Which is the most effective naming convention for optical sizes?

I have thought about this for a long time, but I have not reached a solution that fully satsifies me. So, as I am in the process of making final decisions with font naming, I wished to ask for opinions, especially to the more experts on the technical and production sides.

The fields to consider:

Family name / Style name / Full name / Postscript / Style group

So far, the two prominent naming conventions that seems to be in use are as follows.
I will use "Bold" as an example for the weight.

1a. San Francisco (Apple)

Here, since the intended use for the typeface is for screen and print, the family manages with two optical sizes, respectively named "Text" and "Display". 

This convention does not satisfy me because, while one can consider the relative perception of digital fonts which can be used onscreen (but also the use we can now make directly on printed billboards, large posters, etc.), I would prefer the names to be immediately recognizable in their optical hierarchy, be it numerical (12, 24, 36 etc.) or descriptive (Small, Medium, Large, etc.).

Plus, it does not allow much optical sizes. If I wish to have, say, two caption sizes (5pt, 6pt), three text sizes (8pt, 10pt, 12pt) and so on, I don’t know how I could name them.

San Francisco Text / Bold / San Francisco Text Bold / SanFranciscoText-Bold / San Francisco Text


1b. Adobe Jenson Pro optical (Adobe)

A variation of the above criteria. This kind of naming convention, widely adopted has been the earliest used, by Adobe and then many others (using intermediate categories like "Subhead", or further expansions like "Caption" or "Micro"). Here the main difference lies in the fact that the optical size is specified within the style name, and not in the family name.

As above it does not satisfy me much as it does not allow for precise optical sizes. If I wish to have, say, two caption sizes (5pt, 6pt), three text sizes (8pt, 10pt, 12pt) and so on, I also don’t know how I could name them.

Adobe Jenson Pro / Bold Caption / Adobe Jenson Pro Bold Caption / AJensonPro-BoldCapt / Adobe Jenson Pro Capt


2. Paperback (House Industries)

The family follows the traditional numerical value (in points) scheme, which includes the actual intended optical size for printing in the Family name. I think this is also the convention used at the time for ITC Bodoni, and more recently by Synthview Operetta.

I like this solution, but not the fact that this way each optical series ends up being a family of its own.

Paperback 12 / Bold / Paperback 12 Bold / Paperback12-Bold / Paperback 12 Roman


Given the two examples, an idea which occurred to me, and that could be immediately recognizable by non-designers and non-professionals, would be to use the size codes used in clothing, like:

extra extra small (XXS)
extra small (XS)
small (S)
medium (M)
large (L)
extra large (XL)
extra extra large (XXL)
extra extra extra large (3XL)

and so on.


These could be conveniently grouped to represent a caption range, a text range, subhead and a display ranges (e.g. XXS to S and so on).

So I was thinking of something like:

De Vinne M / Bold / De Vinne M Bold / DeVinne-M-Bold / De Vinne M Roman
De Vinne XL / Bold / De Vinne XL Bold / DeVinne-XL-Bold / De Vinne XL Roman

or, more descriptively, using both the more "universal" cloth size coding and the respective point size for print:

De Vinne M 12 / Bold / De Vinne M 12 Bold / DeVinne-M12-Bold / De Vinne M 12 Roman
De Vinne XL 60 / Bold / De Vinne XL 60 Bold / DeVinne-XL60-Bold / De Vinne XL 60 Roman

Also, could something like this work?

De Vinne / Bold / De Vinne 12 Bold / DeVinne-12-Bold / 12Bold
De Vinne / Bold / De Vinne 60 Bold / DeVinne-60-Bold / 60Bold

I am still very undecided.


Thanks everyone in advance for your expertise and opinions! :)

«1

Comments

  • Brook / Stream / Tributary / River / Delta / Lake / Sea / Ocean
  • Alex Visi
    Alex Visi Posts: 185
    edited May 2021
    De Vinne XL 60 Bold” scheme seems pretty confusing because in this case people would expect numbers and letters to have different meanings.

    If you really have two separate fonts exactly for 10 and 12 pt, calling it just that seems the only way. But otherwise I personally feel like giving the exact point size values can make people think it’s wrong to use them for other sizes, even if it works well.

    Don’t see anything wrong with clothes sizes and descriptive names like Display, Text, Subtitle, Title etc. Clear, 
    not limiting, looks and reads nice, easier to remember than numbers or letters (“it’s Subtitle” vs “was it 5 or 4? maybe try 6”). If you have more options than you can describe with words, that might be unnecessarily hard to market.
  • Hi Alex,

    yes, but it’s not a matter of «having more options than you can describe with words», rather to have names for the specific number of optical sizes, for use in print. And yes, when they are thought chiefly for printed sizes, I do wish to make people realize it’s not ideal to use them for other sizes, as their features are for the intended sizes (especially optical corrections and inktraps), but no one prevents other uses, as even in phototypesetting days one could want to enlarge a small print caption sized text to achieve a graphic effect.

    But considered I might conceive optical sizes not entirely dependent on a printed use logic, I was trying to find a satisfying solution which would illustrate that in this case the size indication has an internal logic which is relative to the overall dimension/environment of use (not just the printed page).

    I would favor the numeric point sizes 
    for historical revivals, as that’s what my optical masters, based on original sizes, are addressing but since one can also use them on screen I was wondering whether relative naming could work, but I do not find satisfying generic names, even if they sound "friendly", when they do not allow me to have enough precision.
    If I have three text-range size masters, for 8, 10 and 12pt (maybe even 14pt), aside from the numeric values, how would you call them?

  • Thomas Phinney
    Thomas Phinney Posts: 2,889
    If you want to have that many optical sizes, and you feel a need to have labels, numeric labels make sense.

    In a current project, for masters we are using S(mall) M(edium) L(arge) XL (ExtraLarge), but for instances, numeric designations such as 24.
  • If you want to have that many optical sizes, and you feel a need to have labels, numeric labels make sense.

    In a current project, for masters we are using S(mall) M(edium) L(arge) XL (ExtraLarge), but for instances, numeric designations such as 24.
    Thanks Thomas, I was also looking at Science Gothic, but I seem to get it does not have optical sizes, just that added axis for "contrast style", right?
    It definitely helps to speak with someone so familiar with Fontlab, as I am a bit confused by the apparent complexity of having Axes, Masters and Instances and I was wondering where it’s important to keep the names consistent (as I would not want to deal with it afterwards).

    But if you avoid numeric or code labels, how would you proceed with a number of optical masters (as traditionally in lead, say at least a selection between 6 and 72 point size)?
  • Craig Eliason
    Craig Eliason Posts: 1,437
    This is an interesting question, one that I've been wondering about myself. I would only add as optical size becomes more often a variable font axis, a "spectrum" solution rather than discrete "sizes" might become more attractive. 
  • James Puckett
    James Puckett Posts: 1,995
    I’ve used small, medium, large, huge for optical sizes. Nobody complains. Although “Bartledes Medium Medium” seems like a silly name.
  • Thomas Phinney
    Thomas Phinney Posts: 2,889
    edited May 2021
    Thanks Thomas, I was also looking at Science Gothic, but I seem to get it does not have optical sizes, just that added axis for "contrast style", right?
    That’s right. The optical size work is on a different project. I considered optical size for Science Gothic, but that would have been five axes and my head would have exploded.
    It definitely helps to speak with someone so familiar with Fontlab, as I am a bit confused by the apparent complexity of having Axes, Masters and Instances and I was wondering where it’s important to keep the names consistent (as I would not want to deal with it afterwards).

    That terminology and those distinctions are inherent in variable fonts—not in any way FontLab-specific. (Those terms go back to GX and multiple masters, even.)
    But if you avoid numeric or code labels, how would you proceed with a number of optical masters (as traditionally in lead, say at least a selection between 6 and 72 point size)?
    Ahhhh. You may be confusing masters and instances here. In variable fonts:
    • an AXIS is a range of variation, such as width or weight or optical size.
    • a MASTER is a set of glyphs at a single point in the design space, for which you have actually specified the positions of the points of the glyphs. Each axis must be defined by at least two masters, although a master may have a role in more than one axis.
    • an INSTANCE corresponds to a font in the old non-variable digital tech. It is just your specification of some particular coordinates and a label for those coordinates.

    So for example, I might make a simple sans serif variable font with just a weight axis. It has three masters: Thin, Regular and ExtraBlack. But it has nine defined instances corresponding to not just those three weights, but a bunch in between (all the standard CSS weight stops, as it happens).

    I could have done that same variable font with just two masters, Thin and ExtraBlack. But for this imaginary design, I wanted to have more control and fine tune the regular, allowing the Regular to be lower contrast, compared to what I would have achieved with just two masters at Thin and ExtraBlack.
  • Alex Visi
    Alex Visi Posts: 185
    Claudio Piccinini said:

    If I have three text-range size masters, for 8, 10 and 12pt (maybe even 14pt), aside from the numeric values, how would you call them?
    Reminds me of the traditional naming system. Perhaps, I would expect something like Text S, Text M, Text L / Text Small, Text, Text Large to be precise enough to refer to typical text sizes about 8–14 points. But for any other sizes still, who knows how big Headline S, Headline M, Headline L are?
  • When I drafted a family concept with optical sizes about 20 years ago, it was natural for me to go for the traditional point size names. They are unique and smart, and can’t get mixed with some other scale (“medium”!). – I ended up naming my cuts Petit, Corpus, Tertia and Titula. Of which the first three refer to 8p, 10p and 16p sizes, just as an orientation.
  • S/M/L seems like the most logical since most people understands them. Also I like Display and Text, although it only allows a few sizes.
  • The user and all related content has been deleted.
  • John Savard
    John Savard Posts: 1,131
    Most of the typefaces I'm aware of that have optical sizes include the numeric size in points in the typeface name. Unless the typeface isn't intended to be used for printing on paper, I would think that this is the simplest, least confusing, and most informative method to indicate the optical size of a font.
  • Thanks much everyone,
    more or less, numeric values indeed seem the best solution for me.

    @Thomas Phinney, can I ask you if such a solution

    De Vinne / Bold / De Vinne 12 Bold / DeVinne-12-Bold / 12Bold
    De Vinne / Bold / De Vinne 60 Bold / DeVinne-60-Bold / 60Bold

    can work on the technical side? I mean specifying both the optical size and the weight in the second field (instead of the family name). Or is it better to call the sets "De Vinne 12", "De Vinne 60" etc.?
  • I use from small to large: Caption, Text, Deck, Display. But then again I come from a magazine background.
    That is perfectly fine, but as I said the only problem I have with descriptive range names is that in my case I could have more than a optical master for Caption or Text etc.
  • This is an interesting question, one that I've been wondering about myself. I would only add as optical size becomes more often a variable font axis, a "spectrum" solution rather than discrete "sizes" might become more attractive. 
    It all depends if you find more convenient discreet, also specifically designed styles rather than the solution provided by interpolation. They are two different things, and even just aesthetically they produce different results.
  • Thomas Phinney
    Thomas Phinney Posts: 2,889
    Thanks much everyone,
    more or less, numeric values indeed seem the best solution for me.

    @Thomas Phinney, can I ask you if such a solution

    De Vinne / Bold / De Vinne 12 Bold / DeVinne-12-Bold / 12Bold
    De Vinne / Bold / De Vinne 60 Bold / DeVinne-60-Bold / 60Bold

    can work on the technical side? I mean specifying both the optical size and the weight in the second field (instead of the family name). Or is it better to call the sets "De Vinne 12", "De Vinne 60" etc.?
    I am not clear on what “the second field” is, in terms of internal font data or your display here. Seems like you are specifying both optical size and weight in the third, fourth and fifth fields here. But without knowing what exactly those fields are, it is hard for me to advise.

    That said, I am guessing your fourth field (DeVinne-12-Bold) is the PostScript FontName, and it seems weird to me that it has two hyphens. The hyphen should be after the family name, only.

    Your fifth field has no spaces, but the only name that needs to be without spaces is the aforementioned PostScript FontName, so this seems odd to me.
  • @Thomas Phinney: Sorry, I mistyped, it should read as follows:

    De Vinne / 12Bold / De Vinne 12Bold / DeVinne-12Bold / 12Roman
    De Vinne / 60Bold / De Vinne 60Bold / DeVinne-60Bold / 60Roman

    as I specified in the beginning of the thread, these are the fields:
    Family name / Style name / Full name / Postscript / Style group

    My question was whether it could work using both the optical size and the weight nomenclature in the Style name, so you would have a single De Vinne Family with many styles. The alternative would be as follows, with the optical size in the Family name (and thus having a family for each optical size):

    De Vinne 12 / Bold / De Vinne 12 Bold / DeVinne12-Bold / Roman
    De Vinne 60 / Bold / De Vinne 60 Bold / DeVinne60-Bold / Roman
  • Thank you Craig!
    Yes, that was my rationale too. And I agree with Rui that, as long as we stray away from use in books and on paper, if we have to adopt descriptive names but relative to size, this would be a good solution (also aesthetically/economically in terms of name space).
    I prefer a XS/S/M/L/XL rather than, say Micro/Caption/Text/Subhead/Headline.

    But for my current project related to historical revivals, I’ll probably go with precise lead point sizes.
  • John Hudson
    John Hudson Posts: 3,204
    edited May 2021
    Claudio, it helps to remember that there are two different ways of grouping 'families' in OpenType: one way, corresponding to name table ID 1, groups into traditional 4-style families—Regular, Italic, Bold, Bold Italic (RIBBI)—, the other, corresponding to name table ID 16, groups into larger, more flexible families. This is a bit occluded by the way FL7 and Glyphs manage names internally.

    If I were setting up an FL7 source for a 12pt opsz bold master, I would probably do it like this:

    The 'Family name' is the name ID 16 overall family name. The 'Style group' is the name ID 1 family name for the 4-style RIBBI family.


  • c.g.
    c.g. Posts: 54

    Also, could something like this work?

    De Vinne / Bold / De Vinne 12 Bold / DeVinne-12-Bold / 12Bold
    De Vinne / Bold / De Vinne 60 Bold / DeVinne-60-Bold / 60Bold

    In this case, Windows users will be able to use only one of the two styles (I suppose the last one installed).

    According to the above convention, I've set two styles (Text Bold and Display Bold) of San Francisco as follows and installed both. Windows only sees one style (open image in a new tab):



    The SF Pro Text style will appear only if you uninstall the SF Pro Display style.

    Claudio Piccinini said:
    @Thomas Phinney: Sorry, I mistyped, it should read as follows:

    De Vinne / 12Bold / De Vinne 12Bold / DeVinne-12Bold / 12Roman
    De Vinne / 60Bold / De Vinne 60Bold / DeVinne-60Bold / 60Roman

    as I specified in the beginning of the thread, these are the fields:
    Family name / Style name / Full name / Postscript / Style group

    My question was whether it could work using both the optical size and the weight nomenclature in the Style name, so you would have a single De Vinne Family with many styles. The alternative would be as follows, with the optical size in the Family name (and thus having a family for each optical size):

    De Vinne 12 / Bold / De Vinne 12 Bold / DeVinne12-Bold / Roman
    De Vinne 60 / Bold / De Vinne 60 Bold / DeVinne60-Bold / Roman
    This could work:
    nameID 1: De Vinne 12
    nameID 2: Bold
    nameID 4: De Vinne 12 Bold
    nameID 6: DeVinne12-Bold
    nameID 16: De Vinne 12
    nameID 17: Bold
    12 and 60 will appear as separate families in font menus.

    If you want to group them as a single superfamily, change nameIDs 16 and 17 (and nameID 6 if you want):
    nameID 1: De Vinne 12
    nameID 2: Bold
    nameID 4: De Vinne 12 Bold
    nameID 6: DeVinne-12Bold
    nameID 16: De Vinne
    nameID 17: 12 Bold
    For style names that are not Regular or Bold*:
    nameID 1: De Vinne 12 Black
    nameID 2: Regular
    nameID 4: De Vinne 12 Black
    nameID 6: DeVinne12-Black
    nameID 16: De Vinne 12
    nameID 17: Black
    * In this case, better to clear the OS2.fsSelection bit 5 and head.macStyle bit 0.

    Anyway, San Francisco is not a good reference for naming, unless you want to cut off Windows users:

  • Thomas Phinney
    Thomas Phinney Posts: 2,889
    edited May 2021
    John’s advice looks solid to me—and Cesare is right about potential name conflict issues. You can’t (or at least, shouldn’t) have two different fonts which are identical for both name IDs 1 and 2.
  • Stephen Coles
    Stephen Coles Posts: 1,007
    I concur with those who favor S/M/L. It works for all media, is easy to understand, and is not too prescriptive (Caption/Text/Display) or specific (12/24/72pt).
  • I think prescriptive is better, because there's not enough typesetting education out there. (And once you're good enough you can confidently break the mold.)
  • Claudio, it helps to remember that there are two different ways of grouping 'families' in OpenType: one way, corresponding to name table ID 1, groups into traditional 4-style families—Regular, Italic, Bold, Bold Italic (RIBBI)—, the other, corresponding to name table ID 16, groups into larger, more flexible families. This is a bit occluded by the way FL7 and Glyphs manage names internally.

    If I were setting up an FL7 source for a 12pt opsz bold master, I would probably do it like this:

    The 'Family name' is the name ID 16 overall family name. The 'Style group' is the name ID 1 family name for the 4-style RIBBI family.
    Thanks much John, that helps!
    Cesare's suggestions dive deeper but I think I must have to study! :-(

    @Stephen Coles: I agree, I might end up with the clothing-sizes-like solution.
    Although this continues to call into question whether we are actually designing for specific print sizes, or if we are trying a flexible approach with an internal logic (small for small text on screen, text size for the average reading text on screen, etc.) but this becomes vague. I mean that if I am designing an optical size with attention on how it might render at an actual, physical size, I can’t accurately take into account its behavior in similar "small sizes" onscreen. Or, if we are reasoning relatively, in small text but on a billboard.

    This is relevant to me, as it is something which always frustrated me with digital scalable type, as I realized how beautiful, fitting and appropriate were size-specific designs in hot metal (when printed). I do not mean anything "esoteric" or "for the type specialist", rather a very simple consideration occurs to me: if a user enlarges a design with adjustments, inktraps et al., which is conceived for, say, a 5pt size, it will not look so great when printed in 72pt size (or used onscreen in similar fashion).

    I will think about this a bit more…
  • really dislike the point size specific naming (e.g. 12/24/72pt).

    I think type designers often forget that graphic designers aren’t the only people that interact with fonts.

    I find that client’s are confused when I tell them they need to purchase or use [Font Name] 12pt. It becomes especially confusing when the font isn't used at that specific font size.

    This problem occurs both with clients that aren't designers, and when you have to interact with web developers that aren't well versed in typography.

    I favor the slight ambiguity that comes with the S/M/L naming convention because it provides a general idea of intended size which can either be reinforced by context or simply behave as a suffix without being overly suggestive.

    The ambiguity allows a designer to "break the rules" without the name of the font working against their design decisions. I will sometimes style a "micro", "caption", or "6pt" font really big to accentuate its crude qualities. If that makes its way into a style guide, it can be really confusing to non-designers.
  • @Matthew Smith Totally agreed on the point size approach being terrible. Also that ambiguity is useful... but I think things like Caption/Text/Display incorporate that while also conveying the qualitative difference between types of reading (which S/M/L fails to address). It's not just about size, the behavior is different.
  • The ambiguity allows a designer to "break the rules" without the name of the font working against their design decisions. I will sometimes style a "micro", "caption", or "6pt" font really big to accentuate its crude qualities. If that makes its way into a style guide, it can be really confusing to non-designers.
    Would you have done so with lead typefaces, in print? The fact that a thing allows to be misused or poorly used, or that you can freely alter mathematically defined outlines without restrictions, does not mean you have to encourage arbitrary usage of form.
    If I design a typeface for a specific point size, or with specific qualities, I take into account elements and develop its design according to the design goal.
    By manipulating things not intended, you can be creative, but at the same time you do not encourage recognizing specific qualities, individual qualities, proportions, relationships among elements.
  • @Matthew Smith: To be precise: my question was not about what we might "like" or "dislike", personally I do not think exclusively (or not even chiefly) of graphic designers, I think about ordinary people, and about the page and its internal proportions. The screen comes only after that for me, and if I do design something primarily for the screen, it will have specific qualities and will be designed with different considerations.
    I thought my premises in asking the questions were clear, otherwise I would not have even mentioned the naming following the lead sizes.