This is an experiment I did a while ago, and now decided to revisit. What do you think?
I previously designed Latin, Cyrillic and Greek, and upon this iteration I added Armenian.
This last one is tricky, with the forms often protruding to the right. I tried to make them bump into following letters gracefully, but that didn't work all that well. The uppercase variants contain more moderated but also experimantal forms.
https://1drv.ms/u/s!AulK0yIHtt_3ju4-dzWiHr6tn33FAw
Comments
I'm a bit bothered by the round endcaps of the counters; I suspect they would look better if they were less elongate and closer to the circle.
I forgot to mention, though, not that it matters much, that the font was created as quadratic curves. I just thought I liked the shape of the straight-angle-control-point curve for this design and sticked to it along, which helped me make uniform curves quicker. I now converted the font to cubic Beziers to try this out. Maybe a good step at this moment, time to manipulate further and polish?
I found the letter /C surprisingly difficult to get right. I tried to let more light inside, but that just spoils the fun of it. Right? Only maybe it looks more Thai with the curls rounded inside.
I like the second C a lot, but it could be read as a Q. I have no trouble reading it in SPICE, but you might want to test with a larger volume of words.
I’m amazed that the /a works, but it does!
I also put up some serifs on /P and /F to make /T feel less lonely. By the way — does the /P in the middle work? I somehow favor the one(s) on the right, as it compliments /A/B/R better; the left one was always kept just in case (the lower one
@Christian Thalmann @Dave Rowland
Does it make sense to develop both square and round endcaps? I somehow cannot decide on either. It is a small detail but repeated everywhere; could you please explain how one or the other works better?
Edit:
Did I get C-arried away?
5 mins later
The /P works much better with a CCW spiral and no foot. I like the foot on /T but not on /F.
The new /C are very hard to read indeed. I mostly thought of /L at first.
I also read something like «Piccould» instead of PICCOLO. Maybe you could get rid of the cheat in /O and give it a full circumference and a filled rectangular counter.
The /E might profit from soft corners on the left side and hard ones on the right.
As for the counters, the problem is mostly with those very blunt outer curves (like the bottom left of /a) that feel completely disjoint with a streamlined slit cap but feel like a natural extrusion of a rectangular slit cap. I would suggest sticking to rectangular, and if you absolutely must have a curved variant, make the aspect ratio wider than tall.
Two new versions of /O.
I guess the third one works against /U? But the /U and /V dislike each other. I rounded the bottom of /V a bit for what it's worth. I mean, we are supposed to recognize letters by their tops, it's bad if we have to look at the feet to tell them apart. Any ideas?
And the corners on /e. Below I type ~wAvY~ old school. Comes out almost like an /f_f ligature.
Two versions of /Oslash — one with the «cheats» to go with the /O with the cheat, and one cheatless. I prefer when there is connection between the O and the stroke, but with a full height integral this would require cheating muchly.
I dropped in a new /m, based on armenian /ayb actually
There is more cheating in the Greek as on the last page of the pdf. What do you think of the cap-style accents incorporated into the letters? I've heard in all-caps settings accents are only kept on the word initial letters. If I say my face is unicase, do the accents work, especially that the Latin ones are incorporated in a similar fashion, only centrally above the letter? I guess we would need to hear from someone who won't read into this by accident.
I'm saving all those failed attempts for a separate version of the font, Radiator Extreme. Or maybe Xtrm, for it will be harder to read.
This is what I could do to /M, and /W by the way.
Feels somehow streamlined, though... Maybe I could just adjust the M and W by giving their endcaps extraordinary leeway.
The lowercasey /h was a recent development to accompany the /t. The latter was a tribute to Coolvetica, I guess. It was always meant to be used with discretion, but if so, maybe it should be eradicated.
I suppose you mean something like this?
If I remove the incision from /C as well, it will no longer differ from D apart from the rotation. Correction: the middle counter in /D goes up 10 units higher to reinforce the presence of the vertical stem.
I could try to get all the counters to the minimum and keep them square as you suggested before (if that's what you meant; maybe you only had curves deep inside the letters, not open counters?):
The /g is a full-fledged cheat. It has thinnings to accomodate all the tucked in terminals. Having thickened the thinning, I get the impression the inner stroke is now optically wider than the outer ones—but perhaps it's because I've seen it so many times with the thinning? Anyway, I don't really like the uppercase G either, because of the image-ground reversal thing.
I don't even know if it's better or worse with the curves straigthened. Maybe better? The sight of two curves facing each other is somehow unpleasant in the middle of a letter, right?
Yes, jackdaw! Sorry, my bad.
I guess the top one better?
Anyway, feels awkward, because the golden rule has been that it's not a stencil: rather monoline. Except for /8 which couldn't be connected and not look like S, and now the full metal alchemist /O.