Dynamic Text for Spatial Computing
I’ve used virtual reality on and off the past couple of years. Every time I re-enter it I become frustrated with the user interface in the operating system and other applications. The interface designs which are legible on screens, posters, and books do not translate well into spatial environments.
Take typography. It’s an excellent interface system for flat surfaces. Originally designed for the mass reproduction of texts on paper, the craft has always hailed legibility as its modus operandi. The seed purpose of type was to make information distributed and thus accessible.
But type often becomes less legible when looked at, tilted, in 3D space. After all a book has a “sweet spot” for optimum clarity which runs perpendicular and center to the page. Try to read a book by looking directly at the edge face of the page and you’ll have a hard time. Similarly in VR you must be at the “sweet spot” to use and see the interface properly. If the interface is not properly aligned, you’ll have to stumble blindly through your living room in real life just to get a look at the virtual window.
In the spirit of experimental typographers and graphic designers, who have run parallel to the rigid traditions of typography by destroying and rearranging the rules of type in their own playful ways, I set out to research the optimal spatial typeface.
I quickly discovered I was late to the party. The problem of clear type in 3D space had already been figured out. Highway Gothic was a typeface commissioned by the US FHA in 1948 for use on road signs on American highways, later to be superseded by the Clearview family of fonts —whose development rode on copious scientific research.
However, it’s important to see that the display of the copy varies depending on where it was located on the road. By the foot of every stop sign for instance is written in white paint “STOP”. The STOP on the ground is vertically scaled aproportionally to the rest of the face. From a bird’s eye view the font looks skewed. But, from the perspective of a driver, driving straight toward the word, and elevated a few feet off the ground in their vehicle, the text reads clear as day.
VR interface design could learn a thing or two from the US road system’s typography. In particular the STOP ground signage holds a lesson: type must take perspective into account. It’s obvious that an optimal type in 3D must be dynamic. In order to be legible, it must adapt its very shape to the viewpoint of the individual looking at it.
3D video games have long encountered this problem and have all dealt with it in the simplest way: place the text above the avatar’s head, and turn the text to always face the camera. While this technique gets the job done, it is an inefficient use of space, and breaks the illusion of three-dimensionality, which is a no-no in VR. In the example above, one can see the exploits of such a system, as the text can uniformly cover the screen as a wall of spam in a bona fide DDoS attack upon graphic design.
The solution I think is similar to the concept of an “Information Landscape” conceived by Muriel Cooper. It is the idea that text does not have to be constrained to flatness, but can effectively work in space, with appropriate presentation. In an information landscape, viewer and text engage in a dynamic tango, relating and acknowledging each other and achieving more intimate meaning.
That all said… the following is a demo example of what I think a dynamic font may look like. It can be used in virtual 3D space so that the viewer can have an easier time reading and being with the text. The properties of such a font are listed so as to…
- Be legible at many angles
- Adapt dynamically to an individual’s point of view
- Retain aspects of perspective to not break immersion from VR
- Have a small graphical footprint