Hello everybody. I'm designing a contemporary, Elzevier-inspired serif typeface and now I've come to a rather interesting point—old-style figures. Normally you'd expect a stressless zero that's almost monolinear, but that it relevant for the basic "Regular-Bold-Italic-Bold Italic" model. I have a Light and a Black master—that's a kind of a problem, because in black, such monolinear stressless zero is too light and too out of style. I suppose that I could use other solutions for old-style zero such as bottom point raised above the baseline and/or reversed stress. What do you think?
Well, the contours are yet nowhere near to the final release quality, but for now I need at least to elaborate the ideas of the glyphs.
That makes it harder to judge things like the size/effectiveness of the oldstyle zero, although the design direction seems viable.
My own preferred form of the OS zero is where the left side is thick while the rest is thin. Example: http://www.identifont.com/samples/berthold/Whittingham.gif
Hrant H. Papazian said: That's an interesting idea, I think I'll try it as well. Although its for modern serif and I've got transitional one, I think it's better to design all possible versions and then to choose.
I originally designed the Brill text 0 with the heavy side on the left, but was unhappy with how feeble it made 00 look at the end of a number. Since 0 occurs at the end of a sequence far more commonly than at the beginning, this seemed better, at least in this typeface. [With regard to historicism, both forms occur in 17th and 18th Century writing manuals, but the form with the heavy stroke on the left is much more common.]
Of course, if one cares about consistency in modulation patterns, this form of 0 is only viable in typefaces with this kind of pattern. If you're building a typeface design around a modulation pattern that doesn't involve this model of thinning and thickening, this form will stand out at least as weirdly as a monoline 0.
[See how I did all that without mentioning the word 'stroke', Hrant. Just for you.]
Certainly in a monoline design a contrasty zero is rarely good, but to me no worse than the classical "thin ring" model.