August 28, 2003
Why 3D can't work.

From the "Information Visualization" Dept.

I was chatting with some designers today and we started talking about next generation interfaces. One designer was thinking about 3D User Interfaces and how that would be a cool way of abstracting information display.

I disagreed. My argument goes like this:

Humans don't see in 3D, but rather something like 2.5D--we see a 2D image with an instinctual understanding of the concept of depth (the 3rd dimension). Only after we move within a space do we start to understand three dimensional spaces (try walking in a maze of mirrors). Among women, spatial cognition is also not a typically strong skill (and I would say that is probably correct in most humans as well)--so thus you've just handicapped 50+% of your potential user base. Therefore the thought of using simulated 3D, without a real simulation of optical depth just doesn't make sense to me.

Alas, I don't think we'll ever see that 3D Unix interface in from Jurassic Park used in Windows.

It also goes against one of the most basic design rules: simplify as much as possible or KISS.



Posted by taitoh at August 28, 2003 04:59 PM

Why stop at 3D ? Some people (with practice) seem to be able to maneuver through (virtual) 4D spaces.

Posted by: David Cary on September 23, 2003 12:21 PM


I come from a Kinesiology-based (study of human movement/performance) background. So I'm rather biased. My frame of reference revolves around the notion of the "Gibson Affordance" -- a theory about how our brains interpret the surrounding environment. The basic tenet is that in order to perceive our environment, we must interact with the environment. We need to move. Studies in human movement and motor learning seem to support the idea that movement is essential to the processing of the external, immediate environment (although there is just about as much data that suggests the opposite).

It's probably a bit of both; you could probably train yourself too. This would explain why we see children being able to work in virtual 4D environments (they have an excess of neural mass anyway.....)

4D spaces? That blows my mind. It does.



Posted by: Tai Toh on September 26, 2003 01:02 AM

In terms of "next generation" interfaces, there are some interesting things around the corner.

Microsoft UI research has made a hybrid 2D/3D-motif called the TASK GALLERY. Multiple desktops abstracted as paintings hanging in a gallery.

David Gelernter from Yale's Computer Science Department has been trying to use "multi-dimensional" scaling to increase the effectiveness of the Desktop / WIMP (Windows, Icons, Menus, Pull-downs) metaphor. He's created a product called LifeStreams that allow you to organize your "state" via a time-line.

He's further enhanced it to have a more "document-centered" approach; it uses a timeline & file-type to manage documents / events. He calls it Scopeware.

Personally, it's inevitable that there will be a move to a "Document" centered OS / GUI. If there is a 3rd dimension to be used, it will most likely be a timeline based solution similar to Dr. Gelernter's work. I also think gesture recognition will supercede the "point-and-click" method. There's also voice recognition as well....



Posted by: Tai Toh on September 26, 2003 01:31 AM
Post a comment

Email Address:



Remember info?