Comments on: ��why so trippy??

Question: ¿¿why so trippy??

garethspor — Sun, 08 Jul 2007 16:45:43 -0800

What kind of interference-pattern phenomenon have I got going on here?

My interest was piqued by that representation of Danny Hillis's Thinking Machine architecture in an old back issue of Whole Earth Review. So I plotted out some images of large grids of evenly spaced points in 3-dimensional space and found all sorts of cool (what i believe you would call) 'interference patterns.' I then applied a simple gaussian blur to the images and animated their rotation, resulting in the trippy youtube video. (all images and video in the link from the main question)

Surely I'm not the first person to play with these sorts of images? Any math/art geeks out there wanna share some information on what I've got here?

By: DarkForest

DarkForest — Sun, 08 Jul 2007 16:55:01 -0800

You see this in regular arrays of objects, like tombstones, of which this is a poor example. Don't know what the effect is called.

By: niles

niles — Sun, 08 Jul 2007 16:59:54 -0800

I'm pretty sure the wikipedia entry on Moiré patterns would be useful.

By: Kadin2048

Kadin2048 — Sun, 08 Jul 2007 17:10:14 -0800

Seconding a 3-D version of a Moiré pattern. I've never seen one like that in anything in real life, but it seems like what you'd get, if you could imagine a three-dimensional array of tombstones, as in DarkForest's picture.

By: effugas

effugas — Sun, 08 Jul 2007 18:09:50 -0800

Yes, you're seeing a Moire effect, but I can go into a little bit more depth than that. Three dimensional scenes have something called Parallax, in which things farther away from you move slower from your perspective than things nearby. What's happening is that you're projecting a set of points, and the nearby ones are moving in sync with the nearby points, while the far away ones are rapidly falling out of sync (since they're moving slower). So as you rotate the cube, the rear points are moving from wherever they are, to either closer or farther away from the points up front. When they happen to cluster, things get dark. When they happen to spread out, things get light. As you can see, the pattern isn't exactly random -- thus, Moire effects.

By: garethspor

garethspor — Sun, 08 Jul 2007 18:26:10 -0800

Ok, so i understand the basic principles here, the farther away dots appear smaller, or more tightly packed, so, higher frequency. The frequencies interfere with each other to produce the resulting pattern. Basically, when all the dots line up, less of the dots appear in the projected image, resulting in a lighter area. Vice versa, when the dots don't line up, more of them appear in the image, producing a darker area.

The pattern of dark areas and light areas is what I'm interested which is why I added the blur to the video to accentuate the shading effect of the dots. What I'm curious about is the regularity of these areas, which seems to adhere to some sort of underlying geometric principle. Anyone have any clues?

By: Reggie Digest

Reggie Digest — Sun, 08 Jul 2007 18:31:48 -0800

It's just a Moiré pattern. (Put one window screen on top of another; rotate.)

But, for what it's worth, it's the most kick-ass Moiré I've ever seen.

By: Reggie Digest

Reggie Digest — Sun, 08 Jul 2007 18:34:31 -0800

underlying geometric principle

Click niles' Wikipedia link and scroll down a bit.

By: O9scar

O9scar — Sun, 08 Jul 2007 20:37:45 -0800

This PDF (5.4 MB) discusses the effects of aliasing on a simpler 2D pattern - it's one I remember seeing in Scientific American back in the day.

Check the aliasing entry on Wikipedia too.