### Project 3 – How 30 balls move- Xing Xu

by xing @ 11:33 pm 29 February 2012

There are three different system(expel, attract and block) which there are 30 balls in each system:

In the first one, each of the ball in the system will randomly pick another one in the system and make attraction with it. Thus later the whole groups will move around. In the second one, the ball will do the same thing but in the opposite way(to expel the other ball).  In the third system, the ball will always go into the middle of two balls. They act like they are trying to block the sight of the other two balls. The idea comes from the Improvisational Acting class I took in the last semester. I remember that my teacher, Brenda Bakker Harger, asked us together into the center of the space and pick two people and try to form an equilateral triangle with the other two people. But the actor is not gonna let the other two people know who did the actor pick. So I am thinking it will be fun to let the balls to simulate the system. Also it just like how people behave in the society. They try to get away from somebody; they try to get close to somebody; they try to block the relationship of the others.etc.. It maybe interesting to add more behaviors to the balls.

For this program, I was using Processing with the physical system “TRAER.PHYSICS 3.0″.

It is very interesting to see the process of the movement and the stage when it gets stable.  Here are some of the screen shots.

## 1 Comment

1. Improvisational acting — you pick two people with whom to form a triangle, but they don’t know it. Nice starting premise.
But the bouncing off the walls makes this look if this is a simulation of people in a space,
I’ve done this exercise – it’s a great inspiration for this project! But I agree that you could show the results more clearly. Why don’t they ever converge on triangles? Are there certain formations or behaviors that converge faster?
Like your first project, the visual representation you settle on doesn’t clearly demonstrate your intent. This may be clearer by simply adding a character class rather than a ball.
Genuine inspiration. But perhaps you should have the line stroke flash red when a triangle forms from intersecting balls.
*yes!

It’s hard to understand what is going on, i have to admit…

I’ll admit that I’m not quite sure what’s going on, but like the robot project, seems to be some great opportunities here to explore giving human traits to inanimate objects or, for you, abstract geometry.
There’s a small bit of maneuvering, but the wall physics DO seem to be dominant.
Even just hiding the springs would make them seem much more like characters.

i think you can also do things with the colors for the balls.

Do the balls have any autonomy? Do they have steering behaviors? It looks like a basic physics simulation, which doesn’t give the impression that the balls are human actors. I think this is only interesting if you were actually simulating human or human-inspired behavior. It’s not an issue of forces between particles; even if you only had one particle, I think it would still look like a bouncing ball. For this to be interesting, each particle needs to be interesting on its own.

The lines do make some interesting patterns, but you can’t really see them beacause there is so much drawn on the screen that all looks the same.

I agree the lines create quite beautiful patterns, like an active mobile, but it’s too simple to keep me engaged for very long. Up the force on your physics x1000 or get some serious visual styling in there.

maybe if the particles were color coded to show which ones are following eachother? it is really hard to understand the relationship between particles because there is so much going on that all looks the same

you might want to check out tableau machine: http://dl.acm.org/citation.cfm?id=1409685

I think the visuals look good. really?

does each node lock onto two others and then stick with them the entire time or does it try to find new people to form a triangle? I’m having trouble relating physically to this due to the abstract nature of the presentation… Maybe use images of people or icons, etc… better modeling of motion using perlin noise could help… Modeling behavor of people that’s unpredictable… like stopping suddenly, or moving smoothly at one moment… maybe the other takes time to adapt and react causing the triangle to get out of sync… I don’t get the sense you spent much time on this–that you took a demo from a physics library and formed triangles.

I really like the premise of trying to simulate the behavior of this improv game. I think for the animation you might want to try some easing function so that the nodes accelerate and deccelerate instead of just instantaneously changing velocity. Another constraint that might help the algorithm is some avoidance focrces for the particles to which they aren’t connected. So each particle attempts to maintain this relationship with the other two in its triangle, but it ignores all the other particles in the room. What would the behavior look like if each particle were trying to maintain that trinagular relationship without running into any of the other particles? Another problem may be that there are too many particles on the screen to really discern movement patterns.

i agree with golan, i’d like to see it without bouncing, but instead the flocking coming through the other side, or maybe if they were smaller it would be easier to see what was going on

I cannot really see the elements of improvisation in the simulation.

Comment by admin — 29 March 2012 @ 7:02 am