Mechanical Magic

Solomon Little

The last ANC project for any animation student undergoes is to create an animation that they feel is on the limit of what they can produce. We are to thoroughly investigate and explore our skill-set and polish a single thirty second animation to a point that our inherent perfectionist is satisfied.

I may of gone a bit further

This is me - Solomon Little

Welcome back to the Weird Side

Mechanical Magic is short animation that shows a robotic arm re-activating in an old warehouse and doing what it does best, throwing a fireball across the room. It was created as an proper exploration into using full simulations as a centrepiece for an animation as well as optimising and controlling fluid dynamics.

So why this?

Well, I have always had a fascination for both magic and machines, and a rather endless hunger for media that contained both, so for this project, I decided to see what I could do. With a recently found interest in simulations, I approached the project from the perspective of attempting to make the fireball into the most impressive visual spectacle that I could, and to attempt to pull back the focus from items around it to better show it off.

The Lab

The first section in the animation was the pan through the lab, to both establish the shot, and to help convey the more industrial take on the magic.

The focus on this room was mostly to get the props themselves up to scratch from a model perspective, as well as having proper texturing. It also gives a place for the camera to snap back to, to help inform the audience that a aspect of the fireball has changed.

This part was heavily inspired by works such as the WarThunder animated trailers, and works by Dima Fedotof.

The Warehouse

The main room was created to have a similar aesthetic to a empty testing warehouse, as I felt it best fit the more utilitarian feel I was trying to achieve with the animation. It also allowed the focus of the camera to be easily divided.

By making use of a massive, empty room, I had the full ability to play with a variety of ways to set up the fireball movement, particular in the beginning when I was still unsure of how exactly the fire was going to move. This also allowed me to strategically place lights to help minimise the total spill, to better show off the colours that can come with a more minimal focus on the lights.

The ARM

The robotic arm was an interesting challenge, and ultimately one that I had to compromise on. While originally I had planned to make use of a articulated hand that I had modeled and rigged, I couldn't fix some of the bone constraints. Instead, the summoning hand was created.

This part had a few interesting parts. The way I had set up the animation for most of the arm was pivot animation held in place by parent-child relationships. While this made animating most of the arm really easy, it had some unexpected effects for the hand itself. The main problem was that when the base of the arm pivoted around, it would drag both the mesh and the bones at the same, causing the mesh to deform, and removing the hand mesh from the relationship caused pathing problems. So in light of all of this, I instead decided to model out an alternative hand that could also be pivot-animated.

THE FUN STUFF - a little bit of fire

And now we have the meat of the project itself, the fire simulations. This was quite a bit of fun to work on, and it has created some really interesting behaviors and constraints that made me rethink the project multiple times to get a result that I liked out.

Right, where to start....

To begin with, I was attempting to use Houdini, but that quickly became not worth the trouble, as I am not comfortable enough with the level of coding that I would need to be able to actually create the variables that I wanted, so I swapped to Maya and the bifrost engine. Using the node system that was in that, I was able to achieve a similar level of control with significantly less complexity. After this, I spent a week or so familiarizing myself with how the fire behaved. This lead to quite a few cool scenes of where the fire behaved in unusual ways.

This then led to the single biggest hurdle of the project, the speed of the fireball spawner. Initial I had set it to move way to fast, and as a result, the fire could not build up enough to actually look like a fireball moving. I then slowed it down, and while the impact look amazing (see below), the actual way that the fireball moved across the warehouse was severely lack-luster.

Fortunately, there was a few ways around this problem, and by scaling back some of the fire, I could get it down a reasonable size to help with actually loading the thing.

Collision 1

COLLISION 2

collision 3

💬 Live Chat with Solomon Little now to find out more about his project.

"Hi everyone! Thanks for checking out my page, please feel free to contact me if you have any questions about my work"

Contact:

My email - Thesamonman@gmail.com

Report abuse