I was approached by Dan Thurmon to do a promotional piece for his speaking tours. Dan is a motivational speaker and author. We discussed a number of ideas and set out to make a clean yet interesting presentation. We knew it would be an effects heavy project, so we broke it down to sections.
As we discussed the project I discovered several challenges I needed to solve.
- Dan wanted to wanted to juggle and have the balls freeze in mid air, at which point he could walk around them, then have them fall out of the air at any time.
- I would have to not only make an animation of juggling, but it would need to transition smoothly from 1 ball up to 5 balls.
- Later in the piece he wanted to show the correlation between the topics by drawing a line between the balls.
- He wanted the animated balls to come to a complete stop mid talk.
- Lastly Dan also wanted a “Danless” version that he was not in that he could use for live situations.
While talking through the process and tossing around ideas, we developed an idea of image frames that would go behind him. I was worried that the image would be somewhat stark if it were just Dan standing on a surface talking to us without some sort of environment. We needed a vehicle that would be clean, visually interesting, and add to the presentation while not becoming a distraction. So I developed a screen ‘system’ that could be driven from the main comp and switch images and move the screens from expression controls.
I needed juggling patterns that would be able to shift to allow additional balls, and a way to work the hands so they would be in the right place at the right time, they would need to open and close at the right time (considering that the frequency would need to shift dynamically based upon the number of balls. More, Dan is a juggler (and I am not) there was also the potential that the pattern I animated would not be correct (in the eyes of a juggler) so the shape of the pattern might need to shift. If I went at this using hand animated cycles, it could have been a terror as if I needed to move one thing, potentially thousands of keyframes would need to be changed, so I decided to animate both the hands and balls with expressions. This led to several other challenges as well…
- How to add balls to the cycle.
- Where to add them from as they would essentially be juggled by a bodiless entity.
- Once a ball joins the cycle, where that ball would sit in the cycle.
- How to make the balls evenly space when new balls entered.
- How to keep the balls from overlapping once in the cycle as this would be collision which is not cool if one is juggling.
Once all that was sorted out, then I needed to figure out how to have the hands match up with the balls.
How to have the hand know when it was done with one ball, and it was time to head back to catch the next ball.
Once matched up, then it was a matter of how to have the hand open and close to catch the ball. Being that the hand plate was technically behind the balls, I also needed to solve how to make it seem like the hand actually grabs the ball with fingers actually gripping the ball. I shot Dan tossing a ball over and over. I stabilized this footage. This result was a rotoscoped plate of the hand, and one of just the fingers which was parented to the hand. The expressions ended up being a couple pages long and slowed renders down a little, but my earlier fears were realized when Dan saw the original pattern and asked me to adjust the juggling cycle. Once the cycle was reshaped, all the updates were already done as it was all programmed based upon the shape.
Like most heavy projects, there were many things learned. These types of things are not easy, and can keep one awake at night. But they spark innovation, and we find more efficient ways of doing things. This was a seriously fun project and pushed me in many ways.
Dan did a talk on TED in which he used the ‘Danless’ version. You can find that video here…