Calluna Sans as an alternative to Scala Sans, but...

Designing a book based on the layout used in one of Taschen's architectural series which featured Scala Sans for body text. Recreated the layout initially using Scala Sans which I'm enamored here - it seems to be the perfect "voice" for the author and subject matter. Of course Taschen's designers created a perfectly proportioned look amongst all the elements, worthy of study and emulation, but what I have is obviously too derivative, and I've begun customizing, starting with a similar font. 

Calluna Sans seemed to fit the bill, after experiments with Martin Majoor's (designer of Scala) and others' designs (Questa Sans, Proxima Nova, Nexus). Calluna is a good approximation of Scala yet with enough distinctive traits to set it apart. However, it's regular weight is a bit heavier on the page and the one light version goes too far. I suppose taking the Calluna 10 pt size I'm using down to 9.8 and slightly decreasing leading is not an option, as it alters the perfectly balanced look between page size and column width that I find the Scala Sans in 10 pt with 14.2 leading here has. Is it not advisable for any reason to use odd point sizes like 9.85 - does it complicate things in any other way? 

Thanks in advance for any feedback.


Comments

  • I've designed lots of publications using fractional point sizes for type and/or leading and never run into any issues, other than complicating the math slightly when determining "space above/below" values that tally up to a multiple of the leading.
  • Nick ShinnNick Shinn Posts: 2,131
    edited April 2021
    I find the most important quality to look for when deciding on point size and leading is how much one’s eye is tempted to travel up and down, rather than across. Several things in a layout affect this, for instance line length, and whether the paragraph style is ragged or justified. The most egregious problems are “rivers” (vertical paths of white space) and “islands” (vertical paths suggested by the proximity of ascenders and descenders). So it’s unlikely that a small fraction of a point size will make much difference in this regard.

    As you are concerned with the overall weight, you might also consider experimenting with adjusting the tracking plus and minus, or even the width of the space. Not to mention nano-stroking the text, i.e. “grading”. It’s all grist for the mill.
  • Joshua LangmanJoshua Langman Posts: 65
    edited April 2021
    This may or may not help you, but for whatever it's worth, when I redesigned the database School Shooters .info about eight years ago, I decided to pair Scala for text with Calluna Sans, rather than Scala Sans, as a companion sans serif. You can see the two faces working together in this PDF document, particularly in the charts and graphs. (The graph artwork was generated in Excel, but all of the type in the graphs was set in InDesign.) Here is another document where Scala and Calluna Sans are working closely together.

    I chose Calluna Sans rather than Scala Sans because it had a bit more typographic "presence," just a tad more personality. They are both humanist sans, but Calluna felt a little bit "more" humanist, if that makes sense. I did have to adjust the size of the Calluna to match the Scala, which is especially important in some documents where they appear together on the same line.

    The web interface for the database also uses Calluna Sans, but the primary text face is Skolar, not Scala. If I were to do the whole thing again, I would likely use the same serifed face for the website and the PDFs, but there were other considerations eight years ago, including screen resolution and what was available on Typekit (which we no longer use).

    To actually answer your question above, there is nothing wrong with fractional point sizes — points are an arbitrary unit of measurement — and I frequently go to two decimal places with type sizes. That said, I have found that when a layout feels a little wishy-washy or unsure of itself, returning to traditional typographic units and sticking to them, especially in the arrangement of empty space, can sometimes give it more solidity and conviction.
  • Joshua LangmanJoshua Langman Posts: 65
    edited April 2021
    [duplicated comment]
  • Joshua LangmanJoshua Langman Posts: 65
    edited April 2021
    One more thing about Calluna Sans is that it has gorgeous, distinctive numerals, both OS and lining. The numbers have these wonderful open counters, and the industrial serifs on the 1 sort of balance out the humanism of the other numbers.

    Here is an example of a magic sheet, which is a schematic representation of all the stage lights in a theatre, used by theatrical lighting designers. I have been using Calluna Sans for my magic sheets for years, partly because the individual numbers are so distinctive and easy to read in the dark. (I wanted to attach a PDF, but it won't let me.)


  • SilasSilas Posts: 19
    To be honest, I wasn't expecting such generous feedback here - thank you. Had actually been expecting to hear my original post was too verbose and wandering. Though I miss the site "Typhophile" sometimes, Typedrawers appears to be the real successor.

    Nick Shinn, grading in the direction I need to go (to slightly lighten the weight) seems not to be possible, unless I'm missing completely how grading is done. One could take Calluna Sans Light and thicken it, but I imagine this may upset the form too much, no? This is for the entirety of the body text.

    Joshua Longman, the description of Calluna Sans as just a bit more "humanist" than the humanist Scala Sans seems very appropriate - an impression I shared. There's more angularity to some of the details in Scala. 

    At any rate, I'm proceeding cautiously. The author is a noted architectural historian with a very refined eye and is a true bibliophile. She's mentioned how upset she was by a previous publisher (Wasmuth) which utilized an absurdly heavy font for the English portion of her text in a bilingual (English and German) publication. 

    I've tried decreasing the point size fractionally as suggested, away from 10 points, and lightly adjusting tracking. It's so minimal, but 9.9 or 9.85 makes a small, barely perceivable difference along with widening the tracking by 1 or 2. 

    I notice that once I go deeper into ID's Justification settings to work away some hyphenation, most of what succeeds in doing so shortens line length and then across the page you wind up with an overall darker look.

    I'm concerned we're going to fall back to Scala over this whole character weight issue. And then the totality of my design is just too derivative of Taschen's. One thing which sets it apart is the use of extensive and highly integrated shoulder/margin notes, which I've set just 1.2 points smaller but at 72% black. 

    Time to nail down these decisions as it's holding up work. 


  • If you are able to post sample pages of your layout, we could probably provide more helpful feedback. And if Calluna Sans isn't working for you, there are many other humanist sans faces out there — something which, only a very short time ago, was not the case.
  • Nick ShinnNick Shinn Posts: 2,131
    To lighten a font in InDesign, for instance, go to the Stroke palette and apply a tiny amount of <Weight> in the background colour, presumably white. Of course, you will need to experiment to get the desired effect.

    I have used this technique, for instance in beefing up Perpetua for footnotes.
  • SilasSilas Posts: 19
    edited May 2021
    duplicate post. edited by @Paul Hanslow


  • SilasSilas Posts: 19
    edited May 2021
    duplicate post. edited by @Paul Hanslow
  • SilasSilas Posts: 19
    Joshua Longman, I'm attaching all the pages here (10 spreads) in identical layout but with Latin placeholder text. A few of the superscript reference numbers in the body text for the shoulder notes are missing. Otherwise, it's a very close facsimile. 

    I've let go of the grid - just overwhelmed at the moment on how to tie this all together, so the baseline will be seen shifting and the shoulder notes not matching a set number of body text lines.

    Nick Shinn, I really must investigate/learn this grading technique - really not sure what impact this might have at the printers, but thank you. What do you think of the weight in the PDF?

    [EDIT] Ok, now I see I'm "not allowed to upload files in this category." So attaching screenshots. Print format size is 21 x 26cm.

    Very appreciatively - Silas 


  • SilasSilas Posts: 19
    edited May 2021
    If a moderator can delete the 2 duplicate posts above, that would be helpful. 
  • John SavardJohn Savard Posts: 1,088
    I took a look, finally, at the two typefaces under discussion here.
    What struck me immediately was how reminiscent of Gill Sans the typeface Calluna Sans was.
    Then, out of curiosity, I poked around Google Fonts to see if any of their free fonts belonged to the same general category of typefaces as these. It did not seem that was the case. I noticed a lot of typefaces there inspired by News Gothic, and even a few inspired by Futura, and even one that reminded me of Akzidenz Grotesk...
    Oddly enough, though, the only ones I saw that were a bit like those two  were fonts primarily designed to handle the Devanagari script, such as Mukta.
  • SilasSilas Posts: 19
    Uploading here as a PDF spread, at actual size.
    Appreciatively,
    Silas
  • Joshua LangmanJoshua Langman Posts: 65
    edited June 2021
    Hi Silas,

    It's hard to comment on the nuances of the typography given only the low-res images above, but the two things that strike me are (1) the measure (line length) of the main text seems a little long given the size of the type, and (2) I would not normally set a sans face justified. Certainly people do, but to me sans faces as used for running text feel inherently informal and therefore best set flush-left. Also, ragged setting goes better with your blank lines between paragraphs, which are likewise informal. Justified setting implies traditional indented paragraphs.

    If this were my layout, I would try shortening the measure and playing with the indications for paragraphing. I know this doesn't address your question about type size, but again, I can't judge that from your images.

    Josh
  • SilasSilas Posts: 19
    edited June 2021
    see below
  • SilasSilas Posts: 19
    edited June 2021
    Have tried again to upload here as a PDF spread, at actual size, and have requested permissions to do from admins, but so far, not possible. It can be zoomed into with good resolution. 

    Josh, in one way, this sample is not perfectly representative. It's the introduction written by an architectural historian, preceding the text proper, which will be illustrated in line, often shortening the measure. I suppose the overall impression then will be quite different. Also, the large number of shoulder notes are in the intro only and have been a unique challenge. The measure(s) mirror the Taschen model, but lack the illustrations common there as well. Both are justified on both sides, but the comment/recommendation certainly makes sense to me.

    Appreciatively,
    Silas
  • SilasSilas Posts: 19

    I think all will be surprised to see my returning to this 2.5+ years later. The amount of research that went into all aspects of this book has been extensive, and the new annotated and illustrated edition doubled in length over the original 1954 OUP edition. Recapping, it’s 21cm x 26cm, vertically orientated, hardback (a robust replacement for thousands of copies that entered libraries in the original OUP edition). Body text type is still Calluna Sans, 9.9 pt, tracking widened by 1, with 47 lines per page, and a text column width of 13.4cm, based off a Taschen budget series model of the same size but with Scala Sans.

    I return to sticking points/points of concern that attempts at resolving were made at the very start (when early mockups were done to garner support and funding), but now after completing most of the rest of the work, still remain.

    The humanist Calluna Sans seemed very promising but, as noted, it was a bit heavier in weight than Scala Sans used in our Taschen model, and Calluna’s light version was too thin to substitute, so we took 10 pt down to 9.9 and expanded tracking by 1—subtle but it made the text block appear a tad lighter and closer to Scala Sans. This might’ve added to another issue, however, as the measure of the Taschen model was already at the max for comfortable reading (actually a tad beyond at 84.8 characters per line, rather than the normal ≤80, and yet it still “worked”) but our book, with identical column width, averages 87.8 characters a line, a difference of 2.9 characters. The salience of this small difference of 2-3 characters rings absolutely true. It comports with earlier tests covering up that many characters with a blank sheet and seeing noticeably relaxed readability with just 2-3 characters less. Alternatively, narrowing the column by 0.5 cm makes a world of difference, resulting in an average of 80 characters per line, a more conventional recommendation, but that would alter layout.

    Alas, constraint-trapped now. Utterly redesigning the book, with its hundreds of images and illustrations, plus hundreds of sidenotes, precisely placed, is likely a non-starter. Not that all elements are locked in, but a majority. Ideally, a less dramatic intervention is sought that won’t require full redesign. The way out may not need to be so complex. The experience of others here might, I dare to hope, point out a low-resistance path.

    Options that come to mind...

    Towards the start, I once contacted Calluna's designer, Jos Buivenga, who helpfully modified the en-dash so liberally used by the author, removing bumpers and thus extending it slightly. This was so kind, but a request for a new weight between Reg. and Light seems unfathomable. Yet what works in Scala Sans at 10pt might have worked in a slightly lighter weight version Calluna Sans at 10pt if it existed.

    I adjusted  H&J long ago when focused on other criteria (looking extensively at work by Jost Hochuli at the time). My H&J  changes included some glyph scaling (frowned on by some, I know) which resulted in fewer hyphens and pleasing cohesive spacing of letters and words. However, I note now, in most cases it resulted in more text on a line. Its visible here in photos shared of the Taschen model and a scale printout of some of the edition’s text; seems that something in the H&J differs between them – that there’s a greater degree of word spacing in the Taschen. Settings are pictured.

    Is there a way to force InDesign to limit the character count 84.8? I’m seeing requests for this on the Adobe forum, but not a means to do it.

    Shaving a millimeter off the column width, perhaps, in conjunction with some recommended change in H&J may get us to 2–3 characters less per line. But narrowing the columns any more results in too much of a narrow tower of text. Expanding tracking by some egregious amount does it of course, but is not aesthetic.

    I’ve worn many hats in this publication, and so much of the work is now done, but the pressure to bring it together now before disappointing the distributors and other stakeholders is on. The book was always notoriously difficult for readers, and no disservice must be done to it.

    Apologies for the length. What would others do, or, more specifically and importantly, what do the very experienced eyes of the type experts here see in the photos?

    Very appreciatively - Silas



  • Given all the constrains you've described, your setting seems readable enough to me. It's difficult to tell just from the photo, but the sleight heaviness of the Calluna in relation to the Scala may actually help readability a tad.

    The first edition probably used only justification by word space. Your justification seems to be an improvement over that.

    The measure is wide in relation to the type size, but that seems to be inherent in the brief. Of all the metrics that one can juggle in composition, character count per se is one of the less important ones, in my view. Perhaps the time has come to leave it alone?
  • This doesn't address your question about character count, but you might consider letting the H&J settings go a little narrower with the word spacing and trying to limit the expansion of the word spacing. You have a few lines where the word spacing looks too wide to my eyes.
Sign In or Register to comment.