(Cross-posted from
Typophile)
Recently, I've been toying with the idea of returning to the project that brought me to Typophile in the first place: To make a characterful text font. My first attempt failed almost immediately, but I've learned a lot about type design in the meantime.
I'd like to make Traction the female hiking boot of oldstyle fonts: Sturdy and functional, a grippy all-terrain profile, yet organic and human, and within the constraints of practicality as elegant and sexy as possible. The main "bite" should come from bold, almost exaggerated serifs, reminiscent of slab serifs, and broken strokes that reinforce the glyph where it is most needed.
My main inspirations are Quadraat, Marlene, and most importantly, Satyr. I'm avoiding looking at those typefaces now in order not to plagiarize them too much. Hopefully my own take will ultimately stand on its own rather than look like an obvious derivative.
Here's the current status quo:
Comments
/c and /e are a little too narrow.
/q has same problem /b does although not as much.
/g /s /ß -- that sharp corner on the bottom right is contrived and does not belong; it also adds too much weight.
Otherwise I like the look. The caps should be interesting.
I completely agree on {b} and {q}, which I've now slimmed down. I can't say {c} and {e} looked too narrow to me, but I made them slightly wider now and certainly don't mind the way they look now.
I find myself disagreeing on the {g s ß}, though... I find that corner quite natural, æsthetically pleasing, and fitting with the font theme.
As for caps, I've just started experimenting on a bunch of them...
Cheers
I've replied to your suggestions as part of this post on Typophile:
http://typophile.com/node/104308#comment-559668
(I'll save myself the trouble of HTML-wrangling the images over here...)
I like the looks of the font at large sizes a lot, but I am not satisfied with its performance at small sizes. It's legible alright, but the overall weight of the cut I made feels too heavy for extensive reading. I guess I will have to define it as the Medium weight, with a Regular and a Book at lighter weights for body printing.
This makes the next step all the more necessary: to render a spectrum of weights down to Thin and up to at least Bold. Unfortunately, that's something I've never done successfully — the Backstein spectrum was much easier to implement, yet I wasn't fond of the results. I'm actually thinking of outsourcing this job to someone more experienced. Here's a teaser I made for reference:
Can you recommend any online resources for making different weights of a font? Briem is not very thorough in that respect — he focuses on the Bold exclusively, and leaves out a lot of information (e.g., what should happen to the sidebearings in the Bold?).
BTW, here are some alternates for {a} and {g} that I was going to use as the default in the Italic, but ended up considering too crazy:
http://www.cinga.ch/type/Traction_specimen.pdf
Anything that needs to be addressed before the work on the different weights begins?
I like the smallcaps too by the way. The /germandbls (Ringel S) stand out a bit though, it seems a bit to wide and the contrast appears a bit odd to me.
I've uploaded a new version of the PDF specimen to the old location. On page 9, I've set a few words of a Roman text in Italic to see whether it is now distinct enough. What do you think?
And here are the new /ß designs:
Furthermore, on page 10, I've taken the stylistic alternate /a of the Italic for a ride. It's rather exotic, but I like it. Maybe the right foot needs to go a bit lower, though.
I'll have to smooth out the /G's curve, and the beard-shaped terminals probably need sharpening, but other than than I like the first impression:
Hi Christian! As you perhaps know from this topic, I’m investigating Renaissance standardizations and systematizations of type.
One of the punch cutters I’m especially looking at is Robert Granjon. I noticed some accidental similarities between your patterns for Traction and those which show up in Times (New) Roman, and which find their (indirect) origin in type from forenamed French Renaissance master.
Your type is somewhat bolder and your pattern deviates at some points, and a bold version was never made by Granjon, of course. However, your bold and and italic seem to follow different patterns, of which some remind me a bit of the ones made for the Linotype composing machine.
As you can understand, this is very interesting stuff for a researcher!
Anyway, I was worried Traction might be a bit too tall for its own good (whereas Satyr has this soothingly wide stride), but if its proportions match those of Times, they can't be all that inconvenient.
@ Scott: I like ligatures in general, and thus just put in all the usual suspects. I also like having a ligature to use in my own name.
You’re very welcome, of course. My posts definitely mark my interest in the origin of sources.
‘Interesting to see similarities to Times New Roman […]’
Actually, I think the similarities are quite amazing! Obviously you took the long and tedious road, but the similarities are so striking, that you almost could have opened Times (New) Roman in a font editor, put roughly 10 units around the contours, tweaked the contrast and contours a bit, by for instance making the stems slightly hollow, and shortened the serifs. But, of course, this would have implied an infringement of the copyright of T(N)R and would have hampered the originality of your hand.
Times (New) Roman itself is subject to a similar case, in which Morison made original sketches, which Lardent claimed never to have received, and of which –according to Tracy– the resulting 327 series showed unexpected similarities with the Plantin 110 series. It’s all amazing…
‘Wouldn't you find this level of rhythmic similarity between just about any two text fonts, given that text fonts can't be too originally shaped lest they trip up readers?’
Well, there are some exceptions IMHO. For instance the for centuries dominant Jenson/Griffo/Garamont model, which shows up for instance in Adobe Garamond (mostly based on Garamont’s Parangon Romain) is a bit different, but does not seem to trip up readers.
I always had the feeling that T(N)R is related to a different set of French Renaissance ‘display’ proportions, which for instance show up in Van den Keere’s Canon Romain. I could be wrong though.
I see Traction's most iconic area of use at mid-to-large sizes in branding, advertising etc., where its organic curves and grippy detailing can shine, but I do want it to be functional at text sizes too. The parallels to TNR suggest that it will do just fine.
Well, there are clear differences between for instance Jenson’s roman type and the ones cut by Griffo when it comes to details, but the types of the latter clearly follow the structures of his French precursor. But because you mention it specifically, let’s focus on the /d for a moment (as researcher I’m always curious, you know).
I took the /d from Times New Roman (on the left), expanded the contours with ten units (in the center) and tweaked some points. For comparison I distilled a /d from the PDF you posted on Typophile (on the right). The subsequent four steps below took me less than five minutes. I only looked at your original and I definitely did not trace it, so the outcome and your /d differ a bit still. But the differences are not larger than the ones between the different Times-versions from Monotype, Linotype, Scangraphic, Berthold, etc., I reckon.
If you were capable of coming so close to the proportions and metrics of T(N)R in Traction without even looking at T(N)R, this is absolutely amazing in my opinion!
Note that most rounded letters were originally a bit wider in the early stages of Traction. The screenshot below shows an older instance of the /d superimposed on the current one. I tweaked it after I felt the counters of the rounds gaped too open compared to the /n-type characters (the /b was a particularly prominent offender). In the same vein, the originally extremely narrow /a relaxed a bit to fit the overall rhythm.
TBH, I don’t think that the similarities between Cambria and Times New Roman are the result any ‘concrete subconscious model’ floating around in Jelle Bosma’s head, but that this has everything to do with the fact that Cambria was meant to replace TNR as default font in MS Office.
I just pointed Jelle to this discussion (I know him since 1978, you know). Perhaps he is willing to comment.
Well, now you have David’s blessing, to be Frank, I know my place and it’s time for me to be off in my not-completely original Trabant (New) Roadmobile (freely interpreting here an e-mail I received from a renowned Dutch type designer today) and to travel back to my Parallel Universe, where I will be cheaply approximating parts of different type ever after. Arrivederci!
Until today I never compared the proportions with Times NR. The heights are similar, with Cambria having a larger x-height. Of course the horizontal proportions and most of all the sidebearings are quite different.
When experimenting with a range of widely used ppem sizes (used around 2003), you end up with optimal pixel x-heights, and ascenders for these sizes which translates back into outline heights. Select suitable features and work your way up.
Times NR became such a widely used typeface because it was selected again and again for a wide range of usage's. Its proportions must be such that they hit a sweet spot, which allow it to be used with linefeed large and small, lines long and short, good printing and bad printing, for books and flyers. It even got selected as standard text font for laser printers and ultimately for Windows. No wonder that its heights are quite near the heights you get, if you deliberate search for heights that work well for a screen font.
Anyway, Rainer suggested the name "Subtraction", so that would work nicely. ;o)
In your place, I would just finish it and release it, it will be seen as entirely different from the text family anyway.