A few days ago, I used the tools at the xtranormal site to create one of those silly digital puppet movies. I think they would prefer to call it something like a “3D animated movie generator,” but I’m sticking with “digital puppet” until they improve the quality of the generated movies. I realize I could have put their tools to better use, but the best videos I’ve seen are still pretty primitive.
(To their credit, the artists who created the digital content seem well aware of the limitations, and they’ve created characters and sets that play off the primitive style rather than clashing with it.)
All this reminded me of an emerging artform known as machinma, which is animated video created by using a computer game as a production studio. This makes more sense than it might sound like at first. Every modern computer game is essentially composed of three parts: The game engine, the digital content, and the game itself.
The digital content is basically all the bits and pieces you can encounter in the game: Every building, every character, every vehicle, weapon, exploding barrel, smashable crate, and item of clothing. These further break down into textures, shapes, forms, and sound effects. The game engine is the software that manipulates these pieces of digital content and displays them to the player. The game itself is built from the digital content and brought to life by the game engine. It brings the plot and the challenge that makes it a game.
If you can figure out how to run the game engine and load content without actually following the plot of the game, however, you essentially have a 3D digital studio. You can load scenery from the game and have your character walk around in free form. The game engine will handle your interactions with the environment and show you what it all looks like, with lighting, shadows, fog, and special effects.
In games that allow a fair amount of unstructured free-form play, people could make interesting movies just by recording their actions in the game world. These often demonstrated tricks within the game. For example, the vehicles in Halo are indestructible–as are the player models, although they can still die–and players soon noticed that the distance things flew when blown up was scaled to the amount of explosives used. This lead to a little experimentation:
And then there was the Star Wars game in which players could give their in-game characters some dance skills, which lead to abominations like this (the dance action starts at about 1:20):
I’m pretty sure that most of the content you see in these two videos was animated within the context of the actual game. However, with game engines such as Valve’s Source engine, you can pretty much completely take over the game, mixing and matching content from all over the game, or creating your own content from scratch and dropping it in.
One of my favorite examples of this is the Civil Protection video series by Ross Scott. The episodes follow the slightly surrealistic adventures of Mike and Dave, just a couple of regular guys who are stuck working as cops for the hated alien Combine that has conquered the earth.
This video–the most recent in the series–is built from bits and pieces of the Half-Life 2 world. But no cityscape quite like that exists in the game. Ross Scott has removed stuff he didn’t want, rearranged things, and populated the world with people and vehicles that fit his light story better. And I’m sure I would have remembered if there had been a donut shop.
The creature encountered at the end is from a previous game and never appears in Half-Life 2, although anyone familiar with the first game will not be surprised that it all ends in pain.
I can’t remember the name, but many years ago the Disney software gaming group released a game built around this idea. Players could set up stunt scenes on a movie set, complete with in-game movie cameras, and then try to perform the stunt for the cameras. This was way before modern game features such as general physics engines, lighting models, shaders, motion capture, 3D sound, and so on, so it didn’t look very good. Also, it was strictly built as a game. I don’t think the creators ever considered that people might want used tools like this to make real movies, with characters and plots.
Ken Gibson says
The “Halo Jump” video really takes me back…
Red vs Blue was one of the first times I saw well produced content using a video game as a production studio. Looking back at it, what made it spectacular was its ability to speak to the people who were engrossed in the subculture of that game.
The xtranormal site is different in that it is neutral, allowing a producer to break out of any specific game culture to create stories. The problem now is that they obviously didn’t put the time and resources into the world that most game companies do. I wonder if there is a market for high quality tools such as this yet. It would be nice to think so.
Glenn S says
Hi there.
My approach to “a more realistic Xtrnormal” is to apply the same modder’s ethic of traditional Machinima to Xtranormal itself. To see some of this in action, visit my Youtube channel.
Here is my 2nd most recent clip.
I think Xtranormal is the best tool out there for one-on-one dialogue scenes like this. And I cranked this out in a single evening. I just can’t get into Halo machinima where people are in helmets and emote at the same level as the Power Rangers.