Why 3D in user interfaces goes wrong 09:38 on Tuesday

The last five weeks our two-man office has worked on prototyping a flash project with a somewhat 3D user interface. Might have something to do with why I chose to post that quote about Brookstone creating a 3D store last week. I also implied that recreating a real life experience in 3D does not work.

I want to clarify that I do not think all 3D is bad (just most of it). Using projected 3D in a user interface goes wrong when it becomes a limitation — and in a “virtual world” meant for achieving something useful, it always does. I’m willing to change my view in a snap if anyone cares to give me a good example.

3D can be used for other UI purposes too, eg. the way it’s been used in Windows Vista or OS X (the cube transitions…). During the building of the proto we found out it can work very well, but it takes a lot of labour and discipline to avoid “getting trapped in 3D”, ie. being able to switch off the rules of the projected 3D space when it makes more sense for the user. This might be why (I think) concepts like BumpTop fail.

One Response to “Why 3D in user interfaces goes wrong”


  1. Henkka Says:

    3D UI when used in an “augmented reality” concept seems quite usable, however implementing such is not trivial.

    For an interesting location based 3D art idea I can highly recommend William Gibson’s new book “Spook country”.

    It lead me to think there could be a reason why you would want to limit some information’s availability to some location. It would be easy to implement with today’s technology. Just make some site available only when accessed through a certain VLAN access point, for instance. Note the difference of limiting instead of offering information based on location.