Wagashi: Humanist (?) Sans Critique Request

Harry Wakamatsu

Nice to meet you. I'm Harry, a hobbyist font designer. I currently have about a year of type design experience, my most successful font being Motion Control (distributed for free on Dafont) that recently broke 100k downloads.

Wagashi (first named Ginkgo before I discovered Adobe's Ginkgo font) is my first semi-profssional font, boasting four different writing systems. However, when I submitted this font to MyFonts, it did not fulfill their design requirements, and they redirected me here.

Therefore, I must ask you a favor: could you point out the flaws in this font, in addition to suggestions for improving the design?

An apology: I could not get LibreOffice Writer's export PDF function to even remotely work, I must ask you to download the font here and plug it in to the OpenType Font Tester.

Wes Adams

I'm curious if you might share more information regarding your inspiration. Humanist designs generally reference the formal, upright calligraphy prevalent in italy around the end of the 15th century. http://en.m.wikipedia.org/wiki/Humanist_minuscule

Type designs, particularly Italian ones around that time, interpreted the written forms, changing them into a more unitized system for composing text. During the sixteenth century, admiration of the early humanist types spurred an adoption of their forms across Europe, most notably in France. Sometimes the work of this second generation is also termed humanist.

Today the term still applies to new designs that are modeled on the structure and principles of humanist calligraphy. Humanist writing cannot easily use typographic serifs because the shape of the nib does not make such a form natural to write. Humanist sans serifs leave them out.

Most humanist sans serifs, beginning around the 1980s, investigated whether a sans serif design could still be a humanist one. Prior to that, sans serifs typically used a different set of proportional guidelines, borrowed from modern and slab serif models. Excepting early twentieth century designs by notable calligraphers like Gill, Koch, and Van Krimpen, sans serifs were mainly grotesques.

None of that really counts against Wagashi or suggests that it is not formally interesting. There is nothing necessarily wrong with a design from a different point of view. Handwriting based types do have some natural advantages, though. For example, where a bowl joins a stem, the stroke is usually tapered, improving clarity.

Fonts designed for small sizes usually make adjustments in the service of legibility which are not absolutely essential for those functioning in large sizes (though they are still appreciated). By thinning strokes were they join, you can improve the appearance of most many letters, particularly those with diagonals.

Austin Stahl

I agree that thinning the strokes where they join would make a huge difference. On a lot of the letters where a curve is meeting a straight stem (h,m,n,b,d,p,q,u) it looks like the curved stroke is thinning in the wrong place: it's actually getting thicker as it meets the stem, but then getting too thin at the top of the curve.

The other thing I notice is that it seems like there are two different approaches kind of competing with each other. Something like the /s, with its flared stroke endings, feels kind of informal and writing-influenced. But then some of the others, like /b /d /p /q, with those flattened sides, have more of a tech-y geometric feel, which reminds me of something like Klavika. I feel like you need to push it more in one direction or the other.

Ray Larabie

Don't be too aggressive with your kern classes. Using the lowercase c as the master for so many other lowercase letters is pushing it. When it doubt, leave it out.

Watch out for stuff like grouping the ogonek kern classes together.. seems like bad news. You can solve a lot of overlapping lowercase accents with spacing. For example: J caron, just pad the sidebearings. Same idea for idieresis. Solve problems first with spacing, then with kerning.

You seem to be kerning the right side of the E quite a bit which can sometimes indicate a spacing problem.

Using the H as a parent for the J, d, i, itilde and f-ligatures doesn't make any sense. A flat H shouldn't require kerning anyway... if it does, you need to work on spacing.

Strokes don't need to make logical sense but you should at least make adjustments so they seem to harmonize better. If you print out the entire glyph set and stand away from it, you can identify certain glyphs that stand out as being too heavy or light. I can see 0436, 0438, 0E1E, N, W, k, v, w, 0416. Maybe consider working out the spacing on the A, T, L, F, I, P, W,W,X,Y,Z. I don't think they were spaced in context with words.

The rounds could use some reasonable overshoots. Most of these problems could have been solved by testing paragraphs, making notes, making adjustments and repeating. I'm guessing refined.

Keep working on the X, x... They're difficult but you have to keep working until you get them right.

A criss-cross lowercase w is challenging to get right. Try to find some successful examples of lowercase criss-cross w's and try to see how they did it. It's such a dense letter that it requires some finesse to get the right weight balance.

Thomas Phinney

I think you could benefit from understanding overshoot and other forms of optical compensation. I covered some of these here: https://www.youtube.com/watch?v=LR-CG5eB3nQ

Thomas Phinney

Some of these issues like the "V" are pretty inor. But the basic things about overshoot and thinning strokes when they join at a sharp angle are pretty important.

Deleted Account

...and Typedrawers is a real names forum, I thought?

LeMo aka PatternMan aka Frank E Blokland

Hi Harry,

TBH, I think you have to start with learning the basics of type (design) first. This simply will cost you a number of years, and then you’re not immediately a type designer still. I know the route; I got my first lesson from Gerrit Noordzij in 1978 at the KABK, and are lecturing type design since 1987 there (I’m actually the longest serving lecturer at the Royal Academy).

I consider it extremely friendly of other list members that they start pointing at details here, but when basic skills and knowledge are missing, this will become labor of Sisyphus IMHO.

Thomas Phinney

It is a bit hard on yourself to do so many glyphs without refining them first. The usual route is to do a few and make them better, and gradually increase the glyph complement as you refine the glyphs. When you present a huge glyph set like this, you've actually made a ton more work for yourself. "Fixes" that might have just shown up in one or two glyphs are in dozens instead. Ouch!

This applies not only to the glyph shapes, but the spacing as well. The spacing as it sits is very tight, and arguably desperately in need of an overhaul. If your straight-edged letters are 40 units, your letters like T and the right side of the L will need to be much less than 30 units, and letters like AVWX will need to be much less than 25-30. And while your rounded letters have partially flat sides, they still could get a bit less space than your completely straight letters.

(Note; I still have 11 spots left (out of 200) in a free online webinar intro to spacing on Tuesday, you might want to consider that. http://t.co/n7qxVNrtPe)

As a side note, Adobe does not have a typeface named Ginkgo (but Linotype does).

Harry Wakamatsu

Hello all,
Thank you all for the advice! I've: 1) fixed the joints of the b, d, h, et cetera, adjusted the curves of the Cc and Ss, 2) slightly changed around the VvWwXxY, 3) added 10 unit overshoots, 4) scratched the Greek and Russian parts (I was unhappy with how the Greek part looked, anyway.), 5) completely rekerned the whole font, and 6) fixed the stress that was going at the wrong angle.

I have uploaded the updated font here, and will soon ask you again for help on the serif variant. Thank you once more!