Quador – a squarish serif


I’m working on a serif display-font, it has constructed elements of a slab serif, but also some of the Old Style genre.

The idea is, to design the characters more squarish – I like the constructed look and the the counters are kept more open.

And I’m not sure about two parameters: How far can I go with the superellipse shape and the stroke contrast?

What do you experts think? Every comment is welcome!

Thanks, Arne



  • What makes you approach this as a display type? Looking at your design a whole array of reference typefaces come to my mind that have one or another feature you are also experimenting with, and none of those types were primarily aiming to be for display use. I'm asking because in regard to your questions the use case and size probably will give the best pointers to decide in terms of shape and stroke contrast. One the one hand features like the angular insides of the n shoulders echo something of a Swift, and seemingly aim to advance text readability. Some of the more strong and wedge like serifs make this design rather suited to smaller than larger sizes, in my view. On the other hand, the extreme ellipse can really only work when the point is to capture the eye's attention. I don't think the problem is how far can you push each aspect, but how to combine them.

    Overall I find myself glancing at the serifs at lot, which seem to have a couple of different appearances and come off as somewhat undecided and inconsistent. To my eye also the overshoots, especially in the caps, seem like they would need to be more pronounced.
  • Hello Johannes,

    thank you, I agree with your statement and this is very helpful! It’s true that the /n is similar-looking to the Swift, but that was not the typeface I was looking at. It was my own typeface „Quitador“, which is more strictly based on the superellipse. Finally I don`t want to make a new „Quitador Serif“, but still toying around with these rectangle shapes and Old Style elements. – and yes, not only for display, it’s for text.About the serifs: Could you precise a little bit what you exactly mean, terminal, spur …?

  • Hi Arne,

    As you perhaps know, I investigate patterning and I noticed some similarities between your Quador and Gerard Unger’s Swift. Of course, it is always possible that this is purely a coincidence. In any case, IMHO the most difficult part of type design is to develop one’s own idiom. The hand of Gerard Unger is very recognizable, like the hands of Johann Michael Fleischmann, Eric Gill, Hermann Zapf, and Jan van Krimpen, to name a few illustrious type designers.

    I am lecturing type design at the KABK for exactly 30 years now and although it is nowadays very easy to investigate digitally stored typefaces directly, I still think that the best way to develop one’s insight and exclusive hand is by starting completely from scratch. Of course, this requires a very long and intensive study before a professional level is reached, but this approach guarantees a unique outcome, which is the only way to prevent our wonderful, precious, and historically deeply anchored profession from dilution. However, maybe I am hopelessly old fashioned and perhaps you did start from scratch anyway, despite the similarities.

    September 2015 I wrote this on Facebook: ‘On 10 September 2015 Adrian Frutiger, one of the absolute giants of the type-design profession, passed away. I attended a couple of Frutiger’s talks in the course of time, starting around 1984/85 when he gave a presentation on his type designs in Amsterdam. I only talked with him once, in Basel at the ATypI conference in 1986. I showed him some of my first –analog– type designs (this was eight years after my first lesson from Gerrit Noordzij). I recall that despite the fact that he considered my work authentic, he was annoyed because I called myself a type designer. In his opinion one was a type designer after a *long* period of practicing and refining one’s hand, and acquiring knowledge and experience. Adrian Frutiger was right, of course, and his remark made me realize that type and its designer have to mature. Although I was technically capable of releasing my type designs digitally towards the end of the 1980s I took my time until 1993, when I released DTL Documenta.’

    Best, Frank
  • Hi Eric,

    thank you for your comment: first of all: This is a first sketch and a try how can I involve these rectangle shapes in a serif? It´s only the first idea, not a finished typeface. My intention was not stealing from Gerard Unger. I played with the shapes of my Quitador and there was nothing digitally stored. But I see, I have to find another way, it´s too similar, but I havn't recognize that before. I met you 1990 in Amsterdam, when I made my graduate at the ANRT in Paris. Since that time I pay attention on type-design, but I was more interested in constructed alphabets and sans serifs. So you can say after 17 years this my first try to develop a serif typeface and of course this is a new learning process and that is why I posted it on typrdrawers.

  • Christian ThalmannChristian Thalmann Posts: 1,521
    edited April 2017
    Frank, you need to stop accusing people of plagiarism, it's rude.

    Arne: The /g strikes me as out of character. Not only does it not fit the other characters, but its top half even looks like it's from a different font than the bottom half. The rest looks like a pretty solid start to me, and reminds me a bit of Cambria and Charter.

    The spacing of the round caps might need a bit of tightening, and /C and especially /G could use a bit more, I don't know, muscle tension?

    In your exploration of parameters, my tastes would tend toward the rounder and higher-contrast options.
  • I saw some of your other fonts and their quadratic shapes (which I liked a lot) and did not think you were aiming to reference Swift, but the effect comes to mind. Especially since for me personally the display character of your face is not the prominent aspect, yet, anyway.

    About the serifs: Could you precise a little bit what you exactly mean, terminal, spur …?

    A couple of examples that caught my eye: The a terminal seems very soft and round compared to the abrupt d or u terminal. The x-height serifs, like i or n or j seem very triangular, but the baseline serifs seem very pointy, flat and elongated. Some of the cap serifs look very inward facing and trying to balance the glyph, like with A or V, but others seem very dominant and pronounced and expressive, like in E or Z. And the curved caps are not consistent to one or the other, for example C or S: One serif looks more thin and wedge-like, others more triangular and with a lot of volume. And the teardrops in a, c, f and r all look similar, but too different in tension, angle and weight of the ball terminal. Also the spurs in b or q are not in sync with the thickness of the strokes otherwise and seem very pointy and sharp. This is, of course, only a subjective impression, and I write them without too much analysis - I suppose "first impressions" are helpful only to a degree, but alas, this was mine.
  • Frank, you need to stop accusing people of plagiarism, it's rude.
    Hi Christian,

    We’re a bit on different wavelengths, I reckon. I am only observing, investigating, analyzing, and questioning. Of course, we can switch here instantly to details mode (which I consider the easy part because consistency matters regarding shape, contrast-flow, contrast, and spacing are purely relative to the concept), but I think it is relevant to discuss the concept first. Your remark make me wonder how your type-design tutor(s) handled the originality and idiom issues.

    Best, Frank
  • Hi Frank,
    I am only observing, investigating, analyzing, and questioning.
    Nothing wrong with that.

    What I find rude is your base assumption that pattern similarities between typefaces are almost certainly due to plagiarism rather than convergent evolution.

    Of course, you know orders of magnitude more about patterns than I do, so maybe you're seeing some compelling evidence here that I don't. Then again, having been at the receiving end of one of your inquisitions, I know you can be wrong about these things.

    I also find Frutiger's behavior in your anecdote rude, for that matter.
    Of course, we can switch here instantly to details mode
    There's nothing wrong with concept mode per se.
    Your remark make me wonder how your type-design tutor(s) handled the originality and idiom issues.
    My type-design tutors are called Typophile and TypeDrawers. :Þ

    As far as I'm concerned, any typeface drawn from scratch without reference is an original, even if it ends up with strong similarities to existing typefaces. Whether or not it is unique enough to hold its own on the market is another matter.

    Cheers, Christian

  • Hello Frank: I don´t think, that we can separate concept and detail mode. Details are part of the concept and design. Especially today, where we have much more text-faces than 30 years ago and the scope is getting smaller. Take a look at your own typeface „Action Condensed“ compared with „Univers Ultra Condensed“. Similar concept, but different details. There is no plagiarism, you made your own interpretaion of a condensed sans serif very well and with many different details to the Univers. – But without looking on details it looks similar … 

  • ... I'm sorry, then my example is from Erik van Blokland
  • PabloImpallariPabloImpallari Posts: 527
    edited April 2017
    One of the reasons there are many similarities in patterns may be that the lowercase alphabet is much more standardized than the uppercase alphabet.

    Once you have you /n then pretty much every other lowercase follows suit (there can be some ±1 variance, but is small) whereas there is much more room to play with the proportions in the uppercase (up to ±5 is normal)*.

    Typefaces that defy traditional lowercase patterns are very rare. An example may be Platform.

    But in the Uppercase, there are more options. From classical trajan-like proportions, to the homogenized modern ones like Helvetica, or the almost mono OfficinaSans... and you can play with everything in between, see for example the puzzling proportions of Kabel uppercase

    Plus, in the era when typefaces where fitted into 18 units systems, lot's of /W and /M got squashed and narrowed thus conditioning people's perception, hence natural /W like Caeiclia's /W may seem wide if you compare to most Mono/Linotype fonts.

    * ±1 and ±5 where /n=32
  • Your remark make me wonder how your type-design tutor(s) handled the originality and idiom issues.

    That is a bit rude again, Frank. Your comments are welcome, but I would suggest that we discuss with more respect and tolerance to younger designers and don’t attack people personally.

  • Christian ThalmannChristian Thalmann Posts: 1,521
    edited May 2017

    That is a bit rude again, Frank. Your comments are welcome, but I would suggest that we discuss with more respect and tolerance to younger designers and don’t attack people personally.

    I didn't mind that part much. It's true that I lack the holistic education that trained type designers take for granted.

    Pablo Impallari said:

    One of the reasons there are many similarities in patterns may be that the lowercase alphabet is much more standardized than the uppercase alphabet.
    That is my impression as well. How many distinct proportions that are not «similar» to each other, yet work well as a text font, can there realistically be? Given the number of well-known typefaces, is it even likely that a typeface drawn from scratch should land in an empty spot in the parameter space?

    Actually, doesn't the LeMo principle echo this view? Choose the stem interval, extenders and /o size, and all the proportions fall into place...

    «Platform» is an interesting experiment but pretty much unusable for text, IMHO.

    Once you have you /n then pretty much every other lowercase follows suit [...]
    Well, there is at least one additional dimension in the width ratio of round shapes vs vertical shapes. Garamonds tend to have very large rounds compared to modern-day typefaces. I am also always surprised again when I see the extremely narrow /n of Futura next to its optically-circular /o.

    But let's get back to the OP's project.
  • PabloImpallariPabloImpallari Posts: 527
    edited May 2017
    Platform is used by one of the main telephone companies in argentina for display purposes and it sucks every time, but in all honesty it does convey some sort of we-are-cool hipster feeling, so it does it purpose pretty well. It works as intended, despite the crazy patterns.

    BTW I'm waiting/interested in Frank's views of uppercase patterns standardization.
  • BTW I'm waiting/interested in Frank's views of uppercase patterns standardization.
    Hi Pablo,

    I wrote a bit about this in Appendix 8: Proportions Of Capitals In Roman Type starting on page 309 of this PDF).

    Best, Frank
  • Nathan ZimetNathan Zimet Posts: 61
    edited May 2017
    hey Arne :) I really like the way this typeface is coming out!

    The terminals on the /a, /c, /f, and /g aren't very consistent in size or shape, you could try neatening those up. The top right of the /q is too pointy. Bottom right of /G the curve needs smoothing. center serif on /E is too small. The bottom bowl of the /g you could try a slanted contrast. The upper arm of the /k could stick out to the right more. The terminal on the /J is too small. The bottom left and top right of /Z could be pointier. serifs on /s could be larger, mostly the top one. the curve on the /t doesn't feel square enough to match the rest of the curves.

    Another thing is that some serifs end in points, while others are quite blunt. Something else is that some letters such as the /o are very vertically balanced while the /e leans back and the /s leans forward (this of course will look different in small text). Maybe try making the counter in the /o asymmetric? Bookerly does this and it works very well. If you're looking for other typefaces for reference i'd suggest Janson (which you're probably already looking at, Janson is also called Kis, you might also look at Spinoza ) and Greta Text.

    Could you make a proof with running text or take a picture of a printed sample? The typeface as is, especially the capitals, look like they would work better for text than for display, because the serifs are quite thick. Maybe think about a text and a display version? 

    Found on page 400 of Mr. Bloklands PDF:

  • Second draft: I modified the serifs and tried to get closer to the concept of „superellipsed shapes“ and make it more origin. They are curved now, not so heavy and trianguar. I think asymmetric serifs make no sense anymore, combined with round shapes it looks too confusing to me. (BTW I feel, these curved serifs are more difficult to handle than triangular serifs.)

    More rectangular appearance of /c/e/t/a

    @Johannes: I will not repeat all of your corrections, but followed nearly all and tried to include them into the new design, thank you for that! 

    @Christian: Right with the /g (you had a long experience with the /g at Paramond…?), what do you think now? /C and /G. Exact Spacing later.

    @Nathan: terminals /a/c/f/g/J, what do you think now? /q is new as well, right with the /G/Z/s and especially /t, are you sure /E is too small?

  • Nathan ZimetNathan Zimet Posts: 61
    edited May 2017
    Do you think the serifs could be more ellipse shaped to match the counters more?

    also because I am bad at explaining things with words i've attempted to redrawn your /t and /G to show you what I meant. I think your /G is fine as is now though.

    The /U could be more elliptical. The lower bowl of the /a is much more rectangular than the rest of the letters. I think the counters of the /g are too rectangular now as well.

    Personally, I think in the first draft the sample with less square shapes had much more motion and flow, and would work well in text, while the sample with more square shapes is much more sturdy and would do better for titles. Are you sure you don't want a display and text version of the font?
  • I also liked the less square original version of the typeface better.

    I agree the new /g is too square. Maybe one could also balance out the stroke weights a bit more; the bottom loop looks too light to me.
  • Craig EliasonCraig Eliason Posts: 991
    Note how with the old serifs the overall mass of the head serifs and baseline serifs was roughly even. With the curved ones, the baseline serifs look considerably lighter which seems odd to me.
  • Maybe that was a step too far. I went backt to the first version of /a/e/c.

    @Nathan: Assymetric and more ellipse shaped serifs was a good idea. But I tried to keep them diagonal. I would like to try (and learn) the text version first, a display version maybe later.

    There are two new versions of serifs now: 

    a) serifs like before

    b) assymetric and a little more ellipse shaped

    c) the same but thicker baseline serifs

    Text example:

    @Craig: You were right, the baseline serifs were too light

    @Christian: What do you think about the /g now?

  • Georg SeifertGeorg Seifert Posts: 613
    edited May 2017
    When you play with the shape of the serifs, you might try corner components: https://glyphsapp.com/tutorials/reusing-shapes-corner-components
  • – I know (just visited a great glyphs workshop with Rainer in Vienna)

    It worked perfect, but I had problems to handle serifs on diagonal stems, e.g. /K

    And when I copy and pasted a glyph with corner components, sometimes the corners were disappearing.

    Maybe I´m too stupid for the corners, but working with components is practical as well and nearly the same method.

  • Thicker base serifs work better. Not sure why you want them to be thicker toward the left, though. If anything, I'd emphasize the outgoing side, in accordance with the forward motion of writing and reading.

    The /g has a more even color now, but its eye still strikes me as extremely square, especially compared to the gooey free-form nature of the lower half of /g. Maybe you can make the counter of the lower half a bit more rectangular and that of the upper half more circular.
  • I think emphasizing the left side is better, like Nathan showed. But they have the same height now. 

    maybe the second /g is better:   

  • I made a new draft with some changes, especially the top serifs:

    ABC.pdf 12.7K
  • Georg SeifertGeorg Seifert Posts: 613
    You can draw the serifs like this to prevent this:
  • maybe the second /g is better:   

    Actually, I prefer the more angular loop of the previous /g (left). The smooth loop feels out of character among the other letters. I still think the top loop could use some more rounding, though.  (Take with a grain of salt; this might be purely subjective at this stage.)
  • What do you think about the accents, do they work together? Especially the eth was hard, I don´t fully understand the construction.

    aringacute: I scaled ringcomb as component to 80%. Is that the correct way or do I have to create an alternate like e.g. ringcomb.case or something …?

Sign In or Register to comment.