Don’t have access to a printer to give any real insight, but i did notice some path direction issues (overlaps) see the eth, yen, D-bar, not equal, and a couple other (in all weights). To that I only want to add: superior job on this font. Well done! Randy
Christian: These weights look great. The only thing that stood out though was the numeral 4. It appears heavier than the other numbers on screen and on laser proofs. It is most noticeable with the fractions. I think because it’s closed, it tends to appear heavier than the surrounding numerals.
Christian, sorry to pester, but could you show some samples with the four weights intermixed? Like a block of each with each of the other three used for emphasis. And a question: what made you include the L-bar and l-bar? I’ve been getting mixed signals about whether to include them in my own “base” encoding or not. Oh, and I think your curly braces need more oomph (a highly technical term diﬃcult to explain*). * Credit goes to David Berlow for that one. hhp
Christian: I just emailed you an annotated version of the four weight sample PDF. Hoepfully everything makes sense in my comments.
I’m still chasing down all the gremlins hiding in the curves. Who knew it would be such a pain in the neck? I’m getting pretty close, though. Here is a sample with the weights intermixed. For fun, I turned on the alternates in the Light/ Medium combination. Thanks for the comments.
Anyone have a link to the critique of this? I don’t see it in the san serif section. thanks
Both look solid to me. hhp
Hinting needs some work, though. light/regular lc l, h, p, u have inconsistent rendering in that test, though in 4 weights sample they don’t seem to. Did you lose your hints somewhere?
Christian, can you implement the Spanish characters?
A bit of news: Pill Gothic is featured in the SOTA publication, Interrobang 2 which should appear in members’ mailboxes very soon.
Alavaro, si, he implementado todas las letras necesarias para espa
HI Christian, Nicely done on the extra light. So tricky to balance the modulation! Your obliques look pretty good to me. The following picture illustrates the major distortions that happen when mathematically skewing a font. All letters are thinned vertically as you mentioned (and ﬁxed). The letters that get most distorted are the round ones. You may know all this, but I thought I’d post it anyways for typophile ediﬁcation: 1. The verticals become more thin as noted above 2. The bottom right and upper left (if applicable) thin out 3. The upper right and bottom left (if applicable) thicken up 4. The upper right and bottom left contract (the curve gets tighter) 5. The bottom right and upper left expand (the curve opens up) Also note generally that the distortion gets bigger and the glyph more narrow as the angle increases. In my (limited) experience the tricky and time consuming part is ﬁxing 4 & 5 (usally 2 & 3 at the same time). I was given this tip by Ed Benguiat: The obliqued curve should look pretty close to what it would be if you oblique it half as much. In other words, if you oblique it 20%, make the curves emulate the 10% version. I ﬁnd pasting the 10% version into the mask layer a handy guide. You still have to ﬁx the modulation problem, but it’s a good start. Cheers, Randy
More about curve compensation on italics and obliques at Briem’s website. I think your obliques are ﬁne, Christian.
Superior Eduardo. Briem seems to slip my mind too often. I forgot the rotation the additional degrees! BTW I agree that pill obliques are working. Randy
Obliques: I don’t see much distortion, but I do think they need to stand apart a little more from the roman. Your angle is already steep, and you’ve already made it narrower, so I’d try making the obliques slightly lighter than their roman counterparts, and maybe change some forms too (like make the “a” and maybe the “g” mono, but don’t make the “f” descending — too old-fashioned). And a question: Are fractions for old-style numerals supposed to be full-height, or really small? hhp
Randy, your illustrations are legendary—always helpful. I had seen the explanations on Briem’s website (one of the most helpful sites I’ve ever seen). For these obliques, I didn’t do any curve correction; the angle seemed slight enought to not make a huge diﬀerence. The squarish forms of the letters help also. I did thicken the vertical strokes, however. As for the fractions, I haven’t done a diﬀerent set for the OS numerals. Should there be a diﬀerence? I agree that the 4 1/2 combo looks a little odd.
Odd…. What happened to the font sample(s) and the critiques? David
It is now 6:01 a.m. and I have stayed up all night correcting the curves in my obliques.
(Silently chanting to self: Go Christian, Go Christian, Go Christian. It’s your birthday…)
I should have added to the last post, your obliques have the special duck sauce now! (a compliment) Well done. And the number one way to tell you’re addicted to type: (Silently chanting to self: Go Christian, Go Christian, Go Christian. It’s your birthday…) Joe, enroll in a 12 step program immediately :-) BTW, the number two way to tell you’re addicted to type: Describing oblique curves as *special duck sauce* R
Here is a rough condensed version. I still need to clean up some weight issues (The N, for example), and a few unruly beziers (always the s). I have added some special duck sauce to the curves in the oblique, though. I haven’t corrected the distorted weights on the diagonal stems, however (see the K). The bold is clunky as yet; I need to cut the connectors sharper. I apologize for the large PDF. It’s only one page, but, even subset, 10 ttfs can be heavy : /
Don’t look at that last pdf. This one is much cleaner.
The letterspacing here is looking increasingly too tight with increase in weight. If you have a weight axis going with the stems just getting fatter equally in both [horizontal] directions, and no increase in character width, that explains it. I’m a big fan of uniwidth fonts, but beyond a certain weight the forms will start looking uncharacteristically too narrow (or too wide on the light end) when you try to maintain set-widths. (Was that too terse?) hhp
The characters on the top are justiﬁed (in other words, it’s not a uniwidth font). However, the spacing is slightly tighter (2-3%) for the bold.
A new ‘s’ shape occurred to me today. I’m trying to decide if I should include it as an alternate, or make it the default. It is more extreme, so in some ways makes the font less versitile. On the other hand, it makes the font more interesting, and, I think, justiﬁes the ‘a’ and the ‘g’. Of course, the old S’s would be included in the “Stylistic Alternates” OpenType feature. What do you all think? Does it make the font less useful; does it make it more interesting, or both?
It’s an odd balance to maintain — unique personality vs mainstream usability. Personally, I think the new S adds a bit too much unneeded quirkiness to the face. Maybe this is due in part to the fact I’ve just become used to the more traditional S in previous versions of the fonts. For some reason I don’t ﬁnd the existing ‘a’, ‘e’ and ‘g’ characters to be as radical a departure from the rest of the character set. I certainly wouldn’t abandon the character, as it deserves to be an alternate.
I’m intrigued…what does Pill Gothic look like?!!
I disagree with Grant. I think that ‘S’ will make Pill sell. It doesn’t have to change the way the font is marketed (as a full- family text-capable sans), but with this ‘S’ it becomes more original and will catch the eye of the all too common impulse buyer.
I also like how the straight stroke becomes less obvious and the ’s’ settles right in at text sizes.
I have to say I ﬁnd the new S/s discordant. I wouldn’t even be sure about including it as an alternate. hhp
Follow this link to get the skinny on Pill and other missing threads: http://www.typophile.com/forums/messages/30/18545.html?1066163062 Randy