Tabular Figures Width Consistency

Michael Jarboe
Michael Jarboe Posts: 265
edited September 2015 in Technique and Theory
Consistency between an entire typeface families' tabular figures width seems like a feature that would be desired by typographers, but there are many typefaces where each style's TF's width differs between styles. Is there a reason why typographers would not want set width consistency across an entire family? Is the reasoning to not design all styles with consistent tabular set width so each font style's individual TF's have a more harmonious relationship with its proportional figures?

Comments

  • Is there a reason why typographers would not want set width consistency across an entire family?

    Because it's hard work to make all the weights fit comfortably into a single advance width, especially when your family runs from very light to very black? Because many type designers would rather focus on pretty forms than think about users' needs?

  • If you have set width consistency across a range of weights, you essentially end up with heavier weights that are more condensed (and tightly spaced) and lighter weights that are more extended (and loosely spaced).  Whether that is acceptable or desirable is a big tradeoff. I don't think the utility of making all the fonts share the same widths is really worth it. I suspect most type designers have come to the same conclusion, because shared set width across weights is a very rare feature in a (non-monospaced) typeface.
  • Mixing styles in tabular material is very common. it's usually only done on individual rows, columns, or cells and the goal is to emphasize or separate that data by setting it apart.

    It's not the end of the world if your weights aren't tabular width -consistent. A slightly wider bold line isn't likely to upset the scanability of the table. Half the time, the bold data is in a column where it only needs to align with itself.

    That said it's not hard to keep the core RIBBI styles tabular width -consistent and let the extremes stray from that.

    Similarly, make sure you're not overdoing your tabular-width character set. I can be easy to forget how they are being used. All of a sudden you're making a mini fixed-width font. 
  • I don't think the utility of making all the fonts share the same widths is really worth it.

    I've designed a lot of annual reports, so my own POV on this may not be everyone's.

  • If you have set width consistency across a range of weights, you essentially end up with heavier weights that are more condensed (and tightly spaced) and lighter weights that are more extended (and loosely spaced).  Whether that is acceptable or desirable is a big tradeoff. I don't think the utility of making all the fonts share the same widths is really worth it. I suspect most type designers have come to the same conclusion, because shared set width across weights is a very rare feature in a (non-monospaced) typeface.

    Thomas, that's exactly what I was getting at, I wanted to hear opinions on that tradeoff and what, if any consensus there was among designers.

    Thanks everyone.

  • Nick Shinn
    Nick Shinn Posts: 2,207
    I don’t see a problem with distorting tabular figures to be markedly different from the corresponding proportional figures, other than all the extra work required. 

    However, I rarely do so for the very heavy weights.
  • I think I mis-read this by focusing on the comment above mine instead of the original question; this is just about the tabular figures in an otherwise proportional font, so I said completely the wrong thing. Both approaches are common (enforcing same tabular width vs letting it vary).

    Apologies for not reading carefully.  :(
  • This has been nagging at me, so I just want to add: most users don't need tab figures that duplex across weights. But those of us who do need them, need them badly. It's pretty painful when you've gotten a client to buy in on a typeface and then find that bolding a row of figures knocks them out of alignment.

    Nick's probably right that you can probably let the heaviest weights follow their bliss, since not many people use ultra bolds for charts.
  • If the extreme weight differ in width, why have tabular figure in extra bold to begin with?