Why don't we hear about more use of variable fonts on the Web?

13

Comments

  • Nick ShinnNick Shinn Posts: 1,867
    If the Universal Sans system offered rectangular tittles, could it make Univers?
  • Dave CrosslandDave Crossland Posts: 1,178
    edited December 2021
    Compress, express, finesse, was my reformulation in 2018 or so, of David Berlow's "go fish, go figure," which is how he explained in his elliptical poetic style the concepts behind Decovar and Amstelvar Alpha, back in late 2016.

    "Go Fish" is the name of a card game, where you draw a card and don't know what you are going to get; the "random style" button of Axis Praxis is a bit like this. But more colloqually, it refers to the phase "fishing around", where you are searching for a prize hidden in murky waters, which is like sliding sliders around to find a style within a variable typeface family that you feel resonates with the 

    "Go figure" is a American English phrase short for "go figure it out," which is to say "go do some arithmetic". It means, instead of setting axes values directly with a slider, you set some heuristic rules, conditionals based on document contexts outside the font, which in turn determine the axes values via some arithmetical calculation.

    An obvious yet cursed example is the optical size axis, where rarely should users set the opsz axis value manually, but rather, set the font-size property, and use that to set the opsz value. (Cursed because while the OpenType spec says the font size should be converted to real physical points and the opsz axis values are such points, all browsers instead resolve the font size to 'css pixels' and set the opsz axis value to the value of px.)

    Other common examples are to change axes values like weight or grade in response to dark/normal mode switching; or to change width based on text length or container size changes, like a button label switching from English to German (as German words tend to be longer.)

    (The Compress benefit he left out, I suppose, because its not very interesting for typography; I mean, in terms of enabling new typography, it is only enabling new typography for designers who were previously told by developers not to use more than X style due to a file-size/bandwidth budget.)

    I've lectured about these in more detail in the last few years, eg


  • Dave Crossland:

    rarely should users set the opsz axis value manually, 

    No, that’s a decision for typographers to make.
    How many users are typographers? :P

    On the one hand, we can say ALL users are typographers, even in plain text environments, since they make decisions about the visual design of text.

    On the other, we can say FEW users are typographers, as few will self-identify as that label. 

    I say that based on how various customers of mine have “misused” the optical sizes I’ve provided in typefaces, and yet to excellent effect.

    Sure, using size specific designs at sizes they weren't intended for can be used to excellent effect, I'm not saying that isn't so. 

    But the vast majority of users aren't able to achieve those artful effective uses, but they are happy when text automatically looks and works better.

  • Nick ShinnNick Shinn Posts: 1,867
    edited December 2021
    OK, but at least users/typographers should be able to easily opt out of the default and do manual adjustments. So often, as Saurabh Sharma notes, user-controlled typographic nuance is missing from the UI of mass market layout apps.

  • Dave Crossland writes:
    How many users are typographers? :P
    On the one hand, we can say ALL users are typographers, even in plain text environments, since they make decisions about the visual design of text.
    On the other, we can say FEW users are typographers, as few will self-identify as that label. 

    I am one of those few and I would suggest that you check a dictionary for the meaning of the term. To whatever extent your observations are true, it still seems rather cynical to me and it begs the question, For whom are type designers making fonts? I’ve been nagged by that question throughout this thread. I suppose it makes perfectly good sense for Google to pitch their fonts at the lowest common denominator, and to think only about web users; yet it makes equally good sense for others, especially those selling their fonts, to operate at a higher esthetic level. But vis-a-vis VF fonts, is Google’s involvement a case of the tail wagging the dog? Or has everyone else become the tail?

    Here’s an argument against the universal automation of size-specific fonts. One person’s judgment of what constitutes a proper regular weight at a certain size will not be the same as someone else’s—at least to those Dave segregates as the FEW. This was clear with MM fonts years ago, but with those fonts, the user made all of the final decisions. There is also the matter of weights intended for the screen not being at all the same as those intended for paper, where the decision is further influenced by printing process and substrate.

    But the vast majority of users aren't able to achieve those artful effective uses, but they are happy when text automatically looks and works better.

    I know, I know; for many a little better (arguably) is good enough. But if that’s true, then it’s likely that type size is not their only problem.

  • Dave CrosslandDave Crossland Posts: 1,178
    edited December 2021
    Nick, indeed, the lack of typographic controls was something that drove me to be dedicated to libre software for graphic design. If you have a libre license to corresponding source code you always have control ;)

    And I agree that all users should be able to easily opt out of each "finesse" heuristic run on their text, just the same as they ought to be able to opt out of default line spacing, length, tracking, size, and even family - but at the same time, having good defaults for these things is important, and even the most advanced users are sometimes in a rush and benefit when things Just Work.

    Scott, please don't confuse my own personal opinion with the positions of my employer! :) I am a strong advocate for @Laurence Penney 's proposal to make a ratio override, as you describe, part of CSS. https://github.com/w3c/csswg-drafts/issues/4430 But I don't believe the default for the optical size axis should be "off", not at all. 
  • To prepare, I spoke with a number of different people involved in type to get their perspectives, and one of the biggest recurring themes was that it seems more education is needed as to how variable fonts can be useful. But in discussing that further with people, a different thought came out, that perhaps we haven't yet clearly understood a set of distinct benefits and use cases, and what would be needed to succeed in each case.
    You’ve had about 50 replies in this thread from different perspectives about why we’re not hearing much about VF on the web. I don’t think anyone has said “the problem is more education”. That’s fine to say, but who is expected to deliver this education? Who will be receiving this education? I kinda get the feeling type designers/foundries are expected to educate our customers, but please correct me if I’m wrong!
    In this discussion, file size benefits have been mentioned—that's compress. There's also been mention of "novelty" fonts; those fall within the express use cases. And there was also discussion of careful typography using, e.g., optical sizing—that's finesse.
    If we were to perhaps give a percentage weighting to the compress/express/finesse benefit trifecta, how do we think it would land? Based on @Dave Crossland’s comment here:
    If you're self-hosting a few static fonts on a little zero-million-views blog, that compress benefit won't matter, but when you're serving the world's most used typefaces, it all adds up
    And historical, pre-VF font sales and observation, would it be something like:
    • Compress 92%
    • Express 5%
    • Finesse 3%
    And begs the question… is the main benificiary actually Google Fonts and others who serve fonts at scale?

  • Here’s an argument against the universal automation of size-specific fonts. One person’s judgment of what constitutes a proper regular weight at a certain size will not be the same as someone else’s—at least to those Dave segregates as the FEW. This was clear with MM fonts years ago, but with those fonts, the user made all of the final decisions. There is also the matter of weights intended for the screen not being at all the same as those intended for paper, where the decision is further influenced by printing process and substrate.


    There is no “universal automation” being suggested: just a change in the default to make it based on pt/px size rather than fixed at the font default. Nobody has suggested saying an author’s ability to choose optical sizes should be taken away. But there is always a default.

    It can be whatever is set by the font as the default (typically 12), it can be the current pt/px size, or it can be some other value set by the app, subject to the range limitations of the font.

    Do you honestly believe that having the default be “use 12 px/pt opsz at all sizes, unless the author specifies otherwise” is preferable to “use opsz = px/pt size at all font sizes, unless the author specifies otherwise?”

    Dave Crossland said: But the vast majority of users aren't able to achieve those artful effective uses, but they are happy when text automatically looks and works better.

    I know, I know; for many a little better (arguably) is good enough. But if that’s true, then it’s likely that type size is not their only problem.

    Nobody said it was “good enough”—just that better is… better. Why not have a more intelligent default?

    Why couldn’t that argument could be used against ever making any improvement to almost anything, anywhere?
  • @Kris Sowersby

    I said that a recurring theme in discussions I had with various people was "it seems we need more education", but I also said that wasn't all that was brought out in those conversations. Sure, there were some who said (more or less), "It seems there needs to be more education, but it's not clear to whom or how." But in other discussions, what came out was, "It seems that way, but that may be missing a crucial point."

    Also note: the topic in those discussions wasn't specifically the question I raised here, "Why don't we hear about more use on the Web?" Rather it was one point in a more general discussion of what is missing before we could say that variable fonts have been a complete success.

    So...

    That’s fine to say, but who is expected to deliver this education? Who will be receiving this education? I kinda get the feeling type designers/foundries are expected to educate our customers, but please correct me if I’m wrong!
    That get's at what I was saying arose from my discussions with people: It's probably wrong to say that better education of customers (designers, end users) about what variable fonts are and how to use them will lead to broad success and adoption of / benefits from VFs because it assumes that a large portion of customers (designers/end users) can be trained to be skilled typographers.

    There will certainly be need to educate _some_ customers, but that's only reasonable for a segment of customers that already have some skill in typography. But for the wide swath of users that aren't skilled typographers, realizing benefits of VFs for providing better typography will require getting more typographic sophistication built into software to provide better default behaviours. (And, as Dave is saying, the skilled typographers should always be able to override defaults.)

    That means more typographic sophistication leveraging VFs built into Word, Powerpoint, InDesign... and into Wordpress or other CMSes.

    If we were to perhaps give a percentage weighting to the compress/express/finesse benefit trifecta, how do we think it would land? 
    Depends: Are you asking about what has been realized so far, at the five-year anniversary, or what we might hope to see in the (hopefully not too far out) future?

    And begs the question… is the main benificiary actually Google Fonts and others who serve fonts at scale?
    Maybe benefits to date have been realized mostly for a smaller number of large-scale products/services. But certainly that's not what I think most of those who were involved in launching VFs five years ago were aiming for.
    • Compress 92%
    Just want to point out that the idea of file size benefits of variable fonts is true only sometimes, since many (most?) projects only use a couple of weights and many variable fonts are bigger than the size of those two files.
  • Ray LarabieRay Larabie Posts: 1,210
    edited December 2021
    I don't know if educating the public is what's holding VF back. Perhaps optical size might require an explanation*, but most of the other common axes are already familiar to anyone who's purchased or browsed fonts before. You get the same weight and width variations but it's in one font...not hard to comprehend once you twiddle the sliders. What you don't see is VF being sold by major vendors alongside elementary fonts for recognizable typeface brands. If customers could see VF versions of Neue Helvetica, Proxima Nova, Avenir and Univers alongside the elementary fonts and could fiddle with the sliders, they could weigh the price options. As it is now, VF is a thing you need to seek out, so it appeals only to the adventurous or knowledgeable. Variable fonts are invisible. Look at Myfonts.com and Fonts.com and see if you can find a VF. Dafont, 1001fonts, Fontspring, Creative Market...none of the high-traffic sites feature variable fonts.

    I know this thread is specially about variable fonts on the web but I imagine many people who are shopping for fonts to use on the web, check out the high traffic font sites to see what's available. They could check every popular site and not encounter a single VF.

    * I've seen lots of VF demo sites where the user can adjust the optical slider, but I think to get the point across, the text needs to scale up and down when the optical slider is adjusted. I think that sort of literal demonstration of the optical axis would transcend language barriers. Even if you don't know the target monitor's size, the customer would be able to feel how the effect works.
  • I’m grateful to Dave Crossland, especially for the link to Laurence Penney’s proposal, and Thomas Phinney for the enlightenment about how the defaults will be applied. That wasn’t clear to me in some of the previous posts.

    My experience since the late 1980s, when “desktop publishing” became ubiquitous, has been that putting sophisticated design tools into the hands of the many has increased the appreciation and value of work that reflects a command of more sophisticated craft. However, it also demanded a refocus of how one marketed one’s services and to whom. For some, though not all, the advent of OpenType created more profitable niches and ways to demonstrate the value of higher and more particular skills.

    But not all designers see things that way and I’ve often found that, even among some of my most esteemed and sophisticated colleagues, whose work is English-language only, there is considerable resistance to embracing new fonts and formats, even when they are demonstrably superior. Those whose needs do not go beyond the standard ASCII glyphs ask, “What’s in it for me?” or say that they already have a body of work in a certain version of a font and don’t want the hassle of changing. They say this even when they recognize the superiority of the new font. That’s what one is up against marketing new, more fully featured fonts. A good part of the problem is the hegemony of the English language. Perhaps pulling the plug on Type 1 fonts will prove to be a boon to font sales and the embrace of new formats, including VF. 

  • I think that these people with huge libraries of Type 1 fonts are a fairly small chunk of the market, however.

    Moreover, the migration away from Type 1 would only help variable font sales if the old Type 1 fonts in question had variable-font equivalents. I think that is a pretty small minority of Type 1 fonts out there. With big foundries from the Type 1 era like Monotype (and its sub brands) and Adobe not offering many of their classic fonts as variable fonts, it seems there is no such route available.
  • Would it work to have a foundry or reseller site on which the customer could dial up the exact setting they wanted and then download a corresponding static font? Would foundries see such a download as fairly priced if it were at the same price point as current static single fonts? 
    Hey, we are just building that. https://blueshell.xyz
  • jeremy tribbyjeremy tribby Posts: 113
    edited December 2021
    I have a bit of a meta-question about this conversation - why is it framed around the web - because of the importance of "compress"? isn't "compress" on the web a bit of a lost cause, or at least a cause that will never re-gain the ground it had before cache partitioning was widely implemented? if I understand it right, it used to be the case that if I visited one site that uses Open Sans, I'd download it once, and I wouldn't need to download it again on another site pointing to the same binary on the same CDN, and that, for (unrelated-to-fonts) tracking and privacy concerns, this is no longer the case as of about a year ago.
    on express and finesse, I see the fullest realization of variable fonts (IMO) so far in the hands of typographers, and off the web, specifically the font modulator that Dinamo made with Collins for the SF Symphony. you can accomplish this without VF, but it never would have happened that way. still hoping for After Effects support. Rob Stenson's Coldtype library is an endeavor that shows a lot of the promise of VF in motion.
  • @jeremy tribby Thanks for pointing out the examples of motion typography by Collins/Dinamo. The link in your post appears to have an empty URL, though, so adding here:

    COLLINS and Dinamo make the classical contemporary in their responsive identity for SF Symphony — The Brand Identity (the-brandidentity.com)
  • ah, sorry, not sure what happened there. it's this tool in particular — assistive to the designer — that I find very compelling
  • Tell us more about that tool. You've linked to a .gif showing things it can do. I see Collins and Dinamo names on it. Is this a private tool they made for the SFSO project? Something available to others?
  • Tell us more about that tool. You've linked to a .gif showing things it can do. I see Collins and Dinamo names on it. Is this a private tool they made for the SFSO project? Something available to others?

    I am not privy to the details / assume it is proprietary. Dinamo mentioned creating it in their newsletter about half way down the page here:
    based on the UI shown, I would guess it parametrizes the generation of static images

    there's also a lovely web demo with sound as an input that they put together with the font, which I think was just for fun: https://symphosizer.wearecollins.com/

    (Tangentially, the Coldtype Python library I mentioned in another comment has some great potential for sound as well: this video was produced with this code)
  • https://sick.agency 

    Slant axis in use on the web (scroll around)

  • https://sick.agency 

    Slant axis in use on the web (scroll around)

    It might be worth noting that the type is actually being skewed via CSS, and is not using a slant axis or a variable font.
  • https://sick.agency 

    Slant axis in use on the web (scroll around)

    It might be worth noting that the type is actually being skewed via CSS, and is not using a slant axis or a variable font.
    Hahahah woops! :) Thanks for pointing this out!
  • Marc OxborrowMarc Oxborrow Posts: 213
    edited December 2021
    Speaking of education/marketing/evangelizing, Monotype is partnering with the creative director of a prominent design firm to talk up the benefits of variable fonts for brands.
  • Speaking of education/marketing/evangelizing, Monotype is partnering with the creative director of a prominent design firm to talk up the benefits of variable fonts for brands.
    Interesting that they disallow registrations from @gmail.com addresses, with a “business emails only” message!
  • They allow registration using an msn.com address. 

    I'll be curious to see what they have to say.
Sign In or Register to comment.