How skeleton based type design could shake up digital type design workflows



  • Maurice MeilleurMaurice Meilleur Posts: 58
    edited March 2016
    People have been writing about this since the 1960s. In all those decades of nobody has created a tool that makes the concept worthwhile.
    Why do you think that is?  Is the concept itself flawed?  Or what?
    I am interested in too.
    as am i. to be blunt, and only slightly exaggerating (i think): virtually always the argument about everything from knuth's metafont to contemporary parametric design and extrapolation approaches boils down to this: it generates letterforms we're not used to looking at. that is not an argument i accept per se from my students and i'm not sure why i should accept it from people i think are smarter than me and know more about type than i do.
  • Nick ShinnNick Shinn Posts: 2,128
    I would be an ideal candidate to work in this manner—were I younger. As it is, I have been meaning to get around to using Glyphs for some time. I am quite content to plod along in FontLab, and devote my energies to the broad culture of type, rather than the hassle of learning how to do what I already know slightly differently, for nebulous creative benefits.

    No doubt there are creative advantages this method has over outlines, and that might interest me, but what are they? It is being hyped as a production tool which does a less than perfect job of imitating other methods, which is not very sexy, IMO. Maybe I could get excited if some latter day David Carson figure were producing crazy shit emergent typography with deconstructed skeleton-based fonts, but how would it differ from, say, work such as Typotheque’s Karloff?

    Perhaps the most authentic use would be in dynamic environments, such as was hinted at by LettError’s Twin fonts, which were designed to morph fonts according to streaming data.
  • Dave CrosslandDave Crossland Posts: 1,389
    On quoate: I haven't found the author. So, to me it sounds like an antient roman quoation. Good job! So, who should mentioned there?
    Sure, its a hackneyed phrase; but in the very specific context of parametric and stroke based font editors, this quote is clearly from Simon Egli's presentation :) So I admit I feel a little annoyed that of the 4 quotes on the page, the quote from the Metapolator presentation is not mentioned, not any links to it :) So I think you should quote Simon Egli, 2014-04-03, Libre Graphics Meeting, Leipzig, Germany, linking to :)

    for nebulous creative benefits.

    No doubt there are creative advantages this method has over outlines, and that might interest me, but what are they? 
    End-user type customization. 
  • It is being hyped as a production tool … Perhaps the most authentic use would be in dynamic environments.

    I think the problem is that type designers don't usually create typefaces with stroke-making tools … [W]hen type designers who are used to working with outlines look at this approach, they see something more restrictive than what they are used to.
    and this is exactly the issue: that we are considering the (re)development of skeleton-based/extrapolative design in terms of how it informs, contributes to, or replaces existing stages of traditional typeface design—that is, a process that as currently understood terminates in the generation of a font, a static and reusable set of instructions for how to draw letterforms. this is even true of the twin cities fonts, if i understand the project correctly: any variation is among different preformed sets of shapes. in other words, as a more extensible cousin to multiple master tech.

    the other way i've seen it considered is as a means of making microtypographic adjustments to letterforms in response to changes in the context of a browser window, such as nick sherman describes in the piece he wrote for a list apart on dynamic hinting. but even there nick is thinking about a process that ends in a static font. and maybe that's why it's so tempting to compare its output to the look of traditional typeface designs.

    but what if the process were considered, not as a design assistant or adjunct, but as a replacement for the font? that is, as a alternative to the way letterforms are currently generated for the screen, something much more like knuth's original idea for metafont? so that in place of static font files, you have a genuinely dynamic process that responds to real-time contextual information—not just micro- or macrotypographic but even semantic information (for example, the age or current turnover of a link, or the tone of the author's voice)—and sends information to a rasterizer without an intermediate stage?
  • Nick ShinnNick Shinn Posts: 2,128
    End-user type customization. 

    Which end-users: Typographers, layout applications or readers?

    And more: 17 font users

  • Ofir ShavitOfir Shavit Posts: 396
    edited March 2016

    @Ofir Shavit I am glad as you that there is somebody else seeing the value in such aproach. We are able to generate well ploted oudline of stroke following a skeleton and currently looking for opportunities to integrate with existing type design tools.
    @Filip Zajac  I wish you the best luck, and looking forward to test your tool.

    @"Mark Simonson" I agree that we face here two different ways of thinking, but there's also a lot of prejudice and antagonism based on non-professional motives, which blocks many type designers from seeing/expolring the potential of such tools/approach, which certainly has limits but great advantages as well.

    @"Nick Shinn" @"Dave Crossland" There are so much "start-user" creative advantages to skeleton type design (at least the FA's approach) I don't even know where to start naming them.

    I'll just point out one fundamental, and for sure controversial, issue: Time.

    Time. Yes, I know what you all think when mentioning time in type-design context, and I agree with you all about it all. Put it aside for a moment, because I'm talking about the new (additional, not replacing) creative advantages real-time synced-skeleton enables, for the type-designer, not the end user.

    First, it is clear that traditional type design process starts by hand, but it only starts there because only when you have your character design digitized you can start examine your ideas in text context, and everyone here knows what it means. I don't know how they've done it in the iron age (can imagine) but fonts today are digital creatures, and digital technology has a lot to offer here, and we're only talking about time at the moment.

    How will constructing and designing type in a fraction of the time and efforts will influence creativity? 

    An explosion of bad typefaces.


    I'm a designer and creative visions strikes me, if I could just project them out of my mind that would be the best design tool for me (craftsmanship will never die, it involves genuine creativity for itself and creativity is god). I could never design type with regular tools because I'm not going to work two weeks just to find out my vision sucks. Reducing the technical barriers (and there are plenty in traditional type-design methods even though most pros has adjusted to them and may not realize its influence on their creativity and creation) frees up more time for design and creative work, clear and sharp. It might take you the same amount of time to complete a typeface from start to end in both methods/tools but you will cross a far greater distance with the faster tool, or just have more time to spend with your family/contribute to the community, which will also have a positive effect on your creativity.

    Now that's just a narrow aspect of the issue. The major thing is how creative can you get? This is an interesting, but very stupid question.
    No. It's a great question with only stupid answers!
    I will try to give one in my next video :)
  • Nick ShinnNick Shinn Posts: 2,128
    edited March 2016
    Time is not a creative benefit, it is a production benefit.
    The amount of time it takes to create a glyph has no bearing on its shape.

    An example of new technology which has enabled new kinds of typeface is the contextual alternates feature, in the way that it makes possible a pseudo-random effect, with no effort by the typographer. Type designers have used this to inform the quality of handwriting fonts, and for distressed effects in display type. Of course, alternate glyphs were available in pre-digital technologies, especially typositor phototype, but the goal was not randomness. 

    Digital kerning was another invention that expanded the type designer’s creative domain, bringing kerning under the control of the type designer—as part of the typeface design—not the typographer (although it could be argued that Linotype had produced kerning logotypes, but these were only for a few “titlecase” combinations).

    I don’t mean to belittle production efficiency, but the question was why hasn’t skeleton-design caught on yet, and my suggestion is because it doesn’t appear to break any new conceptual ground for type designers.

    Consider Wim Crouwel’s CRT designs, or the early bitmapped fonts of Zuzana Licko (before the anti-aliasing of Adobe Type Manager). These fonts addressed technological constraints as creative fields in which novel effects could be developed. Beyond the goal of mimicking existing tools, what is there in skeleton-based technology that opens up new design space? Interpolation already exists, as does “Make Parallel Path” and “Expand Path” (with brush tracing).

  • Ofir ShavitOfir Shavit Posts: 396
    edited March 2016
    @Nick Shinn  I strongly disagree with you.  The speed of work has fundamental influence over the creative process thus the created shapes. It is like the creativity difference between hand drawing and digital outline drawing, it has a completely different right-left brain hemispheres interaction.
    You can see it in the video when I start changing Wes' /c to /e. I would have never get to this shape with both paper or regular outline editing. The creative process of creating it was intuitive and involved with tens of creative actions performed in real time and stimulating each other . The primal factor enabling this was the speed of work.
    Now multiply it 26 times and cross multiply it. Only theoretically you could claim that since the end product is a static 2d shape you could achieve it in any productive manner.

  • Ofir ShavitOfir Shavit Posts: 396
    One more thing, that may answer the question why hasn't skeleton-design caught on yet is that normal “Make Parallel Path” and “Expand Path” (with brush tracing) would be the end of the creative process because after one of these actions you lose control over the outline because it gets "dirty". 
    With FA it is seamless. One Node on the skeleton leads to one Node on the outer outline and one on the inner. 

  • Mark SimonsonMark Simonson Posts: 1,648
    but there's also a lot of prejudice and antagonism based on non-professional motives

    Maybe, but the point I was trying to make was that selling this approach to type designers used to working with outlines is a bit like trying to sell the idea of Lego to sculptors used to the freedom of working with clay. It's easy to see how it could save time and effort, but it's not easy to see how it could be used to create the same kinds of things.
  • Nick ShinnNick Shinn Posts: 2,128
    edited March 2016
    Sorry Ofir, but it’s your argument that is theoretical.

    I’ve mentioned several examples of new design concepts that emerged as the result of new technologies. Please provide examples of new kinds of type design that labor-saving inventions have made possible.

    Time-saving may produce a change in the designer’s qualitative experience of the design process, but the end result is still only quantitative.

    For instance, digital interpolation has produced the situation where new sans serif designs today often have eight or nine weights. But how is that substantially different than Univers, sixty years ago?

    As a type designer and calligrapher, looking at the “C” you’ve drawn, I see only a poor parsing that misses the subtleties of the original pen-work.

    Frankly, this is not the best way to promote new tools. Certainly, new technologies do eventually succeed in mimicking, and then perhaps eclipsing the qualities of old, but the real excitement is in providing artists and designers with new expressive possibilities. 

    FontStruct, a rather crude application, nonetheless managed to do that.

    Show me a different kind of typeface that your software can make possible (or find someone who can), that is the challenge, and if you come up with something, it will go a long way to overcoming the perception Ray mentions, that skeleton-based type design is for “neon” styles.
  • Wei HuangWei Huang Posts: 97
    @Wei Huang’s Work Sans is at least partially built with Metapolator
    A correction: it was prepared to be finished in Metapolator, so it was not made with Metapolator.
  • Ofir ShavitOfir Shavit Posts: 396
    Maybe, but the point I was trying to make was that selling this approach to type designers used to working with outlines is a bit like trying to sell the idea of Lego to sculptors used to the freedom of working with clay.
    Just a little bit, very little. The difference in the outcome is much less significant (and doesn't demand compromising because it is just a short-cut to the same "destination")  while the benefits of productivity are huge. as shown in Ray's /a demo.  <-- @Nick shinn  

  • Nick ShinnNick Shinn Posts: 2,128
    Ofir, I’m not talking about productivity.
    What if I don’t want to get to the same destination faster, but somewhere else, or somewhere I’ve never been before?

    I’m trying to help you market your product in new ways, but you just keep arguing an idea I’m not debating. 
  • Ofir ShavitOfir Shavit Posts: 396
    Les Paul attached a couple of “wings” to his solid-body guitar concept, and played it in a band. People thought he sounded great
    I know Nick, and appreciate it!
    I've been thinking (and experimenting) about it since the first time you've suggested it, but I think it is just not meant for opening new fields for type design. As I say FA is a design tool and it's product is a standard static font (it may be possible to take our technology and try to do other things with it but we're not there), and type is type.
    I see it as a big step forward in improving type design tools and methodology, and I only argue because I do think this alone can and will have a significant influence over the designers creativity... inside the frame of conventional, up to date, type technology.

    Many small steps forward in type design tools have been welcomed and adopted warmly, and here we have a big one and I feel like I have to convince people even to check it out.
  • Nick ShinnNick Shinn Posts: 2,128
    OK, I’ll give it a try.
    I’m starting on a new project, probably in May, and will use FA from the beginning.
    I’ll be doing only one or two weights, but in a wide variety of styles—sans, serif, sketchy, polished, etc.
    Maybe I can use the same basic skeleton for all the different styles, each one considered a “package” of attributes.
    Perhaps I can then interpolate between the different packages, to see what the mash-ups look like, and quickly assess and fine-tune the optimal intersections of the different design axes.
  • Ofir ShavitOfir Shavit Posts: 396
    edited March 2016
    I'll be more than happy to personally guide you (and anyone else, anytime) with FA's tools. It's not like anything familiar. (But brutally simple and easy to work with).
    Looking forward to it!
  • Ray LarabieRay Larabie Posts: 1,374
    edited March 2016
    I can see it potential as a production tool. I doubt that I could save time by designing the Latin alphabet skeletally. Maybe new designers would benefit from it...I really don't know. If this type of system could be crossbred  with conventional type software, maybe as a plugin, it would take away some of my endless symbol drudgery.

    For mathematical symbols I usually employ a uniform stroke width that I call the maths stroke. Dealing with heavy maths Unicode ranges is tedious. Right now, I can keep them monolinear and crudely expand strokes, but I still need to deal with cleanup and interpolation madness which rarely goes smoothly. It quicker for me to make shapes and use regular composites and disposable composites.

    If I could keep these skeletal all through the interpolation process and have them expand when instances are generated, it would save me loads of time. See also: miscellaneous technical, box drawing, geometric shapes, misc. symbols...pretty much anything from 2100 to 2BFF.
  • Ofir ShavitOfir Shavit Posts: 396
    These guys?
  • Ray LarabieRay Larabie Posts: 1,374
    These guys?

    Sure, I could do most of these with maths strokes.

  • Ofir ShavitOfir Shavit Posts: 396
    A muscle skeleton can help here, but the glyphs sync (SX) system less, and I want to give a better understanding of it. I'm working on a new set of videos, going through the very basics - characters width, spacing and optical corrections. It will be a very good exercise for noobs, and will give a taste of the new capabilities I'm talking about. 
  • Typographically, skeletons are dead. Dipping them in paint just make it creepier. Design what we read: notan.

  • Ofir ShavitOfir Shavit Posts: 396
    The only thing dead in many cases is just vision and inspiration, this is one of them.
  • Simon CozensSimon Cozens Posts: 722
    I can see skeletons working for Kanji, more as a way of managing complexity and enormity than anything else.
  • Everything has a place in type design. Yes, even strokes. Just as death has a place in life. My talk at ATypI 2009 in Mexico City was about not shunning strokes, but putting them in their place. As opposed to the pride of place they've unduly enjoyed historically, and sadly increasingly today.

    Strokes to manage complexity? Nothing manages the complexity of type better than not doing it.
  • Ofir ShavitOfir Shavit Posts: 396
    Stroke and Skeleton are not necessarily the same. 
    ... and sadly increasingly today.

    Are you also sad about the birth of Sans serifs?
  • I'm sad that we have yet to properly define what a serif is (because we've been too busy with how it's been painted as opposed to what it actually needs to do).
  • Without wanting to pronounce on the question today of the pros and cons of stroke-based vs. notan-based design approaches, and not wanting to drag the thread too far astray, can Hrant or anyone answer me this: Why talk about strokes as if their prominence in the way we think about type design is arbitrary—or some kind of historical mistake, like a geocentric universe or Lamarckian evolution?

    The affordances of a writing system where I'm dragging any kind of stylus across any kind of substrate to make a mark suggest to me I think of strokes first, and the spaces inside and between/around what I make with the strokes second, because that's the essence of what I'm doing. It's not arbitrary or accidental that we think of it that way, or teach it that way. And insofar as type grows from writing, it's not arbitrary or accidental that we think of type that way. 
Sign In or Register to comment.