Hyperspace, implemented in Java and Javascript, lets you manipulate four dimensional shapes within a Web browser. It's lets you play with hyper-cubes, hyper-simplices, hyper-octahedra, 4D knots, ribbons, tori, and other cool shapes. It's got an "open" API - you can use it to define your own 4D shape directly within a Web document, without needing to do any programming.

Strictly speaking, this is Web4D. But that's ok, right?

The applet lets you manipulate the 4D objects, rotate them around, apply various rendering styles and stereoscopic viewing options, and in general have a grand old time. If you can completely understand what you are seeing, then you may be a mutant, and you will go far in the field of computer graphics.

This work is part of a long-term research agenda to figure out how the brain works. I have three hypotheses:

  1. People just can't visualize in four dimensions. Oh well.
  2. Sure they can, with the right practice and training.
  3. They can all right, but only up to the age of (say) five. Then they lose those language acquisition skills, and the brain hardens.
The answer is significant because of issues that go way beyond seeing in four dimensions. Is the brain a general purpose device, or are our perceptual abilities hopelessly wired in already, still making sure we can evade saber-tooth tigers? Here is a very concrete way to shed light on how flexible are our perceptual and learning abilities.

By implementing this work as a Web-based applet, I'm enabling myself to do widely dispersed user testing to answer these questions.

Coincidentally, all the hypershapes have the exact same colors as this year's SIGGRAPH logo. Which is nice.

- Ken Perlin