Ambicase Modern

eliason's picture

Inspired by the funky Romanian 'A's of the images posted by Florinf (for example here), I have begun to develop a single-case font. Where uppercase and lowercase conventional forms of a given letter differ, I've tried to make hybrid letters that borrow from both forms.

The results are far from graceful, but interesting and more readable than I expected. Does this have any potential?

AttachmentSize
ambicasespecimen.pdf425.73 KB
ambicasespecimen9july.pdf433.44 KB
ambicasespecimen18july.pdf437.84 KB
ambicasespecimen1september.pdf449.61 KB
ambicasespecimen3september.pdf453.36 KB
ambicasespecimen13september.pdf464.87 KB
ambicasespecimen12january.pdf89.68 KB
ambicasespecimen14january.pdf92.59 KB
ambicasespecimen23january.pdf106.69 KB
ambicasespecimen4february.pdf121.29 KB
ambicasethicknesstest.pdf31.19 KB
ambicasespecimen11april.pdf226.54 KB
esandts.pdf20.2 KB
ambicasespecimen20July.pdf253.61 KB
Sindre's picture

I think it's brilliant.

eliason's picture

Thanks! Here it is tweaked (bigger ball), with init and fina forms. First row shows the previous forms of E. (The middle one would still be the default, no-swash E.)
Second and third row shows new ones. Only difference is in the E.init: the incoming swash/crossbar turns a harder corner in row three.

eliason's picture

Or actually descending is kind of nice for the init form.

riccard0's picture

I like the descender, and love the way the swash creates an e!

eliason's picture

Looking through old Bodoni script caps inspired this new D.init.


That also shows U.medi which I like in concept but I can't seem to de-uglify.
And the D.fina which I pulled apart a bit.

eliason's picture

And along similar lines here's a different idea for swash forms of Q:


I really like the concept (a 2-like italic Q form combined with a lowercase q form). Is it readable yet?

creamdonut's picture

I like that A shape very much. It's not either a lower nor an upper case A. Pretty interesting.

riccard0's picture

Interesting Q! It seems really non latin.

Bendy's picture

Wow, that's terrifying*! Nice work. I think it's instantly recognisable as a Q. Maybe you can do something to take some weight off the descender where it crosses the bowl — looks a little dark. As Riccardo said, it does look non-Latin. Looks like you've been studying that Cherokee font someone mentioned earlier.

I like the new U, but I'm not sure I understand it.

*that you're pushing the envelope so creatively and persistently!

riccard0's picture

Now that I think about it, it reminds me of this, too:


;-)

eliason's picture

Some adjusted Qs:

creamdonut's picture

Once again, very nice touch, the Q also shapes q. Really, this is great design.

eliason's picture

Thanks creamdonut!

eliason's picture

Going with Q#5, here it is with some curve refinement and .init and .fina forms.


Here's a better U.medi

Bendy's picture

Wow, that last one is sweet! Shame there aren't more words ending in Q!

nina's picture

Wow, this Q solution is brilliant.

eliason's picture

Thanks!
Allowing the init and medi form a little tail. That makes the glyph less abrupt, fits better with the usually enusing /U/, and may actually make it read more easily as a Q:


(That doesn't become to L-ish, does it?)

Sindre's picture

Boy, that's insane. Both thumbs up.

Bendy's picture

>Allowing the init and medi form a little tail.

Ha, I was going to suggest that! ;)
Are you having those as stylistic alts for the init and medi forms? The features could get very fiddly!

eliason's picture

No, I think I'll jettison the stubby tailless ones.

eliason's picture

Or rather, only have them come out in contexts where an ensuing descender would crash.

Bendy's picture

Ah, yes. Is your coding already quite complicated?

eliason's picture

It is, and I keep piling more into it (just added an IJ trigger, including swash forms, to go off for Dutch text yesterday).

I'm hoping my double encoding trick (http://typophile.com/node/67092) will work, because if not I'll have to double the coding to work the (identical) lowercase letters in.

eliason's picture

I thought in context my newish D.init (first line) looked a little too soft, so I'm trying a new one (second line) that borrows the stem treatment from B.init.


Also pictured: the aforementioned Dutch /IJ/s. I'm deciding between the two .init forms on the bottom two rows. The swash off of the J tail in the last one required lifting the I connection off of the baseline (which I think may be okay).

And yes, the N.fina is not quite there yet...

riccard0's picture

While beautiful, unfortunately I think your IJ resembles too much a ÿ.

eliason's picture

That's true of the proper written form of /ij/, which is one of the things I had in mind. E.g.:
http://en.wikipedia.org/wiki/File:Dutch_handwriting_sample.png
http://nl.wikipedia.org/wiki/Bestand:Aap_noot_mies.jpg
Maybe since that's mostly done in cursive, I should restrict the substitution here to the swash forms.

riccard0's picture

Whoa! I would have never imagined. And, in your first sample, the fact that it's placed between X and Z makes things even more confusing. Now I almost wonder if that's how ÿ came to be.

neverblink's picture

Craig, to me the second /IJ/.fina makes much more sense. If you had to handwrite the shape of the first, you'd start with the swash. I have no problem with the connection being slightly lifted, but as you might have seen in the Dutch IJ/ij flickr-group the Dutch are used to seeing many different shapes for this character.

Bernard B's picture

I'm usually not a fan of what I could call "swirled didones", but I think you've got something great here.
It's funky without being too "intellectual", works for me!

Just one thing, I don't like much your M, the wave in the top of the loop seems very disturbing to me.

eliason's picture

Thanks Bernard for the compliment, and for the feedback. I really appreciate it!
I'll keep tinkering with the M - I agree that it's still rather distracting.

eliason's picture

Here's where the M is after lots of curve pushing. It's similar to what I had a bit ago.


(Top is former, bottom is new.)
I'm still afraid it's just kind of ugly.

Bernard B's picture

For me it's really better like this, but I agree you should work your curves a bit.

eliason's picture

More on M
0 and 1 are what I just showed before.
2-5 are more nutty variations
F is what I had formerly (before trying the loop structure).


2 reads too much like an N on the left, I think.
5 may read most easily of the nutty variations.
Is F actually better than any of these new ones?

riccard0's picture

I'll go for 1, but maybe with a little teeny bit more contrast.

Sindre's picture

Seconded. But I'd keep alternative f too. Those in the middle are just too weird, I think. Have you tried an "M" modeled on the "N" trick (but in reverse), by the way?

Sindre's picture

1 is just smashing. Did you lighten it slightly in the middle? My idea of "N" mimicry was probably not very good, now I think "N" and "M" must be approached differently in a typeface like this.

eliason's picture

That #1 is the same as the previous posting (as is the one below). I still will do some curve pushing on it though it's driving me crazy. Looking at it now, I think I could widen the whole thing.
#6 is certainly readable but just looks bald on top. Here it is (#7) with some serifed corners, so that it approaches that former version in shape (#F).

Sindre's picture

Oh yes, 6 and 7 are absolutely readable. I meant that I think it suits the ambiguity theme that those two brother glyphs have disparate solutions. Perhaps "M" no. 1 could be just a touch wider, but not much. That shape is still my favourite.

Bendy's picture

Craig, my recommendation would be to work on number 1. It's by far the most interesting shape. Failing that, 7 could be worth investigating further, perhaps if you slightly slightly took the narrower proportions of variant F?

eliason's picture

Thanks for the feedback as always, guys!

eliason's picture

Here (in the 2nd and 3rd rows) is a reworking of M#1, a touch wider and the loop is more relaxed and, hopefully, more appealing.
#9 is the same as #8 with the middle vertex raised off the baseline instead of overshooting it.

creamdonut's picture

Although the M is beautiful, indeed it is, I think that in terms of word texture, is very much darker than most other letters, bring it out popping. I agree with Bendy it terms that it would interesting to explore variances between 7 and f (perhaps lose the lower serif in the middle of 7). I actually believe this M suits your swash variant better.

Sindre's picture

[...] very much darker than most other letters [...]

I honestly can't see that at all, and generally I'd say that that requires a print to properly judge. Besides, in a typeface like this, perfect evenness of colour is almost impossible to attain, and in my opinion not even desirable. This is a quirky display face, it should sparkle and pop and dance around.

I see a slight problem in the top middle of nos. 8 and 9, it looks to me like the thin curve is ever so slightly broken and bent by the intersecting thick stroke. That said, and you may call me stubborn, I still dig no. 1. There is something elegantly evil, or perhaps Gothic, about that curl that I really, really like.

eliason's picture

the thin curve is ever so slightly broken and bent by the intersecting thick stroke

Agreed; my plan was always to get the curves right and only then do the offset adjustment. Below is #9 with said adjustment made. It also has a couple of units shaved from the side verticals to lighten the whole.

Hmm, "evil"...? That's kinda the look I'm trying to smooth out of it.

eliason's picture

I've been working on cleaning this up and may have much more to show soon, but in the meantime here are the two latest versions of the /M/s.
row 1: latest version of the looped M, almost identical to the previous post's.
row 2: unlooped one which is a narrowed version of #7 above, as Bendy suggested.
row 3: same looped M as row 1, but set among the swash forms of other letters. (For the moment I'm forgetting init and fina forms)
row 4: same unlooped one as row 2, among swash forms.


I'd be open to using one as the regular and one as the swash, or using the same for both but also providing the other design as stylistic alternates. I'm thinking of also doing the same (either regular & swash, or regular & alternate) for the N, using the form above and the humped "n" form with the pointed toe that I had going a while back.

eliason's picture

Much cleaning up happening, procrastinating on the glyphs and coding I'll have to add to make diacritics in all the various swash forms. :-/

In the meantime, here's a question for you. I've been toying with the idea of thinning the hairlines. Due to the nature of the font, I can see it working well very large, but at those sizes (pt size near or in the three digits) the thins look a bit clunky. One alternative would be to have a "display" cut as an alternative.

Check out the ambicasethicknesstest.pdf attached to the first post and let me know which cut looks better at those three sizes.

(To make the thinner cut I just used Fontlab's "bold" action with negative numbers - didn't do any cleanup on the results yet.)

So question one is: which cut looks better at each of those sizes?

Question two is: if I were to offer both cuts, what would be appropriate labels for them? Usually the thinner would be "Display," but of course even the thicker is definitely a "display" font...

Bendy's picture

Hey Craig, I can't believe how nice this is looking!

I'm personally not a fan of thin hairlines, so I'll avoid giving an answer about which looks best. You could call the large optical size 'poster'?

I noticed your swash s followed by a colon. I wonder if there could be some kind of ligature or interaction between the swash and the dot.

With all the crazy shapes and stroke modulation, this must be a really hard design to get the spacing right.

eliason's picture

Thanks for the compliment and for weighing in, Ben!

I'm personally not a fan of thin hairlines, so I'll avoid giving an answer about which looks best.

So does that mean you as a designer would use the thicker option in every case? (I have to squeeze an opinion out of you since you're the only one who has replied so far! :-) ) Or if you still demur, could you answer at which sizes do these look best (or, which sizes do you think it'd be useful to target with a design like this)?

You could call the large optical size 'poster'?

Yeah, that occurred to me. "Display" and "Poster" — or //blank// and "Poster"...

I noticed your swash s followed by a colon. I wonder if there could be some kind of ligature or interaction between the swash and the dot.

That's an error. To this point, I had created a class of punctuation marks, which "turned off" adjacent swash forms. (Apparently I forgot to add /:/ to the class.) Not an entirely satisfactory solution, but it beat crashes between quotation marks, periods etc. and init/fina swashes. Here's what the colon should have looked like, along with some other examples.

Just the other day I pondered making custom kerning pairs for all those swash/punctuation combos. I hadn't thought of custom ligatures, but I think that'd be too tall of an order for me at this point. Even the thought of doing it with kerning seems daunting to me right now.

cerulean's picture

"Titling".

Syndicate content Syndicate content