I’ve been wanting to play around with Joystick n’ sliders [JnS] for little while now - In the past I’ve used the following technique to animate my heads. JnS appeared to be a more automated way of rigging a head in After Effects - which of course would mean a quicker way and hopefully a more efficient way of getting similar results.
The following is a record of the journey.
I started by taking a sketch a had laying around and bringing it into Illustrator to created a clean, vectorized version.
For now I would not worry myself about the eyes or mouth, I’d focus first on animating the facial positioning.
Position #1: Look Straight
Position #3: Look Right
Position 5: Look Down
Position #2: Look Left
Position #4: Look Up
I brought my Illustration into AE and broke each element into separate layers using Explode Shape Layers. I put the head, the eyes, the cheeks and the mouth each in their own composition.
From there I set keyframes for each of the positions you see above. Since there are 5 positions, you should have 5 keyframes on all the layers that are animated.
In most of the cases, I was animating the Path, Postion and scale of most of the elements.
Setting up the facial positions went fairly quickly. First this first test, I ignored setting up the animating eyes and mouth other then they’re positioning as the heads moves around.
Presently, I’m only using one joystick to created the movement.
Now I’ve noticed a couple of flaws in my setup that I’ll fixe in the next test such as:
The ears. In past facial rigs, I’d create an If/then statement that would turn the ears layers on and off depending on whether the ear need to be in front of the head or in back. In the situation, As I expected, instead of the transparency being on or off, we are dealing with percentage based on the position of the joystick.
The head: The head shape (which I inserted in its own comp) is also used as a mask for the cheeks and any other element that may fall outside of the facial area as the head turn. Inside the head comp I add two other shapes that I animate to give the appearance of .
I also found that position 5 does not “Look Down” enough - so I’ll need to tweet that in the next version.