Tufte’s Envisioning Information is by far the most beautiful text I’ve had to read so far this semester. It certainly establishes authorial credibility. When it comes to color, he promotes the following ideas:

  • Use no more than 20-30 different colors when encoding abstract information. Beyond that it is too difficult to distinguish individual colors.
  • Color can be used:
    • To label (color as a noun)
    • To measure (color as a quantity)
    • To represent or imitate reality (color as representative)
    • To enliven or decorate (color as aesthetic improvement)
  • All colors live in 3-space: hue, saturation, and value
  • Color can improve information density
  • Look to nature for inspiration for color choices

In particular, I payed attention to his commentary on digital displays compared to the screen:

Noise is costly, since computer displays are low-resolution devices, working at extremely thin data densities, 1/10 to 1/1000 of a map or book page. This reflects the essential dilemma of a computer display: at every screen are two powerful information-processing capabilities, human and computer. Yet all communication between the two must pass through the low resolution, narrow-band video display terminal, which chokes off fast, precise, and complex communication. Color can improve the information”

In my experience the interface bandwidth is much more limited by the tools we use to interact with the computer (keyboard, mouse, touch pad), but I understand his point. Even large screens don’t feel like they’re large enough compared to books, and I think it comes down to density – we can easily bring a sheet of paper closer or further away to change the amount of information we’re receiving, but the equivalent gestures on a computer are non-existent, or technically challenging. Even with the advent of high-density, large-format displays (I have two 3840x2160 pixel displays on my desk, one at 32” and one at 28” in diagonal) it’s much easier to read on paper.

But on the other hand, some kinds of visualization are enabled by digital interfaces that aren’t as easy to understand or are simply impossible when represented on paper. Any kind of interaction, for example, particularly panning and scrolling through a 2d or 3d space. And even zooming, which I mentioned above, can work better in digital, as long as there is no constraint that all of the zoomed-subject be visible at once.

I’m excited to learn more about the kinds of interactions we can take advantage of with digital representations. The space is very much still unexplored. While we explore it, we will do well to remember another of Tufte’s recommendations:

“Above all, do no harm”