vector help with multiple attractors

im having some conceptual trouble with what should be a simple problem involving vector summation (i think?).

I am working on a very typical multiple attractor rig setup where i want to rotate a series of panels (think:louvers) along the length of a line, their rotations driven by attractors. Now, the part Im struggling with is getting the "final" vector to look right - as it standsi dont feel that distance - or magnitude - of the consitutent vectors is being considered. i want to be sure that while remote attractors bear some influence this influence only holds in proportion to their distance...

the question i suppose is A. have you seen a good example of this somewhere? or B. can you suggest a good routine for this?

a mass vector addition anda divsion produces a vector that's a fraction of the addition which is not useful ( if oudn this in another user's example)
a mass vector addition feels right only it seems to emphasize longer vectors, even though those are the ones with presumably less "pull"...

help?

Replies to This Discussion

Permalink Reply by taz on September 29, 2010 at 1:34pm

Hey Gabriel,

I think the download for this link works, but for some basic trigonometric relationships check out this studio mode example file:

http://modelab.nu/?p=2817

Also this should be an example with weighted attractor points:

http://modelab.nu/?p=2862

Permalink Reply by Gabriel on September 29, 2010 at 1:49pm

I think I was working with their weighted version and it wasn't working for my particular scenario. In the meantime what I did was create new vectors by unitizing the originals and multiplying them by a factor proportional with their inverse...so that shorter ones become longer and more powerful and vice versa. I seems to have worked though I'm sure there's a cleaner way to do it.

new problem. my planes, upon which I'm building my geometry are flipping for some reason, again probablky something I'm over looking. No amount of plane rotation or other plane manipulation seems to be solving the problem...can you suggest how i can get all my vectors/plans to be oriented the same way?

Attachments:

vector rotation.bmp, 3.8 MB

Permalink Reply by Danny Boyes on September 29, 2010 at 1:49pm

Hi Gabriel, What about using the inverse magnitude to sway Vector direction.

The green arrow is Inverse and the Red is average. Whilst they are pretty much the same when close together only the closest points have a greater influence with the inverse.

Permalink Reply by Gabriel on September 30, 2010 at 9:58am

can anyone suggest - and I'm sure it's really simple geometry - why given origins and z directions, my planes are rotating - see picture. I have made the z and y vectors visible to illustrate that some are getting rotated - you can also see it in the rectangles, some which are one side of the line, others on the opposite side...

Attachments:

planerotation02.jpg, 308 KB

Permalink Reply by Danny Boyes on September 30, 2010 at 10:29am

I haven't looked into this flipping phenomena before but I think it is due to using a curve to create your points on. As a work around GH has a component that creates a box centered on a plane if you used this instead would it help? Or alternately you can array a point to get your base origins

Permalink Reply by Gabriel on September 30, 2010 at 10:31am

funny, that was my hack work around...thanks

RSS