Monday, January 18, 2010

The Great Machinima Debate

Lately there have been numerous blogs that are attempting to find an acceptable definition of what machinima is. This attempt is nothing new, as it has been occurring ever since "machinima" has expanded beyond the use of game engines.

It's a daunting task, to be sure, and I am not going to try to devise this definition. In fact, I have to wonder if, by current standards, an acceptable definition is possible.

See, it was a lot simpler in the past. Machinima was simply video capture of game footage, often directed and edited in a manner which results in the telling of a story. The key component was the fact that it used footage from video games.

Things obviously have changed since then, with several apps that allow the creation of movies in a 3d environment. This certainly falls under a recent attempt by Michel Nitsche to define machinima, which states:

"Machinima is digital performance that controls procedurally animated moving images in real time."

Although I think how he concluded this was very intelligent and well thought out, it's not really correct, is it?

Those of you who are familiar with how the battle scenes in the Lord of the Rings (or any other movie with very large crowds of people or creatures made over the last 10 here to see some familiar movies) were made should also be familiar with the software that was created to do it. The software is called "Massive" .

This software takes several different motion capture files and uses it to generate large groups that respond in a desired way. Here are some quotes from the software's web site to help get across what I am trying to say:

"The intuitive node-based interface of Massive Prime™ allows artists to interactively create AI-enabled agents. The Brain Editor AI toolset gives artists the freedom to build custom responses for the specific behavior they want to simulate, without any programming."    

"Massive's digital stunts are controlled by dynamics that pull from real motion capture data. Filmmakers can direct the motions and reactions they want with real stunt actors and then import these actions into a Massive agent to give direction and character to the agent's performance."
Now I am well aware that there are some differences in the approach of this software in comparison to the tools used to create machinima, but to the layman, Massive, in effect, is "digital performance that controls procedurally animated moving images in real time", and no one is ever going to honestly suggest that what Massive produces is machinima, at least not in the context we understand it., and not because it's "professionally" produced. It is because of the visual divide between what we see in these movies and what we see in machinima. Machinima as we know it does not have the visual polish, realism, or control that something made with Massive does.

I do understand that the intent of Micheal's definition was more about control of a character acting on an individual level, something that Massive does not do or is intended for, but as-is the definition allows Massive to fit in rather neatly.

So, unfortunately, Micheal's definition does not provide an adequate and unambiguous definition of machinima. Any definition that is eventually settled on has to be able to not only satisfy machinimators, but ultimately the public at large. The definition needs to be able to clearly explain to the layman what machinima is. It is for them, not machinimators, that this definition needs to exist. After all, we already understand, at least on a fundamental level, what machinima is.

Don't we?

I wouldn't be too sure. I remember when I first started making machinima, purists always said that machinima was "making movies with video games". Imagine my surprise when I started making machinima with The Sims, which is beyond any doubt a video game, and having these same people writing sims2machinima off, saying it wasn't machinima! These are generally the same people who refuse to count Moviestorm or iClone as machinima, and generally will only accept as machinima movies made with Half-Life, Quake, Halo, etc.

I know that doesn't reflect the general machinima population, but it does illustrate how daunting the task is of defining machinima. If machinimators can't universally agree on at least a basic definition of what machinima was, how can you ever expect to agree on what it is now?

A recent term to crop up is "Anymation", which is meant to be a catch-all to include machinima made with non-traditional methods (such as Moviestorm and the like). But this term isn't entirely adequate, is it? After all, the term Anymation would include anything Pixar makes, flash animation, clay animation, stop-motion animation, traditional drawn animation, etc.

None of these are what we are thinking about when we talk about machinima.

I think one of the ultimate problem starts at the source of this whole debate....machinima itself. See, machinima was never a proper term for what was being done.

Machinima is derived from machine cinema, and no one can doubt it is clever. But even on it's own terms it can't reconcile itself. Cinema made by machines covers just about any 3D animated movie ever made by Pixar, Dreamworks, and the like. 

So even since the very beginning, saying machinima was making movies with video games was not entirely true, and even if it were, in the current environment the term machinima is at best antiquated.

So that is where we stand. We have a general unconscious consensus of what machinima is and what it is not, even if we aren't sure how to say it. But how do we hope to find the true definition? 
Hell if I know. But what I do know is that the first thing we need to do is toss old conventions and terminology aside so that we can focus solely and what we are truly trying to communicate.

We're not re-inventing the wheel, but essentially, we are. At least, we're re-inventing how we look at it. We have to. At least that way we can hope to understand it enough to define it.