thought on Greek right now)?

His site.

I haven’t read all of his stuff and was going to object too, but I did see his comment on Segoe Script in this 2007 Typophile thread:

Of script typefaces out there, Segoe Script is probably the best model for a contemporary informal Greek hand you can find. Jim and Carl did a great job of a difficult brief.

Segoe Script’s regular sigma looking like so:

Granted, its a script typeface with little variance in contrast. (This is where I got the leave it unconnected idea.) Admittedly I am not certain, most other sigmas seen do go to the roof as Craig and Kent note. Maybe compromise and have it slide into the extremum of the curve?

Craig Eliason

I can't help but wonder what a thin-bowl /sigma would look like, à la the revised /six. 
(Probably topheavy.)

Christian Thalmann

...and too small, no doubt. 

Christian Thalmann

Status quo Latin vs Greek:

Craig Eliason

Maybe the middle stems of cap Φ and Ψ can be thinned a bit like their lowercase counterparts. Maybe top part of /delta needn't go so far to the left. Thinnest part of final sigma gets too thin. I still think counter of /U should drop lower. Leg of /R doesn't feel quite settled. Maybe tail curl of /g could have a tiny bit more white?
This is a really fun design and you've solved some pretty big challenges along the way to get it to work!
I appreciate the subtle Upsilon/Y distinction!

John Savard

Thomas Phinney said:

This is quite a challenging style to do Greek for—good going!

I will express agreement. At first, I was not overly interested in this new typeface, despite the effort that went into overcoming its technical challenges. But I am very impressed, even astonished, by its Greek.

As for the numeral zero, however, I still disagree with using a thin horizontal void as the way to distinguish it from the letter O. I realize that the possible alternatives are limited, but one that would remain in the space of conventionality would be to have the white space in the center be vertical but curved to one side.

Jacob Casal

You certainly have done an impressive job! It’s a very bouncy design. That /R leg especially looks odd considering how other serifs in similar situations were handled though. Did you mean thinnest part of regular /σ Craig? At least in my eyes it does look a bit thin where it comes in to join.

Craig Eliason

No I meant at the left side of ς (but I agree with you on σ!)

Christian Thalmann

Craig: I implemented most of your insightful suggestions successfully, but there's only so much I can do about /R/ before it looks like a gnawed chicken wing. I'm willing to bend the rules a bit to make it look solid enough to stand on that leg.

Christian Thalmann

John: Yes, I feel I'm finally starting to gain some intuition on how Greek works. I've had to consciously steer myself away from Latinization in previous typefaces, whereas here I'm unfettered from the need to make things look sleek and clean and can just let its weirdness grow to fill the space.

Oh, and Cyrillic strikes me as horribly boring compared to Greek in this particular context...

Thomas Phinney

Latin is also a bit boring compared to Greek. Greek has a lot going on!

But Cyrillic has a few exciting bits of its own. The /be is pretty exciting.

Christian Thalmann

Trying out alternatives to the chicken-wing /R/... middle is too hippie, but the right one might work... a bit boring though.

Jasper de Waard

Out of those three, I think the original is best. Maybe you could just make the serif on the leg a little bit heavier/wider.

John Savard

I'm surprised you didn't consider something like

Claudio Piccinini

John Savard said:

I'm surprised you didn't consider something like

And more following the weight distribution of /B/. If you mean to keep the leg of the /R/ tiny, it needs to be balanced in some way. It’s the only glyph which looked “amateurish” to me as I first glanced at the caps. If you choose to have “folkloristic/vernacular/untrained” solutions they should find parallels in other glyphs, IMO.
/G/ has a bit of awkwardness too, but the vast majority of the other glyphs are well drawn and maybe in some way hinder your original idea of “overbold”?
Mostly talking loudly, anyway, I don’t want to disrupt your reasoning.

Christian Thalmann

I tend to agree with Jasper here, in that the original /R/ design was on the right track but needed a bit of a fatter leg, even if that breaks with the established thins in other letters. (Though /B/ is also already cheating by having its thin bowl thicker than the thins in other letters.) Here, the /B/ is, strangely enough, the character most closely related to /R/, so they need to agree on a common strategy. The fat-footed /R/s are doing their own thing, leaving /B/ all alone.

John: That extremely thin but fat bowl goes against my grain; even the less extreme example in /five/ already grates at me. I might try something similar with a thin top bowl like /B/, as Claudio suggests.

Claudio: The original /R/ was definitely too light in the leg, but I think the most recent version (before I began to backpedal) worked reasonably well:

I'm not worried about most glyphs being «too pretty», as it were. The goal is not to make things ugly, but to risk ugliness by going to extremes of the common glyph architectures without using the usual compensation techniques. If a pretty solution exists, all the better.

Jasper de Waard

Already looks quite nice to me, but maybe the (bottom of the) leg could also be shifted to the right a little further. Also: shouldn't Q's tail originate from the center? It looks a little like an afterthought to me.

Christian Thalmann

Nah, a little lopsidedness fits the theme and resonates with /B/. I also don't want the gap between the legs to grow too large, or it will start to look odd with respect to the closed counter.

Good catch on the /Q/!

Dave Rowland

I think the tail of Q could work nicely as the leg of R.

Craig Eliason

Dave Rowland said:

I think the tail of Q could work nicely as the leg of R.

I've been thinking the same thing.

Christian Thalmann

It looks less weird than I expected, but works less well as an /R/ than the serifed foot. Somehow I feel the caps can get away with drop terminals when they're out of the way like the tail of /Q/ or /J/, but not on the floor and ceiling of cap space.

It's not a common place for a drop terminal, either.

K Pease

I think the big bowl is vital to keep the /R/ in keeping with the style, but some weight is still needed in the tail to make it read as an /R/. Consider something like the "hippie" one but squashed down into that space, more flat and rectangular, like a little canoe, possibly attached by a stem that goes straight up instead of smoothing in.