Interesting mix.

The J bowl looks bent out of shape and its terminal is sheared crooked.

Nice.
|S| should be a bit wider on the bottom left and a bit less wide on the top left.
Top bowl of |B| could extend a bit to the right.
You could try a non-low-waisted |E|.

Cool stuff.

The "Q" needs work; it would probably benefit from a tail like in the "K" and "R" (which you should definitely keep for the sake of character, although they do need some modulation).

BTW you can't really make the "black bodies" without keeping the spacing in mind - it's all inherently iterative. The "T" is a great case to consider in this respect: how wide it should be will depend very strongly on the overall spacing of the design.

hhp

Thanks for the feedback!

@cerulean: Do you mean that the curve of the J could be drawn better, or that shape is fundamentally out of whack?

@riccardo: Good call on the S. It feels like it's leaning to the left a bit, I think. I'll play with the B, although I don't want to regularize the font so much that it loses its quirkiness/character.

@hrant: ...which makes me glad you like the K/R legs! I tried a few tails on the Q, including some that had more in common with the K/R. I wasn't happy with them, but maybe I just need to work harder.

About the "S": it's leaning counter-clockwise because the bottom-left terminal is too short. BTW you might actually try making the spine uphill - check this out:
http://typophile.com/node/20111

hhp

J: From the vertical to the bottom, the curve is good. Just to the left of center, there is a large-scale kink after which the tail turns up more steeply and more straightened than I think it should.

Changes in this round:

-- Regularized the approach to curves, so O/Q/U now are better related to C/G
-- Widened the W
-- Tweaked the B, gave the E a centered waist
-- Reworked the curve of the J. I think it's a better fit with the rest of the characters.
-- Rotated the S slightly and extended the lower left terminal.

There's a new PDF in the original post, too.

Looking good.

The "B", being so "busy", looks a bit dense. The "D" is suffering from the "bone effect"*, where a curve coming off of a straight segment looks to be bulging outward; you need to pull in the points and extend the BCPs. The "J" is too wide, but also I think a shorter top bar is in order (but not too short). The "L" is too wide; if it does make sense though the "T" needs to be wider; I have a feeling those two letters should meet half-way. The curve of the "P" seems droopy. The "Q" feels cramped at the bottom. You'll also need to modulate the tight joins in the "K" and "R". The "X" and "Z" feel distorted - they might need to be wider. The "W" seems too constructed - try lowering the middle.

* This optical illusion was actually documented very well by Peter Karow many years ago.

BTW, I hope you make a lowercase for this! I can picture it being quite cool.

hhp

Good updates. I would suggest including O/Q alternates, as the old shape had its own sort of unexpected harmony with the face.

The S still has the look of being skewed to the left, but for a different reason: It seems you have moved apart the relative horizontal positions of the top and bottom extrema.

Hmmm yeah, looking at it again I think the old forms of the "O" and "Q" made more sense; I'd actually make the old ones primary* and the new ones alts.

hhp

More changes.

-- Applied the superellipse treatment to B, D, P, R
-- Narrowed the J
-- Raised the stem of the Y
-- Resurrected and refined an early S with curves more in keeping with the direction the type is evolving
-- Moved the tail of the Q away from the body and made its entry more vertical to add a bit more air

There is a certain crispness to the earlier, more geometric characters that I like, and love the idea of including them as alternates. Or maybe I'll design a sister face with that design conceit.

Not sure about a lowercase, Hrant. It would be a challenge, for sure! I tend to think really wide lowercase letters verge on parody most of the time.

New PDF in the original post.

At this point, you should apply the superellipse treatment to the upper part of |K| too!

Remember what Blaise Pascal once said: Things are at their best in their beginning.

Evolution is great, but make sure you do in fact believe in this new direction - don't let it just carry you along. Polish is no replacement for character - in fact it can be anathema to it.

hhp

I'm looking at it some more, and I actually think this is becoming disharmonious... Compare things like the "A" to things like the "D"; with the "D" like that I'm tempted to recommend trying an "A" with a rounded top (like the French used to do in the first half of the 20th century - something that's been revived here and there recently). Now, if you can make two cuts where one is more rigid and the other more fluid*, that would be awesome, especially if each can work for emphasis in a body of the other. But if you plan on finishing one first and then maybe doing another, I would actually revert.

* Even uniwidth being doable here.

hhp

The latest iteration includes figures, which were a challenge to draw. Beyond the tricky construction of some of the characters and the mirrored/not-mirrored thing going on with the 6/9, I wanted to inject a little personality, mindful of Hrant's comment about polish and character. (And I do plan to put the geometric versions in the lowercase slots, so rest easy, HHP.)

New PDF in the original post.

Those nums are quite charming. The "3" is especially attractive - it just needs its head shifted to the right a bit. The bottom-left of the spine of the "2" seems a bit heavy. The "4" is too wide. Try a "5" with a smaller head. The "6" seems unhappy - maybe the top-left curve is too square.

hhp

Thanks, Hrant. The 6 & 9 were much more difficult than I expected. And the 4 has gone through about 10 different angles/widths. :)

Hi gang! Here's the latest iteration. I'd love to have your feedback. Updates include:

-- Wider O/Q and 1
-- Revised S, 3
-- Redrawn 6/9 (Much happier, I think. I wasn't thinking of the structure properly before.)
-- "Narrower" (re-proportioned, really) 4

I has also got a notion to give the arms of the V, W and Y perpendicular endings, rather than shearing them off at the cap height. (I tried the same treatment with the top/bottom of the X and the bottom of the A. Disaster.) Distinctive, perhaps, but successful?

I seem to have lost the ability to add a new PDF to the original post.

Here's a PDF incorporating the latest changes:
https://www.dropbox.com/s/i5hq18vhen08c93/specimen_091112.pdf

Very classy feel, nice and sleek.
|Z| is too narrow. Also, it seems a bit off with the horizontal sharp spikes, if you widen the letter it will become even more apparent. Perhaps it will be better to cut those sharp joints?
Take a look at how you end the strokes, for instance: top ends of |Y|, |V|, |W| are angled, while you've cut them in |X|. I think you should address the |X| and make it consistent with the others. While the |A| should have the bottom ends cut since it's "standing" on the baseline, every other end of stroke in a glyph should be angled (like you did in every other glyph apart from |X|).
|B| still looks awkward to me. Maybe slide out the top bowl even more. Also, considering |B|, the fact that the bowls are so bulgy (what I mean is that they exceed beneath the baseline and over the cap height) make it a bit too bubbly. It works on the |D| but with two bowls in |B| it makes it too bubbly in my opinion and while the rest of the font is classy, |B| adds unnecessary playfulness (so to speak) to it.
Congrats on the design!

The new "S" is trying to be "proper" and failing. Maybe go the other way with the spine...

The "6"/"9" are indeed much better.

The new shears on the "V", "W" and "Y" fit nicely with new softer look of the face, and I can see how it wouldn't work on the "X", making it too much like a symbol and not a letter (although I would expect it to work on the "A"). Here's an idea: make the "X" a diagonal with the two other opposing arms curvy. You've got a lot of that sort of thing in your design already.

hhp

Thanks for the comments, gents. I tried the angled terminals on the X and, as Hrant surmised, it looked too much like a multiplication symbol, and seemed to be floating above the baseline. I'm intrigued by Hrant's suggestion of adding a curve to the X.

Hrant, you're determined to see an uphill S, aren't you? :) If I wanted the S to be "proper," what would you suggest?

Michał, do you think that if the bottom of the B was straight and parallel to the baseline that the curved top would still work? I'll take a look at the Z but am leery of chopping off the sharp points, although as the face evolves there's less of the sharp neo-Deco construction of early versions.

I was also considering either rounding the corners throughout as the default (somewhere between DIN Next and Brandon Grotesque) or offering a rounded version. Thoughts?

I think to be "proper" in the context of this design the spine of the "S" should appear to go perfectly horizontal, probably for a good stretch. Which means: I like the spine of the very first one (the top and bottom of the newer ones are better though) just with a bit of clockwise correction.

"Z": One problem is the weight distribution - the diagonal looks darker than the horizontals.

BTW I'm starting to not like the "W" - it's too close to a perfect zig-zag.

You can do a rounded version, but I would focus on the flat one more. In fact I might even try [a cut with] slightly concave (AKA "cupped") terminals, for extra definition.

hhp

Yes, perhaps making the bottom stroke of |B| straight and parallel to the baseline will work. Or just making the curve just a bit flatter and not perfectly straight - you know it's often all about tiny details ;)
I get the point about the |X|. I think Hrant's idea is the good solution.
I don't think round corners will fit the overall spirit of the typeface. Sharp ones will create black spots and ruin the contrast (that's often the case with |Z| in such expanded fonts with sharp corners). Maybe instead of trying to tame the corners - make the diagonal stroke a curve (double turning, an "s"-like curve), something like (just a quick sketch):

Thanks again for all the helpful feedback. I present more refinements and fiddling:

-- New X, with some offsetting/trickery to give the arms a less steep angle, which seems to help their relationship with the baseline. (I wrestled with curves for hours and couldn't produce anything I was happy with.)

-- New S, which sets aside Hrant's feedback (for now) in search of the S in my mind's eye.

-- Rounded Z, as suggested.

-- Flat-bottomed B, with a (very) slightly higher waist.

Here's the PDF: https://www.dropbox.com/s/gpo673v0u4novsm/specimen_091512.pdf

The "X" is nice of its own, and in an alphabet setting -surrounded by the other diagonal letters- it looks fine, but in a real word it stands out too much, like a pirate map location - it almost feels like it should be in red. :-) On the other hand you might in fact have arrived at the best compromise.

The spine of the "S" is working fine now - but the letter itself has become a little bit butch, probably because the top and bottom are too equal. Try making it slightly bottom-heavy, which the "B" for example already enjoys. BTW, I realize now that the bar of the "E" should come up too (or go back to matching the low bar of the "F").

I think the curve-bottom "B" was better; this new one is out of character.

The new "Z" isn't working for me at all - it's way out of character. The old one seems to be less of a compromise than the new "X" is...

BTW, something else you could try, just to see if it clicks: inspired by the numerals, try making the tops of the "S" and "Z" narrower.

hhp

Just saw this. Nice work. The original Z is better. The only character that I think really doesn't work yet is the S. You might try flattening top and bottom terminals, something like the very extended Dulles Airport Font, which Christian Schwartz digitized: http://www.christianschwartz.com/eero.shtml

BTW if you're trying to make the perpendicular shears work for the "X", the "A" shouldn't be excluded. Oh, and same goes for the top arm of the "K".

hhp

Another thing you've got going here is the reverse contrast thing on the vertical arches. Normally to look the same these have to be thicker than the vertical strokes; here they look thinner. You're not doing reverse contrast generally; the horizontals seem to be a bit thinner than the verticals otherwise. Is this what you want? Maybe you can make it work, but right now the curves, especially the outer curves on the tight vertical arches in B and S look awkward. Or maybe it's just that the vertical arches look pinched in comparison to the other verticals and horizontals. Have you tried going more whole hog on the reverse contrast? Or fattening the vertical arches?

Again, you've got something interesting going here, but I think it can benefit from further resolution.

i think cut ends in all strokes including A, X, V and so on would work better overall. Some letters may suffer then but the consistency will be worth it I think.
The new Z doesn't work indeed.

Thanks for all the comments, folks.

William, are you sure you're looking at an undistorted image? There's no reverse contrast at play, at least not intentionally. :)

Yes, I looked at a blow up of the PDF to check if I was seeing right. The horizontal with of the top vertical arch on the B is less wide than the vertical of the straight stem. To look the same, at this weight, I think it needs to be wider, or certainly at least the same.

Thanks, William. I appreciate your attention to detail. Another reviewer in other quarters made comments after looking at a distorted (squooshed!) preview that gave the font the look of bad attempt at Legato-like reverse stress. Just wanted to make sure that wasn't what was happening here.