Feeds:
Posts
Comments

Posts Tagged ‘Uncanny Valley’

I’m just ruminating a bit, spurred by a pair of excellent game design posts I read last week.

First, there’s Syl’s post about Why Storytelling in MMOs is Overrated.  I love her article, and I’ve wished for a long time now that MMO devs would ease off the reins and let players tell the story.  (Tangentially, Brian “Psychochild” Green’s work on Storybricks looks like a good step in that direction.)  The developer-driven narrative in these MMO things is a mismatch for the game design from the conception, and the devs seem to cling to their sense of authorship too much.  I can understand that, as a creative sort.  I’ve done a bit of Game Master work in tabletop RPGs in my day, though, and ultimately, the game always seems to run better when the players feel like they are in control.  The GM has to keep everything together, but player agency is the heart of games.  Even if it means they do things the GM doesn’t anticipate or even desire.

Second, there’s this gem from The Rampant Coyote, From Whom Much is Given, Much Is Demanded.  The discussion there about graphics and how cutting edge technology tends to create absurd demands rings true to my experience both in games and when I got my college degree in computer animation.

Today, I stumbled across this interesting tech demo from Activision.  It’s, well… creepy.  It’s very impressive, but it’s still not quite right.  Here it is on YouTube:

That Uncanny Valley looms large.  This is one of the huge dangers of chasing the tech edge.  Yes, in theory, with enough money, processing power and artistry, it’s possible to make artificial life that can pass for the real thing.  The cost is huge, though, and that Uncanny Valley is big.

Also, most importantly, it’s relatively easy these days to make artificial life look good in a still frame, but the real test is when it moves.  Motion is ridiculously hard to make, and exceptionally easy to break.  We have an instinctive understanding of how living things are supposed to move and behave, from physics to biology to exceedingly subtle emotional cues.  (See: Lie To Me, Sherlock Holmes, psychopaths, etc.)

This, perhaps more than anything, is what I really dug into when I was in college.  It’s at the heart of the Disney films I always wanted to make, The Illusion of Life that really makes animation work.  (By the way, I highly recommend that book if you have any interest in animation, along with a more recent tome, The Animator’s Survival Kit.  If you can only digest those two books, you’ll be a long way to understanding the core of animation.)  Ultimately, it’s possible for a skilled animator to make a broom or sack of flour (or even a paper airplane) seem more alive than the latest Final Fantasy CGI characters.  Or, as I noted over at Syl’s place, animators try to be conscious of the silhouette, making sure it’s readable at all times.  You can get a lot of mileage out of just the silhouette, as the XBox LIVE Game LIMBO shows:

And really, a lot of what gets communicated has to do with what isn’t seen.  (For a funny riff on this, there’s this take on what LIMBO might play like when you can see more information… but again, selective reveals are what sell the humor; it’s the juxtaposition of what you expect vs. what is “really” there that makes it humorous/scary.)

If you haven’t seen Paperman, go watch it.  Seriously, go watch it and then come back.  (Or watch the embedded one, sure.)

And then watch this, a video about some of the tech behind it.

So, for a relatively simple-looking bit of animation, there’s a lot of tech under the hood.  Some of it is obviously CG, at least to me, having spent as much time as I have watching and producing art and animation, both traditional and computer-assisted.  Still, there’s a lot of work going into this… and it’s all to make a stylized bit of art.  As with the style of The Incredibles, stylization goes a long way to making something play well.  It short-circuits our instinctive evaluation systems, and the errors in animation that pop up are kind of fudged away, filed in mental gaps that we don’t wind up caring about, largely because we have already internalized that these characters are not real, and we don’t expect them to be.

This is how we perceive motion in film and animation in the first place, per the Persistence of Vision theory.  The 24 or 30 frames per second that flicker by don’t cover the infinitely reducible time frames that reality can be split into, but they happen fast enough that our brain accepts them as continuous enough to be believable.  In fact, sometimes less information works better, as evidenced by some of the kerfluffle around the new-fangled 48FPS The Hobbit movie.  All we really need to know is enough to fool our brain into accepting something as real or believable, and then let our imagination and subconscious do the rest of the work.  Perhaps we could call it a “Persistence of Cognition” theory when it comes to storytelling and lore; the reader/viewer invests headspace in imagining the fictional world and how it works, or how they could work within it.  It’s all about leveraging the strengths of the end viewer/reader/player, making them a partner in the experience.

This is why a lot of the high end stuff fails.  It tries to do too much.  Our brain takes it at its word, holds it to a higher standard, and finds it lacking.

Most of the time, especially with art, story and anything that really hinges on the viewer getting emotionally involved and engaging the imagination, less, to a certain degree, actually is more, simply because you’re letting the viewer breathe and take a bit of ownership, which tends to be a multiplying factor in the efficacy of a presentation.  It’s part of that “willing suspension of disbelief” that’s so important to get people to buy into what you’re doing.  There really are reasons not to go into obsessive hyperdetail, not only because it’s a time and money sink, but because it’s also less effective.

Artists tend to understand this instinctively after some practice, since it’s entirely possible to put too much into a piece of art and thereby ruin it.  Hinting at detail is often far more effective than rendering it.  Even Daniel Dociu’s incredible art, which tends to look really complex, is largely suggestive, relying on the viewer to infer a ton of detail that really isn’t there.  Just look at the actual brushstrokes in one of his pieces and compare it to what you thought was there at a glance.  Dociu is a master at implying complexity.  He’s making your brain do the heavy lifting.

Similarly, as any avid reader can tell you, “head canon” and “mental visualization” of words on the page can never compare to a moviemaker’s craft.  They simply function differently.  That’s a good thing, and creative types really need to leverage the supercomputers in viewers’ brains to do a lot of the creative work for them.  It takes trust, and knowing just what to imply and what to make explicit… but there’s a lot of strength in letting the viewer in on the process, even if it’s only on a subconscious level.

They will fill the gaps, if you can learn what to leave up to them.

Read Full Post »

A little knowledge goes a long way sometimes… and maybe not in the direction we’d like.

As I’ve noted before, my college degree is in computer animation, specifically geared to film making.  In the course of earning that degree, I learned a lot of film making tricks and tactics, as well as the extra layer of tomfoolery that computer graphics permits.  As such, it’s very hard to watch a movie these days and not see all the little hacks and cheats.  I can’t help but see behind the curtain because I’m so familiar with what goes on back there.

I have a similar problem with games.  Since I work in making games, usually creating, texturing and animating 3D models, and I’m very familiar with the industry, I see all the little tricks that other game developers use in their games.  Even if I’m willing to suspend my disbelief and just be entertained, it’s a reflex to see, catalogue, and examine how things work, and perhaps more importantly, how they don’t work.

Similarly, since my degree is technically a Bachelor’s of the Fine Arts, and I’ve been an artist who studies art for a long time, I have a lot of experience with art, both creating it and in analyzing it.  I see art problems way too easily.  Even my hobbyist knowledge of astrophysics, physics, math and science makes some things hard to swallow, like the awful science in that recent Star Trek movie.

This is the effect underlying the Uncanny Valley effect.  Y’see, we’re all experts at being human, just by virtue of, well… living life.  When we see something that doesn’t agree with our experience, it just seems wrong, no matter how well-crafted it is.  It need not even be conscious; we notice the inherent wrongness whether or not we want to, and it colors our experience.  Even something like basic kinesthetics can be thrown off, as is the case with the Kinect motion sensor control system for the XBox.  The functionality is wrong compared to what we know so well, and it just doesn’t work.

So when we see something like this, where a psychotic nutjob’s murderous actions are blamed, in part, on video games, as gamers, we cannot help but shout:

OBJECTION!

We know enough about the reality of games from our own expertise to call “shenanigans!” on the media narrative.

Of course, there’s a flipside to this.  If we don’t have personal expertise in a topic at hand, and don’t want to bother informing ourselves about it, it becomes very easy to just go with the flow, accept fallacious authority, and accept whatever we’re told, especially if it’s something we think we agree with and fall prey to confirmation bias.  If we want to hate someone or something, we’ll find reasons to do so.  If the narrative suits our taste, we’ll happily ignore facts.  We embrace ignorance and live in our own little perception bubble, because we’re happy there.  Manipulative agenda-riddled media is more than happy to play along.

This is certainly obvious in politics and the so-called “mainstream media”.  This is one reason why blogging is changing the world and why it’s important to protect in the face of political opposition and Big Brother control; the “news” networks get called on their lies and matters of public policy can get a bit more transparency with concerned citizens involved.  Nothing quite dies on the internet, and it’s increasingly easier to do a bit of research and do a little fact checking.  Of course, even then, so-called “fact checkers” are usually biased, too.  You really have to go do your homework and proper research if you care about something.

Remember the murderer who played WoW?  When there’s a causal link implied by shoddy reporting and poor courtroom procedure, and you know the argument is pure crap, you don’t trust the narrative, and you are right to be distrustful.  You know better, no matter what the talking heads on the magic light box try to tell you.

The really crazy part is when you see through the curtain sometimes, but decide to let it slip back into place later, say, if the same media outlet reports something you want to hear.  They are no more trustworthy than they were before, but this time, since it’s something you agree with, it doesn’t matter what goes on behind that curtain.  The narrative is what matters, not the truth.

WoW subscription numbers down 300,000?  The game is finally dying!  Thanks for the brave reporting, guys!

WoW subscription numbers steady next quarter?  Must be a statistical blip or someone cooking the books.  They are desperate to show they aren’t dying!  Lousy lying media!

So what?  Just sayin’…

Trust, but verify.  Understand your own bias and get past it… at least, if you care about truth.  Sometimes, we just want to be lied to.  Being able to swallow the lies, benign or otherwise, certainly makes it easier to be entertained.

Read Full Post »