Custom fonts that blend two existing typefaces — any other examples?

Raven
Raven Posts: 1
in Type Business
Recently saw the release of Nebula Sans, a custom font for Nebula TV that serves as a close alternative to Whitney ScreenSmart, but the drawing is based on Source Sans 3. It's also released as an open source family. 

This got me thinking — are there other examples of a custom typeface where the brief is essentially "take the drawing of Font A and adjust it to resemble Font B"?

I don't recall seeing this kind of dual-influence acknowledged so openly in type documentation before. Would love to hear if others have come across similar cases.

Tagged:

Comments

  • John Hudson
    John Hudson Posts: 3,617
    Nebula Sans is described as ‘designed to be a drop-in alternative to Whitney SSm’. That doesn’t imply that the typfecae glyph shapes have been made to resemble that of Whitney, but typically means that the scaling and weight/width instance interpolation has been selected to be similar to the weight and proportions of Whitney, and the spacing has been adjusted to match the Whitney metrics. The idea is that one can drop Nebula Sans into designs and templates spec’d using Whitney SSm without needing to adjust anything. The look and feel of Nebula Sans is still very much Source Sans.
  • Thomas Phinney
    Thomas Phinney Posts: 3,114
    edited March 3
    When it comes to metrics-compatible fonts, there are probably a couple dozen that started as Font A and were adjusted to be metrics compatible with Font B.

    Carlito started as Lato, then was adjusted to be metrics-compatible with Calibri.
  • Thomas Linard
    Thomas Linard Posts: 21
    edited March 3
    Given the client company, is there a planned use of Nebula Sans for subtitles?
    I ask this because I always find it strange that fonts intended for use in displaying subtitles do not include ♪ U+266A Eighth Note or ♫ U+266B Beamed Eighth Notes, given that:
    1. many software programs displaying subtitles are unable to compose their display from multiple fonts at once;
    2. many subtitles include ♪, or even ♫ (it's even recommended in several standards, cf. "I.15. Songs" in https://partnerhelp.netflixstudios.com/hc/en-us/articles/217350977-English-USA-Timed-Text-Style-Guide or https://www.bbc.co.uk/accessibility/forproducts/guides/subtitles/#Characters-permitted-online).
  • SCarewe
    SCarewe Posts: 19
    Two off the top of my head:
    Carlito, as already mentioned, is designed to be metrics-compatible with Calibri. Lucas was very pissed about this, as it means copying (stealing?) all metrics, including kerning. I would be interested in the legal situation here, if anybody wants to weight in.
    DB Sans (for the Deutsche Bahn, the German railway) has an Office variant, which is designed to be metrics-compatible with Arial.

    Reading your initial question again, I realise this is not exactly what you asked, but Thomas made me think of the second example.

    Over the last year especially, I have seen more than one custom font a company that was sick of paying Monotype far too much in licensing fees, and thus they opted for a custom font instead, which costs about as much as they were paying per year to Monotype. Looking at the font the company was using previously, it’s more than obvious that the brief was “Make it look like our old font, we just want to cut the licensing fees”. I also got one brief exactly like that: “We’re using Helvetica, but the yearly licensing is too expensive, make us a font like Helvetica”.

    This is a tangent, but I am honestly optimistic about the market for independent foundries, thanks to Monotype. Custom font requests are more often than not made with the reason that Monotype’s recurring fees are just too high.
  • Tofu Type Foundry
    Tofu Type Foundry Posts: 72
    I don’t understand how reusing/stealing another font’s metrics can be feasible since different letterforms will require different adjustments to spacing and kerning. Or are these designs often so close to the original that it simply doesn’t matter?

  • Mark Simonson
    Mark Simonson Posts: 1,778
    edited March 5
    This practice goes back a long way. Consider that Arial used the same metrics as the version of Helvetica (licensed from Linotype) that was included with PostScript so that PostScript clone laser printer makers could substitute Arial automatically when printing documents spec'd with Helvetica.* The substitute face may not be as well optimized to the metrics, but that is not always the primary criteria. In the case of PostScript clone printers, they traded visual optimization to offer a product at a lower cost (partly by not having to pay to license Helvetica), and there were customers who were okay with that trade-off.

    * This was happening c. 1988 or 1989, before Arial shipped with Windows.
  • John Hudson
    John Hudson Posts: 3,617
    I don’t understand how reusing/stealing another font’s metrics can be feasible since different letterforms will require different adjustments to spacing and kerning. Or are these designs often so close to the original that it simply doesn’t matter?
    It’s possible to produce designs that differ in all sorts of ways — stroke modulation, construction, idiom — while retaining similar proportions and hence spacing/kerning.

    Also bear in mind that historically mechanical typesetting technologies tended to be metrically constrained, so lots of designs ended up on similarly unitised widths or otherwise adapted to the way the machines worked (e.g. the infamous Linotype non-kerning f), and those metrics were directly translated into the initial digital versions of those types. So while people regularly point to Arial as a design that was produced to be metrically compatible with the version of Helvetica that shipped on PS printers, fewer have examined the metrics of that Helvetica and realised just how wonky the sidebearings are because of inherited unitised widths.

Categories