Well into the production stage, after modelling and UV unwrap I have undertaken the rigging and skinning of my robot character. Unlike both my previous experiences (which are few), and my current notions of this process, the process will rely on a hirearchy and link system.
One common aspect however is the need for controllers. As has been repeated many times in the past, I have been instructed never to animate mesh. The reason for this
Aside from this controller system it was important for my process to first understand the major difference between an organic rig and skin pipeline and one of a hard-surface modelled character. The major difference stems from the deformation of an organic character as opposed to the rigidity of a hard surface character, in this case my robot character Georgio.
This difference is obvious but research was required as to how I would go about Georgio’s animation setup. In practice this setup has involved pivots, a hierarchy of scene objects and lastly animation constraints.
In principle this is a simple method, and results in a system whereby all parts of Georgio have a 100% weighting during the skinning phase. This of course eliminates the need to creat realistic bends and deformations of body parts. In-experience in rigging and skinning therefore should logically have less impact on the ease to which the final setup can be animated. Specifically Georgio’s range of movements should not be restricted by any areas that were not skinned to an optimum level.
The rigging and skinning of Georgio has however taken longer than expected. Along the way the consequences of such things as neglected xform reset has reared its ugly head.
In order to start this rigging/skinning workflow the model must be separated into the individual parts which will be rotated during animation. For Georgio this meant seperate objects for head, torso, leg and arm segments, finger joints and foot with ankle.
Once independant of each other it has to be decided where would be appropriate for rotation for each indepdenent piece. This involves affecting each objects pivot and placing in the correct point which allows correct rotation of joints.
After this it is now possible to create controllers, and using the align tool, match a controllers pivot with the object pivot. This essentially means that during the animation phase rotation of a controller will cause rotation around the designated point on the object. This also means that the controllers pivot does not need to sit in the centre of the object in order for rotation to occur in the correct location. After spending much time placing these pivot points in their correct position i am now more familiar with using the pivot adjustment functions within 3DSMax.
I have also reached a point of understanding that during the modelling phase it is important to model in a way which accommodates for the alteration of pivot location. This is particularly important for finding centre points within an object for rotation. A mistake which I had made is neglecting to have a symmetrical line down the torso, making it more of a tedious task to accurately re-position pivots to their needed spots.
The model could have benefited from foresight and planning in this way in other areas, such as creating the neck as a cylinder not tube, not only to make the object more contiguous but also to be able to find and place the pivot at the centre point of one end with ease.
Perhaps it would seem a little insignificant to model edge loops etc into a model for the sake of easy pivot adjustment, but even for the cases where it wouldn’t also benefit other aspects of the pipeline production, at this point in time I believe it is worth implementing modifications for pivot adjustment alone (at least when this rigging system is used).
As I have never had to manipulate pivot points in this way for this purpose I doubt that my mistakes could be avoided completely, however I have taken notes for future reference of considerations during the modelling phase. In fact, the importance of beginning these considerations during the character design phase is now something I am beginning to understand in not only a theoretical sense but also practical.
Despite the obviousness of my limitations of technical aptitude and lack of experience, I anticipate that planning ahead for not only technical capability limitations but also time factors is something I will require and be refining infinitely.
Reflecting on this process, one would need to assess the final production in all its practical intricacies of a character from its creative conception. The struggle of creating a character with appeal and polish is already a feat, but how much compromise should be made in order for an efficient pipeline, and easy (in relative terms) to complete project?
I realise that such a musing sounds dramatic, and my intention of such a reflection is more of a hyperbole than anything else but the ramifications of not considering this at all seem to be even more drastic.
Returning to Georgio’s progress in the rigging and skinning portion of the pipeline, the algniment of the pivots was followed by a painful process of discovering the various effects of hierarchy and constraints that compete with each other.
It is important to place the controllers and objects in hierarchical levels which they will be animated by, with the root of the character at the top. It is also important however to ensure that the objects that comprise the body are part of a separate hierarchy tree to the controllers, with each level in both hierarchies mirroring the corresponding components. This would mean the torso would be at the top of one hierarchy, and the torso controller at the top of another. After some accidents I have seen how this directly affects the intended movements of the character.
I have also learnt that constraints can be stacked, and it is easy to constrain an object multiple times without realisation due to little interface feedback. The model may show odd characteristics of movement but it necessary to open up link info and examine to confirm any new constraint additions in order to delete the unintended additions.
All in all I am satisfied with the new knowledge I take from this new rig and skin experience. in a sense it solidifies what I have previously learnt when rigging and skinning an organic model.