ArenaNet’s Chuck Jackman (pronounced as Scottish as possible) shows off the new dialogue cut scenes in Guild Wars 2 in one of the latest blog posts. Unequivocally, he started the doom of ArenaNet and Guild Wars 2 by having to summon eldritch gods to get the job done. It is possible that everybody will be insane/dead before a beta starts. Still, which do you think is scarier, Cthulhu or an art director that has probably already bested a few of those gods? Anyway, amongst the cries of the damned, Jackman writes:
We also use a layered and additive approach on the face. This allows us to animate a mood or emotion-based “face idle” and layer on the lip-synch animation as well. We can then drop additive facial gestures at the appropriate time in the dialogue. For example, we can have a character looking timid or frightened while they talk. Then maybe something happens to scare them and we can drop an additive flinch animation onto the character’s face, body, or both independently at just the right moment so that they react to what is happening to them in the scene.
I have to say that this sounds really cool, especially after watching some deadpan facial expressions on the heralded Skyrim. An official video of one of these new dialogue scenes gives the full impression. I find two teachings colliding from my untrained pundit’s eye: microexpressions and constant movement. Let’s explore both!
Long ago before there was sound, there was stop-motion. Willis “O’bie” O’Brien found that by moving an object each frame he could create imagined creatures that seemed to be real. O’Brien was not the first to do it, though, but he is seen by many as being the pioneer of stop-motion filmography. His trick was simple: move it each frame. Keep it moving. If it appeared to stop, even for a fraction of a second the trick would be lost. If anything just make a muscle twitch or a hand flinch. George Lucas actually tried to go one step further to “force” a blur in movement using his Go Motion technique.
MMOs have this same problem in a sense. That NPC that has to stand there forever, well in this day and age it would be considered a horrendous bug if she didn’t at least look around, shift her feet, and scratch her head every few seconds she is left alone. It seems like ArenaNet is taking this “keep it moving” principle wholeheartedly with the new scenes setup. Move their eyes, move their hands, fingers, arms, hair, it doesn’t matter. The more movement available to be layered, the larger range of emotions the choreographer is going to have when animating the scene.
If the choreographer can get that simple (KIMS?), yet super difficult principle down, the next step is adding involuntary movement (e.g., microexpressions). Non-verbal communication is arguably more important than any objective, verbal communication. Involuntary, non-verbal communication hits our subconscious. It sets the tone and mood and amplifies whatever the literal speech is trying to convey. It creates the harmony of speech.
One of my most endearing shows is Lie to Me (cancelled). I see the world a little differently now because of it; I am more consciously mindful of microexpressions and body language. They are additive… or possibly subtractive. A head-down shrug when answering a question might allude to uncertainty, but what about that shrug trying to hide a knowing smirk? Arms are the great emphasizers, when I move them to make a point, it’s more important than all the other blather surrounding it. I cross my arms to protect myself or to keep something in.
Let’s take a look back at “Cinematic Conversation B” from the blog article. Watch around the 35 second mark when the hero talks about his rogue-ish friend, except watch the Priestess instead. She raises her eyebrows just a little, possibly in surprise, when the hero says the shady guy is his friend, but then she cracks a small smile at the half-joke to check the supply cabinets. More bluntly, the hero dips his head in gratitude at around the 50 second mark. It could use another iterative pass or two: I don’t like how the hero doesn’t seem to react to the barrage of news near the beginning, and the Priestess seems to push her eyebrow movements to a drunk-seductive level at some points. Still I am being hypercritical for such a brief cut scene I watched multiple times. It is high quality for an MMO. The lip syncing alone even blows recent games like Skyrim completely away.
Just like scene cutting in cinematography is an art, so is this. It’s invisible art. The choreographer is going to be painstaking lengths to make it look like realistic communication by painting in a swath of hues of movement: apparent, constant, and hidden. Ironically, the closer he or she gets to perfection the more it is going to blend in to the entire Guild Wars 2 canvas. I can’t even imagine the disparity in work time vs. player view time for each cinematic. This seems on the level of painstaking.
I asked ArenaNet a quick question about the amount of production time per cut scene time, and Jackman was awesome enough to respond:
The ratio is not 1:1. The system works similar to a funnel (or Beer bong if you happen to be a part of the college crowd) where we front load it up with a number of animations at the start and then that is reduced to a tighter focus on the other end. Since the system is built to reuse animations in multiple scenes, the number of animations we are creating at the end is going to be much smaller than at the beginning. For example, let’s say that in scene A the player character needs to be confident. We then create one or two confident animations for each race and or sex and then use it in that Conversation. The next time a confident animation is needed, we are able to reuse the 1-2 we have already created rather than having to make a new one for each case.
He also hinted at the system specs for some of their modeling rigs. The processor is pretty amazing. It allegedly helps with production time. (This is a trade secret, so don’t tell anybody else.)
I suspect that ArenaNet is going to push this system as far as possible. It would be great to see some of the outtakes when layered animations and expression turn out differently than planned. Still, it is pretty impressive. I took time to watch a few other games’ cut scene animations, and I must say ArenaNet’s is stacking up really well. I can’t wait to see how they interpret asura confident or charr shy.
been brassing and laughing so long