Saturday, June 8, 2013

Priming and Padding explained

Priming/Padding. Why is it necessary?

Visually and technically, padding refers to creating a group node above the object in question, and priming is creating a group above the object in question that is aligned (same pivot and orientation) as another object it is meant to control.

Why these may be necessary is simple.


1. If you have a control object that is already in place and you want to make it a child of another object that is transformed, you can't simply parent the object to the other. If you do, the control object is going to inherit those initial transformations from the parent object. So what do we do. We first group the control object to itself, and now we parent that group (or pad) to the other object. In simple terms, the pad serves as a buffer, in which the transformations of the parent object go to it instead of the actual control object. As a result, we have a control object that still does what it did before, is now controlled by the parent object, but is still clean of transformations.

2. Another scenario is with constraints. With a similar idea, you want to constraint object A so that it follows object B's rotations (so an orient constraint). By simply doing an orient constraint, your now 'locking' the rotate attributes of object A, because its being driven by object B. That might not be a problem if you dont need object A's rotate to drive anything else, but its a huge problem if you do. So, we use padding again. We make object A child of an empty group and we constraint the group node to object B. Now object A is controlled by object B because its local space is determined by the group node above it. But! It is still clean. Awesome!

3. Another really awesome reason why we use padding is to allow for our rigs to be quickly repurposed, or changed. What do I mean. Say your on the job, and you've been working on this awesome rig, and its perfect, your art director loves it, but now he comes up to you and tells you we have this scene, that can only be done with this foot rig. We cant use the original foot, we're bringing in a new foot, and we need it right now. Well, if the foot chain (from ankle to toe) was all in one group (pad), that pad can be constrained to any control object, and as we talked before, the joints will be controlled by those objects through the pad above it. The pad can also be parented under the end joint of the leg (in this case) ,so the foot behaves as the continuation of the leg joint chain. Now comes the new foot. Can you already see how easy this will be? All we need to do, is get rid of the original foot (not the pad), and place the new foot root joint under the pad. And Voila! The new foot now follows the leg, and the control object's constraints still carry over into it.

Watch the video on this topic here: