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

24

Comments

  • Perhaps we should start a new thread on the economics/pricing of variable fonts? Or revive an old one on that topic?
  • @Thomas Phinney I think pricing is relevant to this thread since it may be a factor in the lack of adoption of variable fonts.
  • Mark Simonson said:
    Requiring people to pay for the full family to get the variable version assumes they want all the styles. That may be true sometimes. But there could easily be customers who want the variable version in order to get a handful of specific styles that are not present in the static family. I think this kind of thing will more often be the case.
    Or they are only interested in the weight axis and not the width axis, for example.
  • Kris SowersbyKris Sowersby Posts: 42
    edited December 2021
    I’d argue that your lack of VF requests may also be due to you not really publicly engaging with the tech, as far as I can tell.
    Sure, that’s a valid point. However, I did specifically write about our experience with webfonts as well: we also had zero public engagement with @font-face tech, but were swamped with actual customer demand. 

    This is probably not be the most reliable measure, but it's the 11th most requested feature on their forum, sorted by votes: https://forum.figma.com/t/add-support-for-variable-fonts/325 I don't know how that adds up in the grand scheme, but it's not nothing
    On the other hand, 312 votes from Figma’s 4 million-ish user base is… 0.0078%. Despite that, I remain hopeful!
    If more variable fonts had optical axes, I could actually see this happening. Not because you'd see all 20 styles at once, but because responsive websites that are supposed to be viewed on both 4" and 40" screens of varying resolutions could do a better job of serving the right fonts in the right context.
    I think this is what @Peter Constable was partially referring to in his original question? Automatic, responsive optical styles for websites would be amazing, but realistically within the ability, desire and skillset of a handful of digital typographers. There would also need to be *so* much more testing.

    @Cory Maylett I think you’re on the money with this:
    To summarize, I don't think the scarcity of variable font use in websites is due to one or two main reasons. Instead, it's at least a dozen different smaller reasons that add up.

    Making a variable version of a font family that was not originally designed to be variable is not a trivial or significantly automatable task
    100% agree. Coupled with a general lack of demand… why would a commercial foundry undertake the re-mastering effort of the back catalog?

    I am a bit surprised by the lack of interest in the trend here. If we went from 2% to 10% of web sites using variable fonts between 2019 vs 2020, I think it will be VERY interesting to see the 2021 percentage when those stats are in.

    I didn't realise this was the trend — impressive! It’s hard to know what the driver is though, right? Is it because they're simply a default Google Fonts option now that Marc Foley points out, or is it because the developers are pumped for smaller filesize, or the designers can finally get the weight of Roboto in between Medium and Bold they’ve always desired?

    @Cory Maylett

    Since variable fonts involve designing only the end-of-the-range masters…
    That would be ideal, but is almost never the reality. In short, you can technically go from Thin to Black and generate all the intermediates, but aesthetically you'll likely need a Regular or Medium master to control the middle. It's the same for condensed and expanded styles, moreso for Italics which do weird things across a condensed designspace.

    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? 
    Sure, that could be done. But it would likely introduce a death spiral of technical support problems. Font names, hinting, style-linking, etc etc etc. It's hard enough generating and testing and supporting static instances, let alone letting users freeball it and generating their own.

    But let’s get back the OP.

    @Peter Constable — what was the original desire of the Variable Font working group? Where there expectations of uptake and implemenation of VF? Were there milestones set? Could these perhaps be set against the current use data?

    As a sidenote, I assumed that because VF is essentially a webfont technology, I am surprised to see that none of the main sites of the big 4 seem to use VF:


    Of course they're involved in other way and have had to implement and develop OS's, apps, infrastructure for VF. But still? Not on their websites?
  • It would be very interesting and helpful to hear from people who were at Adobe when commercial Multiple Master fonts were developed and marketed.

  • Florian PircherFlorian Pircher Posts: 171
    edited December 2021
    As a sidenote, I assumed that because VF is essentially a webfont technology, I am surprised to see that none of the main sites of the big 4 seem to use VF:

    These sites use only a single custom font style.
    https://www.adobe.com
    I am not sure if there is a variable version of Adobe Clean.
    There are variable versions of both SF and Segoe. I assume these big sites cannot afford to drop support for old browsers/OS versions, but their fonts would be ready.
  • But let’s get back the OP.

    @Peter Constable — what was the original desire of the Variable Font working group? Where there expectations of uptake and implemenation of VF? Were there milestones set? Could these perhaps be set against the current use data?
    Web fonts and responsive design for the Web were two key things that motivated starting the work. But a broad range of potential use scenarios were anticipated as opportunities. E.g., I remember one person who works on a reading device suggesting they could use it to adjust to user text preferences without needing to download different fonts.

    The working group were predominantly engineers along with some font and font tool developers. There were milestones set for development of the spec, and we anticipated some milestones for getting VF supported in CSS, in browsers, and in OS platforms (modulo not being able to divulge specifics about a company's product release timelines). But there weren't milestones set for broad adoption.
    As a sidenote, I assumed that because VF is essentially a webfont technology, I am surprised to see that none of the main sites of the big 4 seem to use VF:
    The Web was certainly a primary motivator, but I don't think everyone involved was thinking of VF as _essentially_ a Web font technology.

    As for the big 4's sites not using VFs, there are probably a variety of reasons for each company. One general consideration is that some things take time in the face of competing priorities, and in the big scheme of things it's still relatively early. As Simon Daniels said to me recently, type is 500-year-old tech and change doesn't happen quickly.

  • Btw, Jason Pamental has just given a really interesting talk on work he did for the State of Rhode Island to create a CMS system that they could use for state agency Web sites—dozens and dozens—that have used VFs to great advantage.

    https://events.bizzabo.com/356258/agenda/session/687543

  • Nick ShinnNick Shinn Posts: 2,131
    @kris sowersby

    why would a commercial foundry undertake the re-mastering effort of the back catalog?

    FOMO (Fear of Missing Out).
    I have a few designs that have continued to be popular for 20 years, so I don’t want them to fade away because they haven’t been “upgraded”.

    By the same token, there are a couple of my old .ttf faces from back in the day that are rarely licensed, that I have yet to “upgrade” to .otf, but still offer—and the beauty of it is that they still work as good as they ever did! That’s what I like about font software.

    But John Hudson is right; it’s tricky getting the new VF named weights to match the old interpolations—because I tweaked the originals in ways I don’t remember how, to optimize the appearance of each.

    It’s always seemed to me that each weight has its peculiarities that are not optimized by a mere two masters. Example: Futura’s apexes (A, V, W), going from pointed to flat as weight increases. Certainly, an in-between master can address that, but it’s “extra” design work, and you can’t avoid weak zones on the axis.



  • @kris sowersby

    why would a commercial foundry undertake the re-mastering effort of the back catalog?

    It’s always seemed to me that each weight has its peculiarities that are not optimized by a mere two masters. Example: Futura’s apexes (A, V, W), going from pointed to flat as weight increases. Certainly, an in-between master can address that, but it’s “extra” design work, and you can’t avoid weak zones on the axis.

    Or you set the font to change swap from one design to the other (using alternate glyphs) at some point(s) in the designspace. For example, above a certain weight, switch to the flat-apex versions of these glyphs.

    This avoid using intermediate masters, at the cost of having a few extra glyphs.
  • Nick ShinnNick Shinn Posts: 2,131
    Those alternate glyphs are masters.

    The “weak zones” I refer to occur when neither the flat nor the pointed top are particularly appropriate. For Futura, between Regular and Heavy. 
    (I don’t know how Futura Variable handled this.)

    I mentioned another weak zone earlier in this thread, on the weight axis when stroke and counter width are identical. For Frutiger, at weight 67.5.

    Fundamentally, doesn’t full incrementality/slideability have “optical” bugs as well as features?

    **

    The difference may be compared to clocks.
    The clock dial has semantic chunks (“half past…”), but not the digital clock.

    I’m suspicious of the white-bread smoothness of variability.
    No doubt slideability can be designed to, and has many virtues, but the OP asked about slow uptake, and this is a typographic theory as to why.
  • Matthew Butterick has updated his VF article, The Scorpion Express.
  • Ray LarabieRay Larabie Posts: 1,376
    Until the largest founders step up with VF versions of their most used types, I fear that the format will languish.

    Absolutely. Once the staple typefaces have VF versions, demand will go up. A system which would let foundries reverse engineer existing families into VF could accelerate this process. I don't expect a tool where you can feed it a bunch of styles and it burps out a perfectly fucntional VF with no extra work, but specialized software could make it economically feasible to create variable versions of existing typefaces. Currently, for me, transdforming existing typefaces into VF makes no economic sense.

    Even for my typefaces where MM masters exist, creating a VF from them with currently available tools is daunting. My typical MM-to-instances process involves manual tweaking and cleanup of intermediate weights. For example, if the crossbar on the "e" looks correct in UltraLight and Black but too light in medium, I manually tweak it rather than devise a interpolation solution. Since the masters are essentially disposable, it was the most practical way to get the job done. There are countless little edits for intermediate weights. Sometimes if my idea for the heaviest weight didn't fit the interpolation plan I'd create an custom font that's outside of the axis range. Often when I add extra characters such as new monetary symbols to a MM typeface that has already been expanded, it was quicker to use the blend tool and manually clean them up than bother updating the MM. With the ideal, specialized reverse engineering tool, I think it would be viable to turn these familes into VF in a way that makes business sense. With current tools, it would be foolish to attempt.

  • jeremy tribbyjeremy tribby Posts: 212
    edited December 2021
    it's at the very bottom of the article as an afterthought but I think this is interesting:

    Practical Typography uses a variable-font-like trick to serve different grades of the body text font to different platforms. For instance, Windows users get a slightly heavier version than Mac users, to account for lighter screen rasterization. All this can be done easily with today’s technology. But I seem to be the only person on the internet who was sufficiently motivated to try.

    this is true of optical sizes as well - you can use CSS media queries to serve different optical sizes on different sized devices, but it's extremely uncommon (the whitney's website is the only one I can think of where you'll see 3 optical sizes of NHG, but it doesn't adjust along with the screen size). for a few years I heard a lot about "responsive typography" and then it fizzled out. I feel confident opsz on responsive websites is a good use of VF, but I am less confident that there is sufficient motivation, as the author puts it.

    On the other hand, 312 votes from Figma’s 4 million-ish user base is… 0.0078%. Despite that, I remain hopeful!

    ha! to be fair, there are other features with similar vote counts that are already in beta, but I take your point RE: how that demand actually adds up for users vs. internally for figma. for what it's worth, figma has been hiring for this role and describes variable font support as "crucial" to their core product experience. I guess we'll see!
  • Ray Larabie said:

    Even for my typefaces where MM masters exist, creating a VF from them with currently available tools is daunting. My typical MM-to-instances process involves manual tweaking and cleanup of intermediate weights... 
     With the ideal, specialized reverse engineering tool, I think it would be viable to turn these familes into VF in a way that makes business sense. With current tools, it would be foolish to attempt.

    OK, I'm interested in this. I'm potentially interested in building this. Because I don't think it would actually be that difficult. (Bear in mind, I am (a) usually wildly optimistic and (b) not a very good designer.)

    One of my assumptions is that in discussions like this we tend to wonder whether such a tool would do the right thing for the most difficult of cases, but actually maybe 80% of cases are simple. Sure, there are two- or three-axis fonts with lots of manual tweaking going on, but equally there are a lot of single-axis fonts with minimal (if any) tweaking. So, what if we handle the 80% case and accept that complicated fonts are going to require manual handling?

    It should be fairly easy to build a tool which can evaluate what degree of tweaking has gone on: take input of a bunch of styles, examine two or three masters far away from one another, compute the interpolations and see if the interpolated glyphs differ from the instance glyphs, and if so, by how much. I imagine that for the vast majority of glyphs in the vast majority of fonts, they wouldn't differ at all.

    The tool then finds a minimal set of masters which globally minimises the differences, marks up the glyphs which interpolate perfectly and don't need to be tweaked, and then warns you with differing colours in your font editor about glyphs which require checking. It could even create the interpolation layers (bracket/brace layers) for those glyphs for you.

    Would that be the "specialized reverse engineering tool" you're talking about?

  • @Peter Constable — The working group were predominantly engineers along with some font and font tool developers. 

    Therein the heart of the problem. Serious typographers were not part of the group. If they had been, there might have come out of it a consensus—or at least a workable range—of goals that the VF developers could concentrate on to make products that would be indispensable (and, therefore, sale-worthy).

    I can tell you only about the needs of book typographers, especially those who work on complex publications that involve multiple type sizes and weights and may incorporate multiple language scripts. The issues are these:

    • Those of us who use static font families with size-range iterations often find that their relative weights are not harmonious when set on the same page (text, notes). It is often the case that “small text” iterations are chosen to work best at the smallest extreme, say 6 pt., while the reality is that footnotes are seldom set that small, but more often between 7.5–9 point. (I’m generalizing about the specific sizes, knowing that various fonts are smaller or larger on  the body, depending on descenders and such.) Similarly, fonts made for text sizes are averaged to work best at 12 pt, while the reality of many books dictates that the type be set at, say, 10.75–11.25 pt. The result is that the text size is a little too light and the footnote size is a little too dark, and so on with subhead and headline sizes. Hypothetically, this could be addressed in two ways with VF fonts, with the optical slider and the weight slider—just as it was with some MM fonts years ago.
    • The width variable is arguably more important for sans serif fonts than it is for serif text fonts. (Display sizes are another matter.) That said, it would be useful to have the width variable for some text fonts, though the width range needn’t go to extremes. It is, in my opinion, a mistake to think that VF fonts have to have the same ranges of variability for all types. Keeping them within reasonable practical limits could speed development without displeasing anyone.
    • Similarly, the weighting of fonts often fails to achieve what the typographer needs and wants. It is often the case that something between the regular and semi-bold is the right weight for the purpose. VF fonts have the ability to address this to perfection.

    Working typographers are exceedingly sensitive to these kinds of thing. Their experience can help developers of VF fonts have a better sense of their goals. No need to be all things to all people all of the time—and no need to knock your heads against the wall. As someone who is both a typographer and a type designer, it's been sad to see how the type design community has become a thing unto itself, losing sight of its place in the bigger picture—and losing contact with its natural allies.


     

  • @Peter Constable — The working group were predominantly engineers along with some font and font tool developers. 

    Therein the heart of the problem. Serious typographers were not part of the group.

    I described the predominant makeup. I didn't say there were no serious typographers. I don't know how you qualify "serious", but with John Hudson, Rob McKaughan, Laurence Penney, Bianca Berning, Erik van Blokland, Tom Rickner, and probably others I'm forgetting, involved, I think we had a fair level of typography expertise.
    • ... Hypothetically, this could be addressed in two ways with VF fonts, with the optical slider and the weight slider—just as it was with some MM fonts years ago.
    I certainly hoped we'd see optical size make a comeback.

    • It is, in my opinion, a mistake to think that VF fonts have to have the same ranges of variability for all types.
    We never assumed that would be the case.

    • Similarly, the weighting of fonts often fails to achieve what the typographer needs and wants. It is often the case that something between the regular and semi-bold is the right weight for the purpose. VF fonts have the ability to address this to perfection.

    Working typographers are exceedingly sensitive to these kinds of thing. Their experience can help developers of VF fonts have a better sense of their goals. No need to be all things to all people all of the time—and no need to knock your heads against the wall. As someone who is both a typographer and a type designer, it's been sad to see how the type design community has become a thing unto itself, losing sight of its place in the bigger picture—and losing contact with its natural allies. 

    So, you started by saying that not having "serious typographers" in the original working group was "the problem". Yet you seem to be saying that VFs can be helpful to serious typographers, provided type designers are communicating with the typographers that will use the VFs. I don't see a connection to how the original working group came up with the spec for the VF formats. There isn't a working group _now_ that's trying to set who does what with VFs.

    I also think this has drifted away from my original question: Why don't we hear about more use on the Web? Are you saying that type designers have been releasing VFs that aren't what are needed for serious typography on the Web?
  • This is a great idea, I wish I'd thought of it!
    There are several sites where you can create a custom 'static' font from your own variable parameters but you have to possess the variable font to begin with.
    So to enable what you suggest, the variable font needs to sit on the foundry's server and the resulting static font be incorporated in the purchase/download process. Easy ??
    Some day all fonts will be sold like this.
  • Apologies - I should make it clear that my previous post was in response to:
    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? 
    I neglected to read the second page of posts. Obviously quite a hot topic!
  • Thank you, John. You are correct in assuming that I was referring to users of fonts. The terms “typography” and “typographer” have become rather murky in recent years, and I believe that to be unfortunate. “Typography,” according to both the Oxford and Merriam-Webster definitions is “the art or procedure of arranging type or processing data and printing from it" and "the style, arrangement, or appearance of typeset matter." “Type design[er]” and “type technology[ist],” though related terms, refer to separate activities. Peter, it appears to me that the persons you identified as involved with the VF project fall into the latter categories. While I have no doubt that most, if not all, of them engage in typography to one degree or another, my self-reference was to someone whose primary area of expertise is in the design, composition, and production of carefully crafted texts and has had the experience of applying those skills to many thousands of pages over many decades using every technology since the days of cast metal. These are the serious users of type, and I consider myself to be among them. (I am also a type designer with more than thirty years experience, though for the past fifteen years this work has been primarily in Biblical and liturgical Hebrew. The number of volumes in print in which these types are used no approaches 1.6 million.)

    Without communication with typographers, type designers and technologists easily fall into misunderstandings of what typographers want and need. Here’s an example: Many recent Latin fonts have script sets for Cyrillic and Greek. These can be very useful, but on the practical side, the most thoroughly made fonts, with alternate glyphs, swashes, and the like, create very cumbersome glyph palettes that slow production. It’s easier and faster to work with separate fonts for each script—and easier to assign them to style sheets. When the secondary scripts are right-to-left, such as Arabic or Hebrew, the problems can become even greater. Some of the problems could, perhaps, be addressed through the application of Stylistic Sets or locl features, but the type designers and technologists, without extensive experience setting these types and languages, wouldn’t know to think of it. Similarly, none of the four principal ligature features allows for a menu-driven approach (I want these throughout the document, not those), which most serious typographers would like to see because it’s a font-by-font issue and a matter of personal taste, though I realize that this would likely require the participation of the app developers. I could go on and on. You need to talk with us.

    You ask:
    Are you saying that type designers have been releasing VFs that aren't what are needed for serious typography on the Web?
    I can't speak for web designers, as I design for print. But the short answer is "yes" and I refer you to my comments up-thread. The font market is saturated beyond the breaking point and designers are not eager to be burdened with all new fonts in this format, when the ones they wish to see for a particular job are the fonts they have been using. And, as was mentioned by others, they would rather not be forced to purchase complete font families just to get the VF versions. (Mark Simonson got this right.) The principal issue for me and presumably others is that few VF fonts released thus far offer the optical size/weight axes that are most important to me for text.

    To Craig Eliason and Nicholas Garner: 
    That approach (dialed-up iterations) might work for web design, as the determination is apples-to-apples. But for print, it would be apples-to-oranges. Print designers need to test the result in realtime and make adjustments before coming up with the needed iteration. It might take several tries. In historical retrospect, what you suggest is a step backward from Adobe's MM fonts of the 1990s.



  • To Craig Eliason and Nicholas Garner: 
    That approach (dialed-up iterations) might work for web design, as the determination is apples-to-apples. But for print, it would be apples-to-oranges. Print designers need to test the result in realtime and make adjustments before coming up with the needed iteration. It might take several tries. In historical retrospect, what you suggest is a step backward from Adobe's MM fonts of the 1990s.
    To be clear, I was envisioning vending customized instances as a complement to, rather than replacement of, the variable font. So that customers could buy the bottle of milk, or the cow, as their needs and budget dictate. 
    (I take Kris’s points above about the potential problems in naming and style linking though.)
  • ...
    You ask:
    Are you saying that type designers have been releasing VFs that aren't what are needed for serious typography on the Web?
    I can't speak for web designers, as I design for print. But the short answer is "yes" and I refer you to my comments up-thread. The font market is saturated beyond the breaking point and designers are not eager to be burdened with all new fonts in this format, when the ones they wish to see for a particular job are the fonts they have been using. And, as was mentioned by others, they would rather not be forced to purchase complete font families just to get the VF versions. (Mark Simonson got this right.) The principal issue for me and presumably others is that few VF fonts released thus far offer the optical size/weight axes that are most important to me for text.


    Thanks for clarifying.
  • Those alternate glyphs are masters.

    No, they are not. They are glyphs. Just the same as if you did alternate glyphs in a non-variable font.

    The difference is, these alternates are programmed to kick in for specific parts of the design space. They *have* masters—but only in the same positions as all the other glyphs. They do not introduce new masters.

    For example, in Science Gothic, I did three versions of the dollar sign. One is the default, and each of the alternates is active in different portions of the weight/width range. But I still had to design all the same masters.
  • Nick ShinnNick Shinn Posts: 2,131
    No, they are not. They are glyphs.

    Yes, master glyphs. A “master” is something that is used as a fixed rendering to interpolate from.

    For instance, in my Beaufort variable, I have two master glyphs for the Regular weight of A, one with a pointed top, for interpolating with the Thin, and one with a flat top, for interpolating with the Ultra.

    “Master” is a difficult word anyway—gender politics.
  • Interesting: I'm listening to Jessie Chen's ATypI presentation regarding condensed/etc. (literally, outside the box) type design, and a point that came up in discussion is that content designers don't know what would be appropriate uses of condensed type. Sounds like a similar issue to what some have brought up here.
  • I raised this question since Web usage of VFs was on my mind while I was preparing for my ATypI presentation last week, Happy Fifth Anniversary, Variable Fonts. In my talk, I discussed the pace of progress over the past five years, what has been a success, what has not yet been a success, and what are lingering problems.

    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.

    I presented three distinct benefits/use cases that I got from Dave Crossland:
    • Compress
    • Express
    • Finesse
    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.

    In the discussion after my talk, a lot of people seemed to like Dave's three-way breakdown. And Google Fonts folk indicated that they welcome others picking that up.

    So, hear, hear: Compress, Express and Finesse!

  • 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? 

    Universal Sans from Family Type allows to you do exactly that.

    View it here: https://universalsans.com/
  • Given that Universal Serif is in the making, can we deduce that it was a success?
Sign In or Register to comment.