Monthly Archives: January 2012

Stuff Flash artists should NOT do

People that worked with me knows that I kinda freak out when I notice that an artist is doing any of the below-listed things. Following this small little rules can help your relationship with the engineering department and can also make you look cool! The list Do not use Groups Actually remove the letter G from your keyboard otherwise that easy appealing CTRL+G will always tempt you. Using groups always leads to uncontrolled agglomerates of drawings all over the place, instead symbolize them, that way you have them in the library, and you can also reuse them. Done that you are already far ahead of the rest of your crew that didn’t read this. Don’t draw twice the same stuff This is where rule number 1 comes handy, instead use a movie… [Read more »]


Complex Multiple Segments Inverse Kinematics

Yesterday I wrote a Simple class to handle simple Kinematics, but I was not very happy with it, I needed to use more segments, so I started looking around. I found this amazing experiment made with JavaScript (Source code) which really inspired me. Looks like the method used in this experiment is pretty valid and I found mentions of it in this two PDF regarding Inverse Kinematics for a Serial Chain with Joints under Distance Constraints and A fast, iterative solver for the Inverse Kinematics problem. The result of my experiments is this: Click on the stage to change to focused point or press any key to toggle the lock on the fixed points. The algorithm check recursively each point from the focus point and back, left and right. It can be pretty much explained with this… [Read more »]

IK Iterative

Simple Inverse Kinematics

I have been looking for a while with interest at how inverse kinematics works, on Wikipedia I found this page regarding the Law of cosines which sounds interesting, although looks overly complicated. Other methods to take in consideration are the Cyclic-Coordinate Descent (CCD) solution and a Jacobian matrix iterative solution. I have been able to determine the exact position of multiple joints using a mix of those, but the resulting algorithm looks ugly at least. So while I am looking to find a cute one enjoy this piece of code to calculate IK on 2 joints.