Reviving an old project

Franz Gratzer

I'm far from being an expert but to me your Caliventa looks very balanced already and I especially like the "ẞ/ß" in both normal and italics. I hope I can achieve something as finely tuned some day.

Steve Gardner

@Craig Eliason, if it is too contrasty, I think I'm probably too invested (not to mention eager to put this one behind me) to change direction. 

Also worth mentioning that this whole project started (way too long ago) as an accident.  Not sure how I did it, but I misplaced or deleted a node in the /o of another sans project I was working on and ended up with this exaggerated 'pinch'.  That very characteristic formed the basis of this project.

There are some inconsistencies in the italic that I've started addressing, but I don't think they'll have a significant impact on the 'contrastiness'. 

Steve Gardner

Appreciate your comment, @Franz Gratzer.

I'm very fortunate to have had the benefit of some great input from a number of people both here and, perviously, at typophile.

Mark Simonson

@Christian Thalmann Viewing the source for the page, it appears that an ASCII 13 character (carriage return, appearing as HTML entity "
") is inserted after some spaces in your text.

Craig Eliason

Steve Gardner said:

@Craig Eliason, if it is too contrasty, I think I'm probably too invested (not to mention eager to put this one behind me) to change direction. 

Also worth mentioning that this whole project started (way too long ago) as an accident.  Not sure how I did it, but I misplaced or deleted a node in the /o of another sans project I was working on and ended up with this exaggerated 'pinch'.  That very characteristic formed the basis of this project.

There are some inconsistencies in the italic that I've started addressing, but I don't think they'll have a significant impact on the 'contrastiness'. 

Fair enough. If the /Oo with their pinches are fundamental, I'd just throw out there that you could also come from the other direction and add more contrast (more dramatic tapering) to the straight strokes. But I'm fine with you skipping this piece of input too of course!
I know from work on my Cuttlefish types that working contrast into sans types can be tricky. 

Steve Gardner

I've started on the Greek italic.  Had a bit of fun with ligatures, especially involving the /rho.  I've uploaded PDFs showing standard forms, alternate forms and ligatures. 

Despite being fairly invested in the Greek, I'm still not sure if it's strong enough to include so, as always, suggestions are greatly appreciated.

P.S. @Craig Eliason, Cuttlefish is a magnificent name for a variable font!

Christian Thalmann

The Latin Italic is very classy indeed!

As for the Greek, I find the closed /theta/ and /phi/ a bit sterile in an Italic setting. I'd promote the open /theta/, closed /beta/, open /phi/, and possibly the curly /rho/ to default in the Italic, though the former two need a bit of work: The open /theta/ is waaay too open (the overall shape should reflect the closed theta AFAIK, so the entry and exit stroke should both roughly be at half height) and the latter is too compressed, especially in the upper counter.

I'm not too fond of cursive /kappa/ and cursive /pi/ in general (not just yours), so I'd keep those as alternates only.

Is the /lambda_lambda/ ligature a bit too compressed as well?

Disclaimer: I'm not a native reader or writer of Greek and all of the above is extremely subjective.

Steve Gardner

Thanks @Christian Thalmann.  I'd found several forms of the open /theta and closed /beta see https://github.com/MicrosoftDocs/typography-issues/issues/759}. 

I've made the /theta's loop bigger, closed it up and lowered the exit stroke.  I've also made the /beta taller and experimented with a third form (see top image below).  Which do you think works best?

I've changed the default/alternate mix as suggested, and ironed out a number of inconsistencies across the whole set.

Christian Thalmann

I still find the /theta/ positively boorish in comparison with the rest of the typeface. (The same goes for the Calibri version that you linked, for that matter.) The /beta/ still strikes me as too narrow, particularly in the top counter.

For comparison, here's what I have in Ysabeau Italic. That part of Ysabeau hasn't been through review with Irene, though, so who knows how solid it is... at least it shows the proportions of the letters in relation to /omikron/ as I consider them balanced. (Theta might need to be made a tick wider.)

Steve Gardner

I think these are better.  Also tweaked the default /alpha and alt /kappa.  I'm like you in that I really don't like that form of /kappa.  Tempted to omit it.

Christian Thalmann

Yes, /beta/ is finished as far as I'm concerned. The /theta/ is much better, but still inelegant IMHO. How about:

• Making the bottom less asymmetric (i.e., less D-shaped and more U-shaped);
• making the joint of the instroke a hard corner (as in /iota/ etc.) to make it shorter and to help with the above; 
• stopping the outstroke within the vertical stem and keeping it sloped, as in Ysabeau.

Might the cursive /kappa/ might be improved by strengthening its vertical stems? Currently it looks light overall, and the weight seems to be in the joint loops, which is a very uncomfortable place.

Also, is your /xi/ leaning backwards?

Craig Eliason

The tail of the default italic /rho feels limp to me, and its terminal quite different from what the rest of the alphabet is doing.

Steve Gardner

@Christian Thalmann, I've tweaked the /theta broadly as suggested.

I'm getting close to dropping this /kappa altogether.  I've tried taking some weight off of the joints and added some to the stems.  I've also experimented with a different form for the instroke (more closely matches the /gamma).  Whatever I try, I can't seem to make it look like a natural fit.

I've also changed the angle of the /xi slightly.

@Craig Eliason, I've added some weight to /rho's tail.  I've also pushed it right a bit.

The terminal was actually created from the in/outstoke of /alpha, /iota, /eta, /tau, etc (see image below).  Every example I've seen shows the tail ending in a similar (i.e.thinning) way.  Do you think there's a form that fits better?

Craig Eliason

Well I've barely dipped a toe into Greek italics so this may just be my ignorance of the conventions. The next tailed rho I design will be my first!
The /alpha/iota/etc. outstroke reads to me like a pen flick, which I'm not sure applies at the different angle and situation of the rho tail. But if, as you say, something like a blunted end is wholly unknown in the tradition, then I concede it may look weird to a native reader.
Another idea: stick with the thinning terminal, but square off the shape of the whole tail more, so the thick bottom part is more horizontal and firm. That might help lessen the "flapping in the breeze" effect I'm seeing; and it seems to me like this face as a whole is quite squarish; and it might also allow more white into that bottom open counter which may be a bit crowded now.

Steve Gardner

This /kappa... ugh.  One of the main problems, I think, is that the letter typically has a thick diagonal and thin stems.  That would be too anomalous in this font, so I tried adding weight to the loops and straightening the stems.  This introduces a little bit of vertical below the diagonal, which prevents the diagonal looking as though it goes from corner to corner.  Not sure it's 'right', but feels better.

Also modified the tail of the /rho, making it a little thicker and squarer.

Christian Thalmann

Sorry Steve, that top of /theta/ still looks a bit weird to me (but better). Maybe try an outstroke at something like a 40° angle downward? It has to end a bit below the middle to make the counter look half-sized. The horizontal looks forced.

Cursive /kappa/: Maybe narrowing it would do the trick...?

As I see it, the /rho/ mainly suffers from that weird serif that so blatantly goes against the pen rules that you otherwise follow consistently... not sure what to do about it, though. Maybe a blunt flared ending like in /chi/, /mu/ etc.?

Steve Gardner

Adjusted the /theta; and tried an alternative tail form for the /rho, which fits the logic and harmonises with /epsilon, /e & /c.  It also makes the transition into the /rho ligatures less dramatic (see /rho_/chi).

Christian Thalmann

Yes, that /theta/ definitely works for me. Perhaps the spacing is a bit tight, especially in the /theta_nu/ ligature?

That /rho/ looks out of character to me, though.  (And those rho ligatures are going to look dramatic, there's no changing that...)

How about this sort of ending? (The tail shape should probably be more angular to work with this ending.)

BTW, the tonos looks off-center to me on a few of the vowels.

Steve Gardner

I think it's time to chose a direction for the /rho.  Either abandon the 'curly' alternate altogether or go with one of the four styles shown below (subject to tweaking). I've shown the third style in a text setting. It's probably the one I like best (particularly given the nature of the /rho ligatures), but I'm not really that thrilled with any of them.

As for the tonos, I haven't started fine-tuning them yet, so this is just where they land using the standard anchor settings.

Finally, whatever I seem to do with the alternate /kappa, it looks out of place.  My inclination is to either discard the alt altogether, or else modify the default to make it curlier.

Craig Eliason

Agree that the third of those is best. But maybe (for default at least) a regular straight-descender is even more suitable? There are a lot of relatively emphatic straight verticals (er, italic-angle lines) in your italic (αηικμ for examples). 

Christian Thalmann

I think 1–3 are all out of character for this typeface. I'd like to see a pinch in that tail... maybe something like 4 with a 1-like terminal, but with more flaring at the end and a harder turn at the knee (perhaps a sharp corner on the inside outline)?

I do support Craig's suggestion, though. You certainly can't go wrong with a straight rho.

Christian Thalmann

Super quick hackjob of my suggestion (my GIMP is acting up since the system upgrade):

Steve Gardner

Haha... ok, @Christian Thalmann, I'll try something along those lines, but I can't 'pinch' it there as it's equally 'illogical' as the original (4 in my image above), which was what started this search for alternatives.

And I think @Craig Eliasonis right too.  The straight /rho might be a better default solution.

Christian Thalmann

It's not less logical than the pinch in your /sigmafinal/, or the thinning of the tail near the bowl in your own /rho/ designs. A bit of pen twisting is admissible in calligraphy, especially concerning swashy terminals.

Steve Gardner

I've tried that shape, but without the pinch.  I accept there's some wriggle room when it comes to pen logic (though I'd dispute the /sigmafinal breaks the rules—if it does, it's marginal), but that pinch in that location was too anomalous, in my view.

For me, it's 1 or 4.  I'm leaning towards 1 (shown in the text setting) unless there's a strong argument against it.  That said, I think I'll relegate this to the alternate and make the straight /rho the default as Craig suggested.

Christian Thalmann

I agree with your conclusions overall. I think the tail might do with a bit of flaring, though; it looks a bit sterile right now (and a touch too light?). But I'll shut up about it now.

Steve Gardner

Maybe a bit naive/unrealistic, given my lack of knowledge of the script, but I thought I'd explore the possibility of adding Hebrew.

Any feedback (even if it's to tell me to scrap it and start again) very welcome.

Jasper de Waard

I can't open the pdf on my phone. Is it broken?

Steve Gardner

Jasper de Waard said:

I can't open the pdf on my phone. Is it broken?

I can open it on on mine (using Adobe Acrobat for Android) and on my Mac, so assume it's ok.  It's pasted below too together with a cyrillic script I've been working on.

Steve Gardner

Update to the cyrillic, including caps (PDF and screenshot)

.