Very well done! I'm a big fan of geometric faces.

Q's tail's a bit too small I think. Also your lowercase "k"'s arms are a bit ill-proportioned. The upper curve on the figure "3" is a bit dissonant.

Good luck on this one, it's beautiful.

I think this can become a pretty solid font - the idiosyncracies (like the "f") keep it from being too generic. One thing I'd worry about though is the vertical proportions: for diagram work you might like a bigger x-height, although that will make it less elegant. The caps are somewhat large, but that might actually be a good thing for display work.

In the UC, the only problem I see is the "Q": I agree with it being non-descending, but then you should try having an internal part to the tail. Actually, I'd change the "I" as well: for technical labeling serifs would help legibility.

In the lc:
- The "a" looks... dorky. Not sure how to fix it.
- I'd raise the dots on "i" and "j" (except if you'll be pulling in the extenders).
- The "k" isn't working.
- The beak of the "r" is big, but if you keep the spacing loose (since it's a light face) it might work.
- The "s" can have a small head, but I'd make the whole more circular - like in Futura.
- The "y" is too small.
- The "z" is leaning right.

Numerals: to me they seem too mannered.

--

Darker weights: you'll have to do a lot of modulation at the joins.
Italics: a sensitive slanted-Roman seems in order, maybe with some shears at the terminals

Keep at it!

hhp

Awe, Hrant -- I like the "a".

Maybe the "s" is leaning back a little, or maybe it does need to be more circular. What about hybrid numerals?

Jon, are you creating this out of a real need for this type of font, or is it more of a challenging side-project? Just curious. Can't wait to see a text sample...

-Addison

Thanks everybody!

Meir, I appreciate the encouragement, and suggestions.

Hrant, thanks for making such a detailed analysis.

Addison, this is a 'challenging side project' rooted in percieved need. Of course I hope that this can find real use for real need. The definition of the problem, the whys and the rules helps me get out of my own way and concentrate on the doing. I don't just want to say I want to make a geometric sans because it doesn't supply me enough traction.

I am definitely going to re-visit 'Q' , UC 'I' ,'k' and 's'. The idea of a more circular whole for the 's' is useful idea and I want to explore that. I see the tipping 'z' now, funny how that kind of thing can disappear on you.

The 'a'... well I don't know if I'd say dorky, but I also don't think I am totally happy with it. I want to keep it two-stories though, that's for sure. I have a couple ideas I will explore.

Hrant, if my numerals are too mannered do I fix it by encouraging them to be ill-mannered? :-) Kidding aside, I suspect this is because when I was a draftsman I prided myself on my numerals. I'm not sure I have an idea or reference point about where to go with them. Any suggestions?

Can't wait to see a text sample

Me too.

> do I fix it by encouraging them to be ill-mannered?

:->
Well, they're only mannered in the context of the alphabetics.
And maybe "mannered" wasn't even the right term (except for the "3").
Looking at them more carefully:
- The zero is a bit eggy but I think that's in character.
- The one and 2 are great.
- The 3 needs to decide what number it is.
- Try a bowed diagonal on the 4.
- I guess the 5 is kinda mannered too - that point is too strong. At the least make the bar longer.
- The 6 and 9 seem too big in the bowl.
- Like the 5, the 7's point is too much. Plus I might pull in (and maybe flatten) the main stroke.
- The 8 is squooshed. I know you don't want to make it too small though.

hhp

I think Hrant and Addison have the alphabet covered, but I'd definitely argue for oldstyle figures with this face. I love the small curves on 2 and 7, 6 and 9 are gorgeous, and 1 is nicely accented. The slight flair on these numerals really ask for drop treatment IMO. Maybe concave the hypotenuse on 4 and make 0 and 8 more ovoid

On the exclamation point and the question mark, you may want to move the bottom of the stroke higher up. That way when it's printed small, the dot won't morph into the stroke and cause confusion. If that makes sense...

Chase, thanks for the numeral ideas. I was going to try the reverse curve on the 4 but I don't think that it will do so well at small sizes. I'm going to be trying a few things and we'll see where I get.

That way when it's printed small, the dot won't morph into the stroke...
Matthew, it makes perfect sense, good point.

Thanks guys.

Jon

Well, after many months I've stolen a moment to get back to revising this typeface. This is a small step: some modifications to K and k.

I will be grappling with the other letters and suggested modifications outlined here as I find time.

Any comments are welcome.



Jon

That nicely tames the cap size, but I'm wondering what your plan with the descender depth is now.

The new "K"/"k" seem to be right on.

hhp

Thanks Hrant. I still need to get some solid hours to get my head back into this and look at everything.

Descenders...Thinking about balancing technical or performance demands with something that is attractive...

Things that make short descenders sensible are:
presumably limited vertical space in labelling situations, concern with collision of parts of an illustration or diagram

Things that make long descenders useful or okay are:
no long text or immersive reading requirement (not a text face), distinction from other letters(?), attractiveness and distinctiveness as letters and as Kompass (say compared to other geo-sans).

Studying the existing descenders though they look pretty good, and I'm going to try to make some studies soon.

There are indeed pros to long descenders, and in some cases it makes great sense. But the "limited vertical space" con you mention is sort of an extreme manifestation of a much broader functional requirement leading to the need to make descenders notably shorter than ascenders in any "real" text face:

Since economy is always a factor (to some extent), and you need the apparent leading to be just right to ease "linefeeds" but without waste, optimal usage of the vertical space has a direct bearing on functional merit; it's not really that you want to avoid lines touching (which can only happen with very narrow measure), it's simply that the space you give to the descender zone might be better used elsewhere.

As you say though it's a matter of balance, as always. It seems to me that the pivotal character is the "g"; generally it must look slightly cramped for the balance of Cartesian usage versus overall aesthetic integrity to be good for text.

hhp

Jon, I tried to imagine what the typeface would look like in the context of a map or library label. All-caps settings tend to be used often (for headings, tables, etc.) and I think the uneven width of your capital letters would look somewhat awkward. Shrinking the W and M would help; make the narrow glyphs (B, P, R, U, V) a little wider so they'll sit nicer next to the wide round letters; but don't compress the round glyphs or you'll kill the geometric quality. The R needs work -- it looks like a B with the bottom chopped off, and a straight leg would suit your intent better.

The number 1 should have a wider "beak," (and maybe a "foot") for the same reasons.

I like Avenir but it's been getting overused, so I'm interested to see where you take this. I think there is a niche for a font that has the "flavor" of Avenir but is tailored for a certain function. (Hmm, I wonder, what's the flavor of Avenir... does it taste like Cool Mint Listerine, chewable vitamin C tablets, a raspberry Stoli & soda?)

//john

John, Thanks for your input. Perfectly timed as I revisit the many aspects of this project. It is very true that all caps is common in labelling on maps and in libraries, and your advice makes a great deal of sense in this respect.

When I drew the glyphs it was of great concern to me that I not end up with another Futura. I like Avenir a lot, and was making decisions on some of the glyphs like 'f' and 'R' to intenionally diverge from the path.

Your suggestions strike a chord with me and I will give them serious thought.

To my thinking Avenir = Cool Mint Listerine (minty with a little hint of medicine).

Jon

Well I finally carved out some time to return to this.

I have redrawn many of the letters, incorporating many of the suggestions here (an improvement for sure). Here's the deal:
- Narrowed the 'W' and 'M'
- redrew the 's' which is more geometric now, but I'm afraid is too narrow
- Just noticed that the uc 'S' needs to be redrawn too, especially its tail.
- terminated the hook in the 'J' and 'j' earlier and they look nicer now.
- K as noted in an earlier post
- uc 'i' has bars on it but I got scared and they look knobby. Same with 'J'.
- lc 'a' is now a single storey (sigh) which works better.
- 'Q' has a new tail.
- lc 'L' stroke too thick and 'X' strokes too thin.
- also redrew the lc 'y' and shortened its extender.

I took the drawings into TypeTool and generated this font. The sample was made with the newly generated font in Illustrator and converted to outlines for the PDF so I didn't run into any weirdness. I did some ignorant kerning and spacing but I have never used a font editor before so this is obviously crude. No punctuation or numerals have been revised since the earlier drawings and are not in the font at this time.

Are W and M narrow enough now? Should I straighten the leg of the R? Please comment on anything as you see fit.

Download the 20060105_KompassSample.pdf.

A small gif to give you a clue:

Well, now it's redundant. The original two-storey 'a' and 's' gave this face an identity and a reason to exist, and I don't think there was anything wrong with them.

The 'R' looks awkward, but I don't think the leg must be straightened necessarily. I think it just needs to join a little further to the left.

The 'A' and 'V' should match; you can see the problem in 'ARRIVALS.' A compromise between the two angles you have now would probably be best for this face.

Thanks Kevin,

I see the angle problem in ARRIVALS, that's a fair point. I'll try adjusting these.

Well, now it’s redundant. The original two-storey ‘a’ and ‘s’ gave this face an identity and a reason to exist

Am I right in understanding that you think I have moved too close to other geo-sans by changing this ? I suspect I am wrestling with the originality/usefulness problem. Iaadmit that the single storey a is much more pedestrian but may be more useful in the intended application of labelling etc.. I think the s is improved though.

I am going to see if there's something I can do with the R. I like the curved leg and will keep it. I'll try your suggestion about moving the join.

Thanks for your time and comments!

Jon

You are correct, and I actually think it's more of a utility thing! When I saw your initial post, I thought to myself, "Excellent, someone has fixed the legibility problem common to many geometric sans." Especially for labelling. Whether at small screen sizes, small print with ink gain, lacking one's reading glasses, or just from glancing at it too fast, there is a huge potential for 'a'/'o' confusion. If the tiny words in your gif were out of context, on a map, for instance, I honestly wouldn't know if I were looking at 'Rambas Tocos' or what.

Hey Kevin,

Thanks for this. I went to the single story a after a lot of back and forthing and not feeling right about it either way.

I am going to do some studies on the a and see if I can find something that is better. I'll share when I get some good comparisons.

Jon