Today’s fonts

Television and video

There has been no coordinated effort whatsoever to design fonts for captioning and subtitling. A couple of individual fonts have been designed:

Typical captioning and subtitling providers (e.g., for DVD and VHS) use off-the-shelf fonts, like Univers 45 or Antique Olive. Or they might take it upon themselves to “improve” existing fonts. There is no professional oversight of these choices or improvements, and no testing that the fonts are actually superior for captioning and subtitling.

Fonts built into analogue TV sets for closed captioning are low-resolution bitmaps designed by the TV manufacturer. Many Line 21 caption fonts in North America – the name refers to the line of the TV picture where captions are transmitted – still do not even have descenders on lowercase g·y·p·q·j, but they do have italics (and underlining). Teletext fonts used in PAL countries typically are ancient bitmap fonts that are more suited to a dot-matrix printer than continuous reading. Those fonts tend to lack italics, too.

Fonts for high-definition TV sets in North America are, as aforementioned, off-the-shelf choices, almost invariably a set of print fonts that are claimed to be good enough for HDTV.


Film subtitling and captioning are heavily reliant on the technology used.

  • Laser systems that burn off the emulsion of the film, leaving white characters when projected, are limited to stroked or monoline characters, with no variation in stroke width. Some designers have been reasonably creative, coming up, for example, with a Helvetica lookalike in monoline form. There may be no technological way to eliminate the monoline characteristic of these fonts, but they could still be improved.
  • Photochemical systems achieve a similar result – removing the emulsion from the film to leave white characters when projected. Fonts could be custom-made to accommodate the ragged edges caused by this method.
  • Optical subtitling uses a second film that is printed together with the original one. You can use any font, and resolution is unaffected.
  • Nonlinear editing tools allow fonts to be superimposed before a film submaster is made. There is no necessary loss of reproduction quality in such a system, though it is rarely used.
  • Open captioning can use any of the subtitling methods. Typically, though, the laser method is chosen (due largely to the fact that one supplier does most of the work).
  • A newer form of open captioning involves projected bitmaps, similar to DVD bitmaps but not with the same resolution. Any font could be used.
  • Closed captioning in cinemas involves the use of a large display mounted on the back wall of the auditorium that displays caption characters in mirror-image. On the way in, the viewer picks up a translucent plexiglas panel on a stalk. The stalk sits in the cupholder attached to the seat. You position the panel just below the screen so you can read the reflected captions. Currently, the only displays in use are large LED datawalls of three 32-character lines, with each character built from a 7×5-pixel matrix – not enough dots even for descenders. However, that is not the only kind of display that can be used, and fonts can be customized for various displays.

Captioning and subtitling on the Web involve the worst of all possible worlds:

  • A lack of good fonts.
  • No research on what “good fonts” are: We don’t know anything about the specific fonts needed for online reading of captions and subtitles.
  • All the usual problems displaying fonts on computer screens, including format incompatibilities and antialiasing.
  • Inability of the designer to specify a font and know that it will actually be displayed.

Currently, it is difficult to add closed captions or subtitles to an online video stream. The author can specify a font, but the viewer has to have the font. Also, even the most basic typographic attributes, like justification and centering, are not reliable.

Open captioning or subtitling can be done, using any font you like. It is also possible to simply digitize an open-captioned or open-subtitled feed from TV or video, with the attendant problems of blurriness and artifacting in the digitized type.

Other applications

Other captioning applications – billboards, videowalls, installations, video art, kiosks, stadiums – have not been explored in any meaningful way. Some of those applications can get by with fonts from existing technologies, while others may require custom fonts.

Set-top boxes

We can expand our theme a little and include a discussion of set-top boxes, as used in digital and high-definition television. The fonts used in STBs are generally poor and are rarely custom-made for the limitations of the medium. There are no known fonts engineered so that visually-impaired people can use STBs – either as readable normal-size screenfonts or as custom large-print screenfonts for people with very low vision.

Some makers of STBs include fonts in their hardware, while others rely on bitmaps from the downloading source. Sometimes a mixture is used. Improving the accessibility to set-top boxes involves the creation of screenfonts for that application.

Note that onscreen menu systems for home electronics (including VCRs and DVD players) are analogous to the set-top-box segment and, if anything, are a bigger market.