Author: Nathan Willis
The creative forces behind the Orange Blender Project describe it as an opportunity to showcase the power of open source — in particular the Blender 3-D modeling application. Over the course of the last year the project raised funds through arts grants and pre-sales of the DVD, and used the money to sponsor additional development work on Blender, adding the features they needed for the project.
As such, Elephants Dream is a bit of a mercenary endeavor — conceived not as an artistic statement, but to prove the trustworthiness of a particular process and tool chain. That puts it in a curious position: is it flatly outside the realm of straight art, and thus unfair to judge on creative merit alone? Or does it have to succeed as a creative work in order to make its intended point about free software?
On the aisle
| A scene from Elephants Dream. Click to enlarge |
The film itself is plotless; a sequence of seven scenes depicting two characters as they wander through an enormous machine that responds to their thoughts and words to varying degrees. Each scene is visually exciting in its own way, even if they do not connect in a coherent storyline.
In this regard, Elephants Dream feels more like a demo reel than it does an animated short; it is as if the creators are the older character Prooq, dragging the viewer (represented by the younger character Emo — whose name seems account for 60 percent of the dialogue) through a guided tour of What Blender Can Do Land. Had they had twice as much money, I get the feeling we would have seen simply twice as many unconnected scenes showcasing visual effects.
But those visual effects are certainly fun to watch. An elevator scene showcases some large and loud science-fictiony scenery reminiscent of the Close Encounters ships. In contrast, a telephone scene is quiet and weird like something from a David Cronenberg movie.
Indeed, visually there is no weak point in the film. The lighting is spot-on, the cinematography is professional. The characters even look good — they are cartoonish in anatomical design, yes, but the eyes, skin, and even the notoriously tricky hair look as good as anything you will see on the big screen.
The characters’ movements, though, are less refined, and the quality of the movement varies from scene to scene; sometimes quite fluid, at other times more akin to wooden marionettes. At times, Emo and Prooq seem to be waving their hands and twisting their necks just for the sake of being animated. I’m not sure if there is a term for this phenomenon, but it does seem to afflict 3-D animated characters more than their 2-D and live-action counterparts.
Since there is no storyline to Elephants Dream, it may not be appropriate to critique the acting and writing. On the other hand, some interviews with the creators lead me to believe that they do think that Elephants Dream has a storyline, and a coherent one at that.
Suffice it to say that the second time through the film, I wanted to turn the sound all the way down. If there was a story in there somewhere, it fell victim to the editing process. Given the “open movie” moniker that the project applied to this short, I think it would have been better served had it solicited scripts publicly, then selected the best story.
In the DVD drive
| A Blender file from the Elephants Dream DVD. Click to enlarge |
Of course, what really makes Elephants Dream different from other animated shorts is the project’s open source and open content policy. Alongside the movie itself, the DVD contains all of the music tracks featured in the film, and all of the data used to generate the scenes: textures, images, 3-D models, maps, animatics, Blender plugins, Python scripts, and sounds.
All of the data (excluding the various logos and trademarks) is licensed under the Creative Commons Attribution license. In addition to being on the DVD, it is also available online.
Though much of it was created in-house by the Orange Blender team, some content was solicited from outsiders via the project’s official blog. Conversely, the project was very open on its blog during the design and production of the film, often posting snippets, teasers, samples, and so on while writing frequent progress reports.
In some cases, there are multiple versions of each file, letting you see changes made over time. Also included are a set of cut scenes, in various stages of completion; perhaps it is only a matter of time before someone completes them and releases their own cut of the film. And given the license terms, someone could come along and recut Elephants Dream entirely, or lift Emo and Prooq to put into another movie. That’s the beauty of it.
But that is only half the story. One of the Orange Blender Project’s stated goals was to use the production of the short to increase the quality of the open source software it utilized. Over the course of the past year, the project has fostered a significant number of enhancements and extensions to Blender, including rewritten character animation, fluid dynamics, particle systems, rendering pipelines, and compositing. The Blender Foundation mentored a number of projects in 2005’s Google Summer of Code.
All of the enhancements have been added to Blender as part of its regular, ongoing development. A list of the changes made during this time can be found at the Blender Web site.
One hiccup in the project’s “all open source” battle cry was the sound programming, which was eventually processed using the proprietary tool Reaktor. Orange Blender’s Ton Roosendaal told Wikinews that the decision was a practical one, as the team chose to go with outside experts for the sound design. While they would have preferred open source, they did not mandate it.
Children of the Open Movie
Ultimately, Elephants Dream is most successful in those elements that lie closest to Blender’s core competency, and decreasingly successful in those further afield. I think that does count as proof of Blender’s prime-time readiness — and, by extension, that of the other open source tools.
What’s more, the improvements made to Blender and the other open source apps during production are signs of the strength of open source, not weakness. The open source tools were adaptable, flexible, and fixable as challenges arose; even with the limited (by movie studio standards) budget, the tool chain was changed and enhanced while production proceeded at full speed.
That makes the project a success on one of its stated goals — demonstrating that open source is a powerful tool in the professional creative arts. So, the big question now is: will there be more open movies in the future? It looks like there will. DVD sales seem to have generated enough funds to pay the bills, and the Blender Foundation is already talking about other projects.
