 |
|
|
|
 |
|
|
|
|
|
|
To
make life easier the standard wing moves were set up as custom attributes
on a null which acted as a central control unit for the whole bird. The
picture below shows the attribute editor of my null group Caudeopteryx_CONTROLLER.
The second extra attribute (Right_wing_ELBOW) was creteated with
values of -45 to 45, defaulting at 0. Rightclicking on this property allows
you to load the attribute as a driver in the set driven key configuration
(2nd pic, sorry pix)
The 15 joints above the end feather bones detailed on page 2 were selected and loaded as the driven elements. After setting keys at the default position with the controller ELBOW attribute set at 0, I rotated all the joints by hand to their maximum and minimum fanning/collapsing states and set driven keys. Careful attention was paid to the overall position of the feathers (as they were not grouped uniformly - 1 feather to 1 bone etc..) to ensure a tidy representation for the wing at all states of fanning. 
The view of the graph editor shows just 3 of the 15 joints with linear curves between keyframes with x and y axis as the driver and the driven. A great feature is that these curves are editable however for our purposes the ELBOW driver would be properly animated for the velocity etc.. of fanning. Although attributes that are driven may themselves be drivers or be driven by other drivers, it is worth remembering that regular animation cannot be overalyed on top (at least I haven't found a way). You can begin to undertand how these principles could only be employed in just some of our shots. Ultimately, we used a different bird for each sequence depending on the animation and what it had to do. 
Below we can see the ELBOW attribute of my null controller in the channel box. The setting is at 41 which makes for releatively large rotations on the orange fanning bones. You can also see how this also drives the shoulder and elbow joints, collapsing the arm and pulling it in toward the body. The hands are not properly set up here - eventually they too were affected by the ELBOW driver so that the hands pointed backward remaining parallel with the outer wing edge during folding. Interestingly, the wing (although contracted) is still pointing straight out and does not curve round the body snugly and point backward and upward like a proper bird. Well, have a go if you like. We would have used maybe just 3 surfaces for a different, simpler wing and edited between the two in post. Luckily, it was not called for in the Horizon programme. 
Finally, the top view shows the feathers fanned out to an extreme and unatural extent. Again, the innermost group which blend to the body are not affected. The images above also show another custom attribute - Tail_fan - which was a similar parameter driving the fanning of the tail feathers. 
|