Vainilla – Something Old-school

Hello. I’m a recent emigré from Typophile, and I’d like to submit my work for criticism and suggestions. Here is a complete family – Vainilla Avec (“with serif”) and Sans, both in regular and bold roman and italic – for your consideration. The Sans is the more interesting, having angular stress and Garalde construction, but it is not identical with the Avec; they should work together fairly well. The serif is currently under heavy revision; for spits & giggles I’m also showing it; its codename is “Q.” TIA!


  • Please use your real name; this is a real names forum.

    First impressions of the serif font: roman has a bouncy look to it (overshoot problems with various characters), /Q tail doesn't work; italic is too tight.
  • Vainilla avec 'a': low / left curve too light
  • @George Thomas:

    "Please use your real name; this is a real names forum."

    A bit late for that, but point taken.

    "First impressions of the serif font: roman has a bouncy look to it (overshoot problems with various characters),"

    Indeed! I'm definitely struggling with finding a) the right amount of overshoot and b) the right settings for stems.

    "/Q tail doesn't work;"

    You're probably right, but I like that style of tail, and I'm not sure what would work best here...

    "italic is too tight. "

    Very much so. Corrected.

    Thanks for the input.

    -- Michael Vokits
  • @Ramiro Espinoza:

    "Vainilla avec 'a': low / left curve too light "

    Yep... definitely an "opportunity for development," as managers might say.

    -- Michael Vokits
  • "A bit late for that, but point taken."

    Not too late at all, merely change your name in your profile.
  • Heh, just discovered how to fix that as you posted.
  • Current WIP (just the roman and italic, regular width, serif).
  • /e, /c need to come down a little at bottom; /a, /t are too low (see /o).
  • Nov 15 . Seems to work well enough, or maybe it’s that I’m tired of working on it...
  • Christian ThalmannChristian Thalmann Posts: 1,480
    edited November 2013
    Looks appropriately oldstyle from afar! And I like the name Vainilla.

    Up close, though, there are a lot of small things that feel inconsistent from one letter to the next. For example, /g has less contrast than the other letters; /u has heavier strokes than /b, /s and /S too light, too cursive, and hovering above the baseline; spacing is uneven in the Italic (look at the combination /isi or /lv).

    Finally, some shapes strike me as awkward, such as the break in /p /b /d (if anything, I'd expect a break at the opposite bowl/stem junction), the scrunched-up tail of the /Q, or the lumpy /a with its unfinished terminal.

    What software are you using?
  • @Christian Thalman:

    Thanks for the input! Humbling, but oh so necessary. And appreciated!

    /S and /s are really difficult to get right; I could not have passed Monotype's interview, alas. They're still too calligraphic, but their color matches better now.

    /a is another letter I have troubles with; I can't decide between the three shapes I have, but I think the second might be best, as it matches the beaks on /f and /c better.

    /n and /b have the exact same stroke thickness, but optical beats mathematical every time. (I take some comfort in knowing that Slimbach did the same thing in Minion.)

    As for /p and its ilk, their broken strokes are a tribute to Scala. Maybe I should reconsider this feature, no? Maybe make their bowls a bit more angular while I'm at it.

    As for software, I use TypeTool, with FontForge to build PS-flavored OTFs.
  • Hi Michael,

    it certainly wasn't meant to be humbling!

    Yes, both of your new /a designs are better, and it's a good idea to use the same terminal style as in /c. They could still all use a bit smoothing of the curves in the bowl.

    Great job on the /g, it's sexy all of a sudden! I'm wondering whether it might look even better with a /c-style pointed terminal in the tail rather than a perpendicular cut?

    /S /s are better, but still strike me just a tad oblique. The capital might also have become a bit too heavy around the spine.

    Gaps in /b /d /p /q: Fair enough. They don't draw attention at text sizes after all.

    The /i seems a bit unbalanced at large size, since the serifs are so asymmetrically weighted. Similarly, maybe the lower spur of the /b could point a bit more backwards so as to play well with the upper serif.

    Maybe /Y would look better if it were a bit wider? And the tail of the /Q looks nicer now, but still feels disconnected from the body. It looks like an underlined /O to me right now. In the /Qu ligature, maybe the strokes all lie a bit too close to each other?
  • *yawn* Bleepin' tired; now I can sleep. Cheers!
  • Christian ThalmannChristian Thalmann Posts: 1,480
    edited November 2013
    Good job on the /S! The new one is strong and handsome.

    Here's some more feedback for when you've slept. Tell me if you get fed up with it, and feel free to disagree on any and all points. ;)

    The /a still has some bumpy curves; see attached image. There's also a bit of a bump on the very bottom of /g. Best go through the glyphs with SpeedPunk or RMX Tools to localize those and get rid of them.

    I like the new /Qu ligature, but it bothers me a bit that the beginning of the stroke is barely poking out of the /O body. I'd say, either make it poke out decisively, or not at all. The rough brush-stroke intersection of the new /Q feels a bit out of character with the calm rest of the font. How about using the wide tilde you had before, but moving it a bit to the right so as to imply a connection between its wavecrest and the /O body?

    /b looks more balanced now, but the spur feels too heavy. Note that it's a "light" diagonal. Maybe you could transfer some of its weight to the bowl connection instead, which lies along the "heavy" diagonal?

    On the gap in /b/d/p/q: I'm wondering whether it would look nicer if the thin stroke were angled inwards a bit more, making gap between the end of the stem and the bowl more prominent? At small sizes, /p looks flat-topped.

    /Y still looks too narrow to me (lower the branching point without changing the opening angle?). /B is a bit squarish, especially the top of the upper loop.

    The lining figures strike me as systematically too light. /3 /6 /9 look a bit rotated.

  • Michael VokitsMichael Vokits Posts: 201
    edited November 2013

    Yay, another type of feedback to enjoy! Especially since this type doesn't make my ears ring. I really need to turn the amp down.

    Seriously, though, thank you for your time, advice, and all that -- I appreciate it.

    Lots of work this week; this is beginning to look halfway decent. The bows of /pdbq are a bit.. flabby, I think -- the outer curve sticks out too much, rather than tucking in the inner.

    SpeedPunk and RMX Tools look wonderful, but waaaay out of my price range, alas. Instead, lately I've been manipulating node-less curves and auto-inserting extrema. (This is also a great way to embolden glyphs.)
  • Christian ThalmannChristian Thalmann Posts: 1,480
    edited November 2013
    Good changes. This is coming together nicely.

    Minor things now:

    — The bowls of /b/d/p/q don't look flabby to me; to the contrary: The more calligraphic and less circular shapes lend the letters some bounce and turgor. You're right about the stressed part of the bowls being too heavy, though.

    — The gaps in /b/d/p/q are now so tight they disappear at small sizes. I'd either get rid of them, or make them large enough to be decisive. Conversely, the gap in /B feels too large. Maybe typeset a bunch of words with those letters (and /R /P, and /6 /9) and see whether they look consistent.

    — The whitespace under the peaks of the /M is unbalanced. Maybe set the right-hand stem upright.

    — Some of the thick strokes in letters look a bit too heavy compared to the other letters. For instance, the diagonal of /Z, the spine of the /S, and the bow of the /D.

    — The oblique cuts at the bottom of the /v and /N don't quite feel right to me. For instance, in the /v, it looks like the thick stroke is cut off perpendicularly, and the thin stroke is then stuck to it. Maybe if the cut were more horizontal, it would look more like a shared feature of both strokes.

    — /t looks a bit unbalanced to me. Maybe move or extend the crossbar to the right a bit?

    — The right part of you /& feels too technical for your humanist font.

    — /c has a sharp change of curvature on the top point of the inner curve that would benefit from some smoothing.

    Once you're done with the shapes, the spacing could use some fine-tuning.

    The RMX Tools are still in free beta version for Glyphs. You can get a free trial version of Glyphs for 30 days.

  • Alas, Glyphs requires a Mac, which is definitely not in the budget. (Not to mention that I can't get used to a one-button mouse!)

    btw, which /b looks best? I'm not liking the right at all, but maybe it's just the clumsy join to the bowl's top.
    b.png 48.8K
  • Nice work on this one Michael. I'm not too familiar with this type of typefaces, but it looks nice. So take whatever I write with a grain of salt, this is an untrained eye watching.

    To my eye the /u seems to hover a bit in comparison to the other lowercase letters, seems like the bowl/shoulder should be a little lower maybe? Or maybe it should depart from the stem a little lower.

    The /O and /Q appear to be a little lower than the other uppercase letters, I'd push them just slightly higher, to make them appear of equal size.

    Also the /? feels a little out of place to me - most of the letterforms are quite curvy whereas your /? now seems like it could be a bit wider and the top could be bigger in relation to the rest of the mark.

    The /? also seems to have an angle or reversed slant that I don't see in other letters. Same a little bit with the /a, it feels like it's 'falling' to the left a little bit, I think you could fix that by moving the curve before the top terminal to the right a little and maybe make it a little bolder?

    Also I agree with the comments about the gaps in letters such as /b /p etc - either make them a little more obvious or get rid of them. I do like the gaps, gives your typeface something extra (you could also consider to offer them as contextual alternates or something).
  • Hello, all. Many thanks for the critiques! I'm going with Jan's idea of the closed bowls as normal, and open as alts.

    Can I get some feedback on some ampersands? I'm playing with transparent /e/t ligatures, and I think I like the ones with a raised /e and normal /t best.

    Anyhoo, busy week, much preparations for Thanksgiving. I hope the week is treating everybody well. Cheers.
  • Personally I like ampersands as something that can be extra playful within a typeface. I might step on someone's toes here though, but I feel that ampersands can differ (slightly) from the rest of the letters (they often stand out in a sentence or combination anyway). So if you have the chance, be a little playful. I like the /e/t combinations, seeing them as they are now I prefer the same one as you - but it might help to set them in actual combinations of words, letters or numbers.

    Enjoy your holidays!
  • Man, the worst part of the holidays is the week you spend with the flu afterwards. Maybe that's just me. Anyhoo, here's the latest, with a few alternate characters. Wish I could get the fraction code going for the italic. OK, off to sleep again. Mumble flurble squizzle m'shurbling fizzbang.
  • Just for fun, revised the sans a bit. Much nicer. Still basically a Renaissance pastiche with light weight minus the serifs -- the letter modeling is incorrect, alas -- but I like it.
  • The main impression I'm getting is one of unevenness. The weights of the stems and thick parts of arcs appear different from one glyph to the next; there are zero-contrast caps interspersed with high-contrast ones; the stress angle changes between similar letters (e.g., /c and /d); and even the arch shape is completely different between /n and /m. Is that intentional?
  • Wha...? Here's a screen grab from the .pdf --

    Maybe your pdf reader is substituting different glyphs? Gimme a screeny, please.

    Anyway, there are no zero-contrast caps. There are no high-contrast caps, either. /c and /d have almost the same bowl. The /m is exactly the /n duplicated.

    Hmm... I just tried viewing the .pdf in the browser, and I'm seeing a ton of pretty much random glyphs, mostly an uppercase gothic, in random places -- not even full lines. Is that anything like what you're seeing?

    Crap, I really don't need another interesting problem... somehow, installing VTT screwed up my fontview settings, and I can't figure out how to restore them. Regedit won't accept my changes. "fontview *" tells me I have an invalid font file, but calling "fontview *" works
    just fine. Yeesh. Anybody have a solution?

    Anyhoo, I wonder if LuaLaTeX is messing something up. I'll try it with XeLaTeX. Let me know if this works out.

  • Christian ThalmannChristian Thalmann Posts: 1,480
    edited September 2015
    Hi Michael,

    your screenshot looks pretty much like mine. I see the same issues in your screenshot. Allow me to demonstrate:

    There's also some inconsistency among the high-contrast glyphs. Compare the thin stroke at the top of /D with that of /G, for instance. The latter is much thinner.

    /c and /d have almost the same bowl.
    Stress angles:

    This close-up also shows that the curves are very bumpy and could use some tuning. Can you see the two kinks on the inside of the apex of /c, for instance?

    The /m is exactly the /n duplicated.

    No. Here's the /n in red superimposed on the /m in black:

    Cheers, Christian

  • Hi Christian. Hmm, much to think about, starting with what I think of as high- and low-contrast. I must lump everything which isn't Bodoni or Helvetica into a middle area... but I do see your point. The contrast on /C (about a 2:1?) is quite a bit more that that on /E (about 3:2?). Hmm, math time. Let's see, a 30-60-90 triangle in the pen, so divide by the sqrt (3)... 80:46, 1.74:1 -- call it a little over 3:2. Cripes, that isn't much of a contrast at all, is it.

    Gah, you're absolutely right about /c and /d... I'm working on many projects at once, and I've forgotten what I've done to each. I think the last time I touched the /c was in 2012. btw, how did you work out the relative angles of their bowls? What should I look for in determining it for myself? I think I'm only looking at the outer part of the bowl, which is pretty close in both glyphs -- well, except that /d has a bulge s/w. And yes, the curves are really bumpy. It seemed like a good idea at the time... Correcting now.

    As for /m and /n... hmm. I was so bloody certain I'd just duplicated the /n and extended the arm to connect. (No, wait, it was the italic that had that -- this time I copied shapes to be certain!) Ah well, it's a truly lousy pair of curves anyway; I just can't get the transition from straight to curve right -- it's always lumpy somewhere to the n/e.

    Thank you. I deeply appreciate the advice.
  • OK, let's see how I did today.

    Much better, methinketh. No more lumpy curves on /c, /m and /n are consistent -- (probably too consistent, given that /m is just a duplicated /n (honest!) -- but then, you see the same thing in Minion and Trinité, so I'm at least in the land of Established Precedent. The stress angles are much more consistent as well.

    The bad:

    The thins of /E may be too much so in comparison with those of /C and /P.

    /D, now that I look at it, is too monoline.

    The left half of /o is narrower than those of /c and /e, but the latter two are shorter, so I think it evens out.

    The Ugly:

    Jeez, that /a has a pot belly.

    The bowl of /g looks huge. I think I may go back to a two-story construction.

  • Hi Michael,

    on the /n/m: Did you deliberately pick the old /m arch shape over the old /n? Personally, I found the latter matched the style of the typeface better. The former strikes me as rather angular and abrupt; a motion that appears nowhere else in the face. In any case, if you consider /n/m too consistent now, you could make /m just a bit narrower; most faces do that.

    I think the new /E/F/H work much better than the old ones (though the hook of /G is still monoline). There is an ongoing problem of weight consistency, though: The round strokes and some of the diagonals look significantly heavier than the vertical stems. This sometimes even happens within a letter (such as the /M: the right stem looks heavier than the heavy diagonal). Similarly, the thin strokes are not equally thin between letters (compare the heavy thins in /D to the lighter ones in /C).

    I'm also getting a certain "blobby" impression from some of those heavy spots in your round curves. Maybe that will go away when you make them lighter to match the stems, though. Maybe it could also help to think of the inner and outer curves of glyphs as continuous organic entities, rather than dividing the glyph into thick and thin parts joined together.  For instance, your /O and /o look every so slightly "fig-shaped" to me, with a certain pointiness toward the upper left. I think it's due to the diamond-shaped contour of the counter.

    I actually don't mind the bodies of /a and /g. The tail of /g is a bit unbalances, though, pulling too much to the left. I had that problem, too, with my recent serif font (where the double-storey /g was the default solution and the single-storey alt felt weird at first). I found it helps to think of the tail as starting with a pronounced downward motion before curving around. In cursive handwriting and blackletters, the tail actually often starts out with a rightward trend!
  • Hi, Christian. Nope, it's a new shape entirely -- I was trying to get away from the "Bembo pastiche" Frank diagnosed, so I went for a more symmetrical arc. Err, in the newer "FELT" glyphs, that is.
    I still like the Bembo design better, but I think this fits the style, alas. It's also why I ditched the horns on /T /Z /z and /7.

    Yep, stroke consistency is something I'm trying to work on. It really is better to match letters to letters, rather than just on their own (especially since the damn vertical axis shows up so often so often). As for /C and /D... yeah, I completely see that now.

    The bottom is OK, but the top... oy.

    Hmm... I'm afraid you've lost me about /O /o diamonds and figs, but I know your insights far outstrip mine. I just see curves. :(

    But yes, I definitely have a tendency to blobiness. I'd been trying to do something "artistic" -- like I know what that means! -- rather than a nice, simple, clean curve. Plus, I finally discovered that I can simply delete an on-curve point and manipulate handles, letting the on-curve point arise naturally from the shape. Bit of a revelation, that!

    Ooh, and good points about /g. ty.

  • D'oh -- wrong font. Disregard.
Sign In or Register to comment.