The J would fit the style better if it turned up to a flat terminal like C and S. Tail of Q could use some weight. Otherwise, everything looks pretty good at this distance. A valid critique of weight and curvature would require a vector pdf for a closer look.

It looks a little to me like you haven't compensated for the visual stress of horizontal vs vertical, or for visual center in S etc., but if so, it also seems like this face kind of pulls it off, making it a strong choice for sideways and other rotated designs while still looking good upright.

Generally this is looking pretty good, but there are some global issues and a few specific things you probably ought to address before doing the lowercase.

I'd say that you do need to compensate for the horizontal vs vertical. Making them physically the same thickness makes the horizontals look heavier than the verticals. You need to make the verticals about 5-10% thicker than the horizontals (or the horizontals that much thinner) to make them look the same weight. This is just how human perception works, so virtually all Latin/Greek/Cyrillic typefaces take it into account.

Fiddly stuff: The join of the leg of the R needs some slight optical compensation. Same with the intersection on the B. The S looks top-heavy.

Anyway, good luck with it! Doing a typeface is a major project.

Cheers,

T

Thanks for the suggestions guys, I'll play around some more and see what comes out. As for the question of horizontal v. vertical stress, I focused on the particularities of the original signage. Here's a google street image: http://bit.ly/d83vv6 (I would have included my own photos but I'm on a brief vacation in Montreal)

How do you feel about keeping irregularities in doing revivals/homages? This might be a topic for another discussion but when does a typeface's features go from eccentric to problematic?

I hope this doesn't come off as defensive as your critiques are exactly what I need to hear and a 100% appreciated. I'll post my progress and hope to hear your opinion.

jrl

Well, it looks to me like you've already made a distinctly original departure from the inspiration, whose top and bottom curves appear to be plain half-circles. So I'd suggest you do whatever helps make the most of what you've created.

Also, these optical adjustments can be very subtle, and it bears asking the rhetorical question of what makes you so sure they aren't in the original. They can easily be lost in the perspective of a photograph.

There's no single right answer to the question of keeping irregularities. It depends on what you're trying to achieve. How do you want the typeface to be used? Often you may have to trade off authenticity against utility/versatility. The relationship is usually non-linear. You may want to maximize the sum, or you may find one more valuable than the other.

For reference, see the page here:
http://www.adobe.com/type/browser/landing/hypatia/hypatia.html

Note the "flavor vs utility" graph. Same idea here with authenticity vs utility/versatility, IMO. But depending on what you're after, you may value one more than the other, in which case the ideal balance would be different (but still quite likely not right at either extreme, IMO).

Regards,

T

I've attached a sample pdf of the adjustments I've made so far in my original post. I've evened out the horizontal stress to make the stroke weight look more uniform. The N and M feel a bit darker than the rest of the letters but that might just be me. As I said before I'm debating whether or not to do a lower case. I'll work on a few glyphs and post what I come up with.

@cerulean & Thomas Phinney: You both paid me a compliment in your first posts and I focused on being defensive in my reply, which isn't very fair. I've of course kept a copy of the original wonky glyphs because I like them but thanks to your advice I've ended up with something I can seem myself using much more.

Thanks again and I hope to hear what you think,
jrl

Looking really good. Get ready for more picking.

The diagonal of N is severely heavy. The diagonal of Z is tapering from one width to another. The vertical of P is thin, probably an accident.

Now the tough stuff. The upper right corner of B,P,R is lumpy (i.e. its curves aren't matching up). The inside curve is formed well, while the outside curve starts sloping down too early and then bulges out.

The S appears to be leaning to the right; I suggest nudging the top half to the left a tiny bit so it sits centered, and then working on the middle curves more.

I think it's a good idea that you're holding onto the version with apparent reversed stress. It seems like the sort of face that people might like to use for setting sideways, and having a separate version optimized for that would be a neat selling point.