#16 Colour: A quick guide to its use in informative graphics


1.0 Introduction
The most fundamental employment of colour in [qualitatively/quantitatively] informative graphics is to allow the observer to easily distinguish elements of the information displayed. The strict definition of the term “colour” must be cast aside here, as just as useful are black, greys and white. Much of what will be discussed in this post may seem intuitive, but it is the guidelines and their employment in practice which unfortunately escapes the many.

The most important point to take away from this post is: colour used well can enhance and clarify visual information; and colour used badly will likely obscure and confuse.

2.0 Principals of Colour in design

2.1 Hue, value and chroma

Colour designers use three variables to describe any particular colour:

  1. Hue – Fig. 2.1 – The name of the colour (e.g. Red, or Blue)
  2. Value – Fig. 2.2 – The lightness or darkness of that colour
  3. Chroma – Fig. 2.3 – Equivalent to saturation (reducing to zero gives equivalent grey value)
huechart-01

2.1 – Hue values

2.2 – Value, or brightness.

chroma

2.3 – Chroma, or saturation.

2.2 Legibility

If something is legibleit is clear to read. Although this extends to literal clarity, here we’re talking about optical clarity to the human observer. Hue and chroma do not contribute much to legibility, it is the luminance contrast (or contrast in value) of the background and the foreground. Fig. 2.4 shows varying degrees of legibility, due to differences in contrast between the red and black text, and the background.

LEGIBILITY

2.4 – Legibility is a result of the difference in value (brightness).


Legibility, therefore can be achieved in a number of ways. Adequate contrast may be achieved in a either a monochromatic or multi-chromatic image. As far as monochromatic images are concerned (Fig. 2.5), the case is simple and value contrast is the only thing one needs to consider. For colour images (e.g. Fig. 2.7), in the strictest sense of the word, contrast surely gives us legibility, but unlike in monochromatic images, legibility is then not our only concern!

Print

2.5 – Monochromatic image achieving legibility through contrast in greys.

2.1

2.6 – Image shows high analogy of colours and low contrast.

2.2

2.7 – Image shows high contrast and two analogous colour groups.

3.0 Selecting a colour palette

3.1 Review of concepts

So by now you should understand the important colour principals should be employed in the following manner:

  • Select COLOUR to suite FUNCTION
    e.g. I’ll select a red hue for high temperatures and a blue hue for low ones. Intuitive, eh?
temperatures_city-01

3.1 – Colours here have been selected for intuitive function.

  • Use colour ANALOGY to delineate GROUPS
    e.g. Set A = Orange, Set B = Green. Each set may have different chroma values.
THEMPERATURES-2-01

3.2 – Here, the red and blue palettes have analogous variations of chroma (i.e. saturation).

  • Use CONTRAST to HIGHLIGHT
    i.e. New data vs. old data.
contrast-2-01

3.3 – Contrast here is used to highlight the “new” data over the “old” data.

  • Use CONTRAST for LEGIBILITY
    Often useful when labeling
Reducing the diaphragm on the microscope gives better colour reproduction for some camera phones. It also increases the apparent relief od the minerals.

3.4 – White text on a black background gives high contrast, which translates into high legibility.

In most situations where information is being displayed, it is best to limit your colour palette to two or three different hues. Using a limited hue palette, you are forced to increase variation by changing the value and chroma of your core hues. Examples showing theses limited palettes have already been discussed.

3.2 Colour Palettes vs. Symbols

Of course a lot of researchers need to display information in monochromatic images, due to constraints applied by the journal in which they wish to publish the data. The de facto choice for displaying multiple sets of information is to use different symbology (e.g. dashed/dotted lines, different point symbols). However, studies have shown that the human eye is only capable of understanding a maximum of 7 variables in any given set. This means that if you have more than seven symbols or line types, your plot is going to be hard to decipher. As discussed, the same applies to colour. So when it comes to symbols, it is also advisable to limit the number used to around three. This thinking is employed in the much used GGPLOT package for R. The package, written by Hadley Wickham limits the automatically assigned symbols to a plot (PCH values) to a maximum of seven.

3.3 Complex palettes

The idea of restricting your palette to only three hue values, or indeed the number of symbols used, is an ideal one. However, there are often many situations when you simply need to use more. More symbols or more colour? Well the advice from the experts is to use colour. It is much less work, and the mind is much better at distinguishing colours rather than a series of [more complex] shapes. So should you need to display 7 sets of data then colour is the way to go. Consensus on the ground is, that should you need to display more than 7 sets of data, then you need to rethink the way you are visualising your information.

3.4 Colour palette resources

Now you understand a lot more about using your colours, you could go out there and compose your own colour palettes… OR you could use any number of great websites to provide the palettes for you! AND if you use GGPLOT, the first of these (ColorBrewer) can be integrated with your plots! Remember, these palettes are not only great for plotting data, they are also great for providing colour schemes for posers and presentations!

4.0 A word on background colours

Some people prefer using background colours to their plots, commonly pale yellows or greys. The problem is that most colours (particularly preset colour palettes) are chosen to be printed on white paper, so it stands to reason that they should be displayed digitally on a white background too. Not only this, our brain is designed to constantly normalise our palette to a local reference for white. For example, if you’ve worn yellow ski goggles all day, when you take them off everything appears blue. This is also why when you take a photo indoors under a tungsten bulb, the image often comes out orange, more orange than you see it. This is because the camera’s white balance has not been set to match your eyes.

3.5 – Most camera users encounter white balance issues when shooting indoors. This is because the cameras white value is not set correctly. In the human brain, white balance is constantly being corrected so we would see the right hand image.

3.6 – Early computer screens didn’t have enough brightness to show black text on a white background. So the text had to be white and the background black to increase legibility.

Dark backgrounds are only ever recommended when the information is being displayed in a dark environment. The use of light text on a dark background reduces problems of legibility, allowing the viewer to see the information without being distracted by the background. The reason that early computer screens and slides from the earliest projection systems had dark backgrounds was because the displays weren’t that bright, and consequently needed to be used in rooms with reduced lighting. Now, however, projectors and screens are bright enough that this isn’t the case.

2 thoughts on “#16 Colour: A quick guide to its use in informative graphics

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s