Codename "Gouda" (1920's pastiche, more or less)

Michael Vokits

Kind of a mix of Goudy and Dwiggins shapes in the roman in some ways, fairly narrow and spindly. Looks much less so now than it used to, but I can’t think of a better codename. It seems to work well enough on tablet.

I’m still unhappy with the roman /f; any pointers? Darkening it made it look unbalanced next to /l.

Is the roman /Z a bit dark?

I think I prefer the alternate italic shapes of /C/G/S; they match the lc (not just their lc counterparts) much better. I also think I prefer the basic shape of alternate /S over that of normal S, which is strictly a sloped roman.

Smallcaps forms:

I used sloped forms of /G and /Q in the italic to match the roman SC as opposed to the italic caps; is this correct?

Steve Gardner

There's quite a difference in serif weight throughout (compare the /E and /T to the /S and /Z, for example).  Is this intentional?  

Michael Vokits

Steve Gardner said:

There's quite a difference in serif weight throughout (compare the /E and /T to the /S and /Z, for example).  Is this intentional?  

Nope, just an oversight. (Maybe “lack of sight”?) I knew there was something wrong, but just couldn’t point to it. Thanks!

Steve Gardner

Looking at this again, I'd expect the angled stems of letters such as /i, /j, /m, /n & /u to overshoot the x height a little.  Look, for example, at the /uv pair; to my eye, the /u appears too short in comparison.

Also, I agree with you that the /z seems a bit dark, as does the /4 in your old style figures.  By contrast, both the /W and /w seem light.  Take a look how light the /w appears in the word 'wisi' in your PDF (this word also highlights the need to look again at the relative heights, since the /w seems taller than the /i).  

Michael Vokits

2018-10-28, last night’s version.

Did a lot of little things to increase consistency, especially overshoot. Alignment zones and bluevalues mean that you have to sacrifice beauty (ha!) of proportion for consistency in low-pixel/ppem environments, but that’s the game.

I changed strategies on /W and /w; overlapping vees produced either too dark a color (if I didn’t lighten the vees) or too light (I had to make each vee obnoxiously light to make its overall average color/density consistent with everything else). /W is too wide, but it has the same kerning as /V, so that it’ll work with /A on either side (pet peeve of mine).

I have two sets of superiors and inferiors; the normal create fractions a hyphen wide (also the width of each numeral). Is this acceptable? The larger look much better in chemical formulae (and could use some kerning!), but the smaller fractions look nicely discrete — much nicer, in fact:


Enh, what’s the superscript equivalent of sinf? (I’ve been faking it by using .supr and .subs features, but that’s incorrect.)

Thomas Phinney

There is only one superscript feature (unless you also count numr).

The two subscript features only exist because in the original discussions, some people wanted a subscript that went below the baseline, but somebody else thought a distinct subscript that sat on the baseline was important to have.

I don’t know of any need for that (other than as a denominator, but we have that feature as well).

Michael Vokits

2018-11-03. Minor touch-ups all over, but redid /y. Added a rather vast number of dotaccent and other such modded ASCII glyphs, and Latin-2 seems covered. I’ll either complete the IPA set, add a Greek, or add bold versions (maybe smallcaps) next.

This was never meant to be a large project; it was just a toy experiment (hence the bare-bones structures) that just gathered its own momentum. Not a bad fruition for a bunch of fugly, unruly, disproportionate glyphs.

John Savard

I like Goudy and Dwiggins, and I like your typeface. The revised version improves the lowercase f, but in my opinion the problem with it is that the crossbar on the f should be raised to match the x-height; right now, the top of the letter looks too big.

I don't have the kind of knowledge of type design required to advise you instead how to turn what was presumably an intentional experiment to flout convention with the f into a successful one, and thus the only recommendation I can make is the timid one to scurry back to what is most common.

Looking at the pdf sample of the third version, though, with an extended text specimen, I can see other issues. The crossbar on the t, also lower than the x-height, is distracting. Letters like g and o and e have an optical correction bringing them above the x-height... but it's excessive, so the letters look too large, breaking the x-height line.

Michael Vokits

Been a while, but I think I've improved matters somewhat; curves are smoother, proportions better, ink traps were added in tight corners and between curves (/B and such), and I went with a non-kerning /f. If this design works out, I'm planning on maybe doing optical sizes; the new codename is "Ahptikul" in such hopes.

A few brief notes:

I'm still not happy about the roman /f, but at least it's consistent. Keeping its head consistent with the serifs of /a/c/s makes it appear weak, but darkening it sufficiently would make the letter monoline. Just a teensy bit vexing, that. Any suggestions?

The italic question mark is left-handed and the roman right; this helped tame the curves.

And yes, the kerning needs work. No point in fine-tuning that until the letters are correct -- a lesson I wish I'd learned earlier!

Thomas Phinney

This is really turning into a well-developed typeface. Very nearly usable now.

That beak on the /f needs a lot more weight—at least as much as the top serif of /b /d /h. (I would also see what it looks like with a tad more, but probably similar is close enough.

If you have this Linotype-style short /f, you don’t really need f-ligatures so much. The ff looks a bit odd with that beak, in fact. At least, to my eye.

The join on /n and /u seems thinner than almost any other thins, perhaps a bit too much so. The foot of the /t is also a tad light.

Maybe /a is a bit too narrow?

In the italic, the ascender has a much more rounded cornering treatment than the baseline. I am looking at (for example) the el and thinking the direction change looks much better at the baseline than at the top of the ascender. Same for letters such as /n /u, the x-height serif seems less successful to my eye.

Michael Vokits

Thanks for the feedback! I agree with everything said above. Roman /f still mocks me, but I think I’ve gotten to the point of the least pessimal color and shape. Perhaps I’m overly worried about consistency?

Michael Vokits

Here are the relevant details (no need to dl anything).

“Consistency: how much is too much? Are your hobgoblins ruining your design?” (Sounds like a slow news day segment!)

Craig Eliason

Just a few quick observations:
When I see /a/ next to /n/ or /m/ it's very apparent that the latter letters carry far more weight into their shoulders than /a/. I would darken the top of /a/ (start by having the outside stem contour stay straight for longer before tapering over. (It may also be that /m/ and /n/ are too heavy through their humps.)
/g/ seems way too lyrical and smooth to fit in with the other letters.
Straight vertical part of /G/ should be full stem width. 
Crossbar of /A/ might be a bit too heavy (relative to the diagonal thin) and maybe a bit too high. 
Tittles are too large, and I'm not sure their shape is working (roman and italic). 
Closer attention to sidebearings will make assessing letter design easier.
Keep rethinking the italic lowercase serifs. Top arms of italic /E/ and /F/ feel stunted. Top right of italic /K/ and /Y/ should bear more relationship to each other. 
Letter proportions overall are very solid and most glyphs are very well drawn.

Craig Eliason

and four's counter might get too small.

Michael Vokits

Craig Eliason said:

Just a few quick observations:
When I see /a/ next to /n/ or /m/ it's very apparent that the latter letters carry far more weight into their shoulders than /a/. I would darken the top of /a/ (start by having the outside stem contour stay straight for longer before tapering over. (It may also be that /m/ and /n/ are too heavy through their humps.)
/g/ seems way too lyrical and smooth to fit in with the other letters.
Straight vertical part of /G/ should be full stem width. 
Crossbar of /A/ might be a bit too heavy (relative to the diagonal thin) and maybe a bit too high. 
Tittles are too large, and I'm not sure their shape is working (roman and italic). 
Closer attention to sidebearings will make assessing letter design easier.
Keep rethinking the italic lowercase serifs. Top arms of italic /E/ and /F/ feel stunted. Top right of italic /K/ and /Y/ should bear more relationship to each other. 
Letter proportions overall are very solid and most glyphs are very well drawn.

Many thanks for the feedback! A few notes:

/a: yep, I totally see it now. I was going for a mirror/reverse contrast thing with /f, but that really doesn't work when you're trying to keep a bit of a pen feel, alas.

/g: best negative review ever!

Tittles were resized and reshaped; I was definitely trying too hard with them.

Sidebearings... and how. There have been a lot of revisions since I ran things through LeMo’s spacer. Astonishing how much of a difference spacing makes.

Italic serifs. I’m thinking of a more horizontal in-stroke:

On the left is the original, deservedly maligned serif; the middle is that from the latest specimen (and simply the rotated out-stroke); and the third is the proposed new serif (whose bottom is perhaps a bit too flat). Better?

Michael Vokits

First stab at a new roman /g, with the old for reference, more or less in context:

Not thrilled about the inside of that lower bowl... but it’s sufficient for a prototype.

Michael Vokits

And a spec sheet, with boxes to show sidebearings (more or less; LaTeX is being a jerk these days). /g is halfway decent now. (I'm a dogged reviser, not somebody who gets things right the first time. Ever.) Cheers, all.

Adam Jagosz

That new structure for /g is wildly original. Not sure I like it though.

Michael Vokits

Yep, I know what you mean. I was trying to go for something simpler than the previous “lyrical” /g, but kinda went overboard in the other direction. Ah well; back to the drawing… err, tablet!

Michael Vokits

A little better, methinketh:

Shape and simplicity match the other glyphs; some tweaking required.

Michael Vokits

I think the regular roman and regular italic are ready for testing. If nobody complains too much about the design, I’ll probably be working on a bold weight and axis next. Or maybe an optical size instead, since that interests me far more than weight. (And extending the character set; I want a full Latin plus IPA!)