Okay, having narrowed a.salt and g, I better make sure the other ball-and-stick letters don't now need parallel revisions.
(pic shows new a.salt and g)
I don't actually think they do - maybe 'q'?
q might stand it. (And did we do a variant with a flick to the right on the tail?)
(And did we do a variant with a flick to the right on the tail?)
I don't think I did, did you?
But seriously, I just was looking with chagrin at that tail in that sample. I just changed it from a flattened terminal (like the bottom 'l', for example) to a curved one (like the top of 'l'), but it still looks awfully dumb all of the sudden. I could experiment with a flick. Do you have any ready examples in mind? (The ascender of 'b' isn't so bad, right?)
I thought of it when I made several q's to see how wide/narrow I made mine, but I tended always to curve the tail to the right. It's usually less than half as wide as the tail of the j, but obviously facing the other direction.
quite = old
frequent = "doubling up" of descender
sequels = flick based on 'j'
Craig, I personally feel that curve on the descender of your "q" is too soft. I think it should be more of a flick than a curve. (Remember the salt.)
Re: examples, for some reason I feel like listing some semi-randomly picked, right-facing "q" descenders (I've become quite a fan of these!).
Handwriting/calligraphic fonts with right-"flicking" "q" descenders:Sanvito ProKallos ITCMama ScriptTekton ProLucida Calligraphy / Lucida Handwriting / More subtle: Lucida CasualITC Bradley HandMedici ScriptFF Child's HandCascade ScriptLemonade ICG
Other fonts: ITC Berkeley, Hiroshige, Parisine Plus; Patria
Some historically-influenced examples:Carolina, based on the Carolingian Minuscule; Arnold Boecklin; Auriol
I was going to post some actual historical samples, but I need to make a phone call first. :->
Yeah, some of my first sketches were more upward flicks, like #2 and #4 here:
But they weren't working to my eye. Basically, I felt like the short stroke became too much of a distracting detail - at a different level of detail than most of the rest of the font, almost like a rogue serif in a sans font. (I would in fact call many or most of the q-descender-flick examples you cited "serifs".)
Did I dismiss promising possibilities prematurely? Wouldn't be the first time. Should I keep working on #4?
Hm… how about something between #4 and #5 – maybe not angular, but kind of a little curve whose "weight"/extremum is heavily on the left?
Alternately, maybe something like #2 could work if you keep the model of your "n" &c. in mind?
What about the third one but with the curvy ending on the inside (stroke completely flipped)? The dynamic of the stroke is leading back into the bowl and I think the movement should be the other way.
4 and 5 are definitely the ones to develop. 5 in fact looks very much like my normal q when I print (actually the stem leans slightly, so the bottom of the stem is slightly under the bowl).
rough 'q' sketches
4 & 5 are as above.
6 is the "altaira variation" and 7 the "Bendy variation" :-)
Same thing smaller:
6, definitely. Now try (just for kicks) moving the bottom just a hair to the left.
I'm thinking 6 could work out fine; but I think I'd make that "hook" tighter and quite a bit smaller/shorter – it is quite dominant now. To me the "optical size/importance" of the hook looks better in 5, FWIW.
I'd worry that if you made the hook in 6 tighter, it might fill in at text sizes.
Now try (just for kicks) moving the bottom just a hair to the left.
Did you mean the whole descender, or the extremum at the bottom, or something else?
Here's a couple of versions shortening that hook of number six, big and small:
9 might work well.
I meant to move the lower part of the tail, so it has an ever-so-slight NE/SW tilt.
It already does, but I could try pushing it further.
I find the flick of the q difficult to get right too. I don't think this'll be much help but it's more what I was intending to mean before...
Ah yes, like #2 in my quququququ sample, only drawn well!
I think it'd need to be better drawn really though, as that's quite a heavy bottom on it and probably needs some thinning.
I think #9 might work well. The other thing I was thinking about (and tried to describe somewhere up there) is pretty much exactly what Ben posted above.
That looks pretty cool too (and takes up a structure you already have in there).
I think the cleaned up n-like tail (2B) is less attractive but perhaps more fitting with the alphabet than the small curl (9B) - BUT, at least as I've drawn it, it's no good at all at small sizes. Agree?
The eternal question of 2B or not 2B... It does sort of disappear when small, so I'd have to go with 9B.
I think I agree. 9B looks kinda funky large, but seems to work quite nicely smaller.
Edit: 2B could also be mistaken for a somewhat strangely-shaped handwritten "g" with an anemic loop – at least around here.
Yup, needs to work at small sizes too. Go for 9B :)
Well, after all this work, it'll be very exciting for this font to come out! Please keep us informed, eliason.
As for the q, I think that no matter the final design, it would be nice to have a tail-less version available to the user as an alternate.
Keep up the good work!
I'll do just that. Thanks!
Lastest pdf up top (loopysample10.pdf) shows status as of today. I've made a ton of tweaks. Some of the more prominent (or at least the ones I can remember now) are:
- Slimmed down some "secondary" strokes like bar, slash, the short strokes on currency symbols, diacritic slashes, etc.
- Added a little more curve to some of the horizontals like top of T and Z, bottom of E and L, etc.
- Knocked the vertical part of the fives off-vertical
- Removed W.salt
- made many of the dots (e.g. period, colon, bullet, tittle, umlaut) more (but not entirely) circular
- redrew ear of g.salt
- addressed some of the weight problems and wonky curves
I also changed the hook of the f, and with that, much of the ligature setup. Instead of the old ligatures that had that f hook stretch to bridge with ensuing ascenders (fl, fb, etc.), I have designed an alternate f with a slightly shorter hook to go before those letters. The ligatures that connect with the crossbar (of f or t) are the only ones that remain, and are all discretionary.
Here are my reactions to the latest PDF:
1) Wow. This is looking nice!
2) I like the f-ligatures, and I think they work well here.
3) Now that I see it in context, I think the ß is too wide.
4) The tabular 0 (both lining and hanging) is too wide.
5) Looking at the French, I wonder if your apostrophe needs more curvature.
6) There is no number 6.
And, welcome to the "E-named fonts club". :-)
I printed this out, and tried looking at it as if I'd never seen the font before. And I must say, this is incredibly amiable and seems to be working very well indeed. I'll admit that I'm not usually a big fan of "handwritten" fonts, but I do like yours a lot! The tons of work you've put into this really show, as does the joy of making it.
Here are some things that struck me:
There might be some weight issues with certain glyphs, but I can't be quite sure since I'm still battling my printer. Particularly, I'm wondering if the top of the "8" is too light (especially compared to the ampersand, which looks very nice); the small cap "A" in the Latin sample looks a bit dark and the tilde on the "a" in the Portuguese, very light. The percent sign seems rather light too. But like I said, that may or may not actually be in the font.
I like the "Q" a *lot*.
Your guillemets (both single and double) seem rather mechanical. I wonder if they could have more of that handwritten feel.
I agree with Steve that the eszett feels too wide. I think it's especially the bottom part – it sorta looks too much like a truncated "s".
Ligatures and "f" alternates look great!
I think the apostrophe is fine.
The accents look better, but I'm wondering if they're now tapering too much towards the thin end? They look really pointy, which isn't really something you have in the rest of the font. That makes them look a bit aggressive in context.
Btw I think the circumflex could benefit from not being quite as flat, and it doesn't seem to share the contrast of the "straight" accents?
The asterisk is lovely.
I'm quite excited about this. What a nice font.
Thanks guys; your encouragement means a lot to me, and your eyes for persisting problems are invaluable. I'll look at fixes for all of those issues. :-)
And in the meantime, what do you think of a change along these lines to g and a.salt?
(As a special bonus that one shows my demechanicalized guillemots.)
And how's the width on this ß?
(or should I abandon the snaky form and do more of a fold (like the 3) or even a loop (like the KBR)?)
Your alternate a and g look delicious! Yum!
BTW Guillemot vs Guillemet
I've just noticed how well the caps work. Would it be too horrible to widen the W and pull the middle up slightly but not to the top? It looks a bit shrinky I think.
The lc of course looks great too!
Emi looks really wicked in comparison to those other fonts and even against Myriad, which I very much admire. Good work!
I like the new alternate "a" and "g" too. They look smoother without abandoning the premise of the font I think.
But the "ß" still looks a bit weird… too much bulging at the bottom to my eyes. :-/
I think the basic shape could work, but the proportions seem off.
These less-familiar glyphs are so hard for me to judge.
I'll try the W, Ben.
(LOL at myself for "guillemots")
The same new a.salt and g, with a little more context.
The letters are more attractive, but does it work okay with the other letters well enough? Do the new a and g (and q - I changed it too) look like they're falling off at the upper corner? Does the stem coming up shy of the x-height make other things (like the stems of r and n) stick out more?
To be honest, I like the version with the stem coming up (i.e., the top version) better.
Or there's this - an intersection like the revision, but back up at the x-height line:
Same as 8:53 post, but with redrawn (less bubbly) and slightly narrower bowls:
No, stick with the original or the first revision. You could be right about the NE corner though. It makes the shape look shorter than the others, so perhaps a bit more overshoot would do it? I don't prefer 3 and 4.
Still on the g and a.salt; 1, 2, and 4 are as above and 5 is new: is 5 any better, or should I stick with 1?
In my opinion, 2 has very nice forms of a and g, but they perhaps don't quite fit so perfectly as the number 1s. I think 5 fails a bit as the NE corner looks a little clogged, but 1 has larger counters there. 2 could also be a bit reminiscent of Comic Sans, though of course much pleasanter. 5 has lost some of the springiness in the top arches, which gave 1 the vitality and made the a and g harmonise better with d and b.
I think you're right, Ben. Thanks for the observations.
>And, welcome to the “E-named fonts club”. :-)
Um, took me a moment to realise what you were talking about, Nina! Nice ;)
Some more glimpses of the redrawn ß. Will this work (setting aside the need for kerning)?
I think this works a lot better than before, Craig. It's still quite soft-feeling and the difference in width between the two bowls is characteristic, but I think it works.
FWIW (from Wikipedia):
Since you draw a lot of your inspiration from handwriting here: Most people I know write the "ß" a bit like a beta, with a sharp connection between two round bowls – most similar, in this picture, to #2 in the second row, or the rightmost one in the second row from the bottom. There may or may not be a descender. (Note that those ones that don't do the sharp join in the middle still have two approximately almost equally wide "bowls". Also note how the stroke starts at the top-left for the handwritten variants, comparable to how one would write a "p". Do not look at the center one in the bottom row – that just looks stupid. :-> )
I don't know how deep you want/need to go into this; it may just take too much time and effort to do a super brilliant unbeatable eszett, as it's not a letter you are usually confronted with – and then it's kind of a "minority glyph" ;-). I think yours is definitely readable and even rather charming, so you might just leave it as is.
But do make sure the bottom doesn't close up / clog at smaller sizes.
Off topic sorry! The dieresis in your sample there looks a bit close to the a. It would look happier to me to be the same height as your i tittle. Or perhaps you wanted it to be different?
Good catch - I've raised it to match the tittles. Here's some revised diacritics for your perusal, as well as yet another go at the germandbls, this time with a sharper middle and a doubled stem based on p.