What strengths and weaknesses does Machinima have right now? Well, we can learn a lot by attempting to copy other media.
I ran across this parody of the Neon Genesis Evangelion intro sequence yesterday, via WoWInsider, and found it absolutely fascinating. Let me start by linking both the parody and the original:
So - how do the two compare, and why, and what can that show us?
Firstly - the Machinima version looks great. If I saw that as an intro sequence to a Machinima piece, I’d be intrigued and psyched. I’d be expecting a very high-quality production. And that tells us some interesting stuff about how much sophistication it’s possible to get out of a Machinima piece - that by pastiching the work of another medium it’s possible to produce such a good piece, better than 95% of the title sequence efforts out there.
(It also heavily implies that we should be looking to Anime for clues on how best to use Machinima. I’d have to say a big “yes” to that. Machinima bears a lot of resemblance to limited-animation styles like most Anime, and they’ve got a whole lot of tricks over there.)
Secondly - whilst it looks good, there are large parts of the Machinima version that just don’t work. And the reasons why they don’t work are really good indicators of the problems that Machinima’s having right now, and potentially the ways it can get to the next level, and what we should be looking for from tool developers.
Firstly - the various face shots don’t really work, at least for me. In particular, the first shot, of the lead character (I assume), works really well in the original, but not at all in the parody. That’s purely down to the artwork - the lead character’s image in the original tells us a lot about his character, and the subtle animation helps to bring him to life. But more than that - the character looks involved, vulnerable, readable. We can empathise with him and feel for him. By contrast, the WoW character has the same blank, dead-eyed stare that all WoW characters have, and is clearly a generic WoW male - it doesn’t suggest anything, it doesn’t imply an inner life, and that stops the shot from working.
Even worse is the silhouette shot that overlays on that, and this really zeroes in on one of the fundemental points of Machinima right now. The Evangelion character’s image strongly suggests character. He’s a small guy, his stance, with the cocked head and the weight on one foot, suggests that he’s unsure of himself, but curious. By contrast, the WoW character - is just a WoW character. There’s no information there at all. No pose, no emotion, no accessibility.
By contrast, the shot that works is a few moments later - the female seated silhouette. It’s strong, it looks relaxed, alert, curious, and sexy. (And well done for that! It’s long been my feeling that “sexy” is something Machinima could do with a lot more of - but that’s a different blog post.) Interestingly, the message of the shot is completely different to that of the equivalent shot in the original title sequence, but it still works.
And so on. I could go through the rest of the parody, but the point is the same. I’m not criticising the Annoying Quest guys - this is a great piece - but using it to point at some problems common to most Machinima right now.
It’s very hard to get Machinima characters to suggest anything right now just from their appearance. Never having watched Evangelion, I already have ideas about the characters as people just from the Evangelion title sequence. It’s something that Machinima is sorely missing.
What Machinima is lacking right now, beyond whiz-bang effects, beyond terrain and realistic rendering and enormous banks of assets, is the ability to make character expressions and particularly body language and silhouettes - two of the key instruments for the traditional animator - emotional. It’s a hard problem.
I can see two solutions that would be practical to non-multimillion productions. Firstly, motion capture of face and body will solve a lot of these problems. Better rigging will be enormously helpful for facial animation anyway - look at Lit Fuze’s fantastic work with the Half-Life 2 engine to see the results of a talented crew combined with great facial rigging, and Moviestorm’s getting there too - but for body animation, motion capture is probably the most practical solution, and it’s set to get very cheap.
Failing that, the kind of “emotional animation” work that people like Ken Perlin and the NaturalMotion guys are doing might well help. No-one’s yet unveiled any kind of animation control system that will allow characters to suggest emotion through their movements, but if it comes, it’ll be a huge leap for Machinima.
I’ve been playing the new console game Mirror’s Edge recently, and the closing cinematic there (near the end, warning - spoilers) is a brilliant example of what could revolutionise Machinima if it was available to normal users. It’s very simple, really - after a horrible event, one character puts her arm around her sister, and they both look out over a city. But the silhouettes, the emotion that’s conveyed by the protective stance of the one and the huddled stance of the other - that’s storytelling. And that’s done in Machinima. It’s doable - we just need to bring it down to the level where non-multimillion productions can access it.