Good Icon Design

Designing an icon for use in a Google Earth project requires thought about how that icon is to be used. If it is to be used once, as in the Eagle icon in 'Explore Artic Drilling' then you can design something with unusual shapes and colours to be eye catching, I liked this little fellah found in a good article on icon design for desktops here.

However, if you are using a number of icons to mark multiple of points then your design must be different. You'll need to make the icon easy to differentiate from other icons and still visible when small - reducing size reduces clutter. You will also want to put common features into icons that should be grouped together. Lets look at an imaginary example

Taking the excellent explore artic drilling project I mentioned before I took the single use icons they used to mark 'read this for more information on wolves/eagles' and copied them around to simulate a project where you are trying to show sightings of these animals in a field study.

Immediately you can see that the blue borders and use of the same square shape limits the ability to differentiate between the placemarks. I also think the three colour design (brown, white, blue for eagles and grey, white, blue for wolves) looks 'busy' when the icons are repeated a number of times.

However, the blue borders and white backgrounds do strongly mark out the 2 types of placemark as having similarities, which is good as they both represent animal sightings.

In my version I have drawn some new icons (notice I'm no graphic designer!), the two colour design reduces busy-ness and the simplicity of the icon allows it to be reduced in size which allows more positional information to 'read' by the user. Also, the octagonal shape of the wolves (based my late grandfather's favourite soccer team logo by the way) compared to the rectangles of the Eagles allows more visual information for the user to help differentiate the placemarks.

In terms of linking the animal groups together, the light yellow eyes of both animals and the similar background colour sends the message 'these icons are in a group'.

A lot of these rules of thumb come from a major influence of mine Edward Tufte, These seem to be notes from one of his lectures on how maps should be constructed. One of his key points about any informational graphic is that the comparison of the data is key to good design, thus in the Explore Artic Drilling project it is the wealth of oil icons compared to pristine wilderness nearby that is striking.

So you now have some ideas of what makes a good, multiple use, icon but you have no idea how actually to do it. Rest easy gentle reader, I will cover that in a later post. If you want to see how the project works at other view points the .kmz link is here.

I hope that no one reads criticism of the Explore Artic Drilling project from this post, none was intended, I just used it as a handy source of a problem and icons to illustrate an issue.

Wild West of Design

A while back I swapped emails with Keene Hayward the person at National Geographic responsible for the Ivory Wars Google Earth 'Zakouma' project. Whilst I thought their Google Earth file showed some fascinating data, especially about the trail of an elephant through virgin African jungle, I made some points about how the design could be improved, I'll get to those points later in this blog but I wanted to share part of his general response to the criticism with you now.

He stated that whilst National Geographic had a well developed magazine design, developing for Google Earth is new, and we are in a kind of 'Wild West' period at present - everyone is guessing at what is the best way to present data, there are no norms to guide your design. In case you're wondering what a 'norm' really is think about a newspaper: if you're in a new city and pick up a paper you've never read before you already know some things about the content. Sports will be on the back pages, the biggest writing on the front of the paper is a headline and the writing is arranged in a series of columns.

The theme of this blog is to explore the best way of putting a Google Earth project together, I have ideas about what works but part of the fun is that no one really knows the BEST way to design a Google Earth project at the moment - we're all on the new frontier just having a go. Which is why I'd like to encourage you to add comments to these posts or, if that's a bit scary, email me with your thoughts.

Top 10 Google Earth Bad Design Practices

In this post I refer to a 'Google Earth project' by which I mean a collection of multiple elements that have been wrapped up in a file readable by Google Earth (a file with .kml or .kmz at the end). There are some good examples here .

1. Placemark icons that are overly complex and/or have too many colours (A simple 2 tone icon such as the orange blogger icon top left of this page is good)
2. Users encounter a cluttered view when opening a Google Earth project
3. No attempt has been made to give the users an introduction to what data layers they can view in a file
4. Use of acronyms/code names (e.g. 'Sensor X10-7/2007') where unnecessary or where they haven't been explained
5. Too much text used in pop up balloons, no understanding that users will just skip large blocks of text
6. 'Chart Junk', elements that are in a file but actually do not add any relevant understanding and just get in the way of overall clarity
7. Not editing camera positions. Placemarks can each have a unique camera position which defines a custom view, a powerful way of enhancing patterns of placemarks or their position in the landscape
8. Use of overly colourful or thick lines in polygons. Lines have less 'clutter' effect on a view if they are as thin or bland in colour as possible whilst still remaining clearly visible
9. Not encouraging users to turn layers/folders on or off to aid understanding. Users can be advised how to view data to help them understand the message in a file
10. Not picking out key data points or study areas. Users have different lengths of time to spend looking at a project, some will just view what is on the screen without clicking anything, some will spend a lot of time exploring your work. There are also a set of medium length users inbetween, if some key parts of the data set can be emphasized the medium level user may be encouraged to stay longer.