Hi everyone,
I'm writing a paper on animation and got to the following issue.
When I'm watching animated films the ones where they relied on motion capture seem to be missing something. While adding realism to your animations usually is a good thing those characters appear to be missing a certain amount of "personality/life" making it harder for me to properly relate to them.
Could you help me pinpoint what exactly causes this? Does this have to do with acting for animation that is different than for life action, or is it the level of subtle "noise" in movement?
Or am I totally thinking in the wrong direction... All help is more than welcome!
P.S. I heard they dont use any mocap at Pixar (and PDI?) at all for their movies. Is this true, and if so why not? Does this has to do with the visual style of their work or is there a more fundamental/principal reason?
I'm not an expert by any means, but by extensions of common sense and what would be practical, I think both questions can be answered by the fact that broad overacting, exaggerated motions, etc... are more entertaining and more easily understood to represent the actions they are representing.
With motion capture, you literally -are- adding realism, but most of the animation I've seen is successful (both monetarily and from an interest standpoint) because of how unrealistic it is. Kicking a football doesn't make a motion capture person do 40 flips in the air, etc..
As far as Pixar and PDI, there are workers on these boards who could ultimately clarify, but I think the answer is both visual style AND necessity for artistic integrity in making truly animated movements...
There's an article on Slate which addresses this issue.
http://slate.msn.com/id/2102086/
Hi guys,
Thanks for the posts, this is pretty usefull but i'm not done yet.
I did see the something like the slate article a while a ago (only that one was solely about robotics) but hadn't linked it to this yet.
But if this was the main problem than why does it seem to work for some characters? I thought Gollem was done pretty good, although there the creepy-thing might be in favour of his character. I know this was mostly keyframed as pointed out by jason but is sure does LOOK like motion-capture. The Hulk however, while technically no less impressive, seems to be missing those emotional qualities gollum does have.
When looking at the trailer for the upcomming movie "the Polar Express", although the second trailer looks a lot more promissing than the first, I cant help but wonder why it looks so bad... It hasnt got the hyperreal look so that isnt it. Their using this new system to capture the motions where they can get both body and facial capture of multiple actors at the same time enable-ing the actors to have better interaction with eachother. So what went wrong?
In relationship to ScatteredLogical's post.
A Watership down (1978) is pretty realistic animation in my memory (been a while since I last saw it) but that certainly wasnt less interesting. I dont think only cartoony animation is worthwile though i do agree Mocap isnt much use in that area. As i see it slapstick-style animation isnt that often used anyways at the moment. Even in movies like Finding Nemo they are going for "realistic" motion and images. For example i see distance fog, caustics effect from the surface, particles on currents in the water, aswell as the propulsion of the fish and so on.
If you got anything that might be of use in my paper please feel free to add. All rants appreciated :P
PS
I'm still looking forward to some Pixar/PDI animators explaining whether its true they dont use any mocap at all and why. So if you happen to have some time... please enlighten me.
of course if we're going down that road, there's a vast difference between realistic *rendering* and realistic animation. Animation in my opinion is about idealized movement, its about what we need to see to read something clearly, what we *expect* to see when something moves, not what is necessarily reality. I mean, people don't usually present their hands before they pick something up, but it definitely helps the audience understand that this character is picking up that object.
Ender
You have to remember the Slate article which states that we have problems with HUMAN characters that look 99% realistic. Gollum and Hulk are inhuman monsters (maybe 80% human), so we're not in the process of being fooled into thinking they're human. Hulk was missing emotional qualities partly because his character had far less emotional range than Gollum.
You can also check out this article:
http://www.arclight.net/~pdb/glimpses/valley.html
Good point Ender about the difference between the rendering and the motion. Although very interesting, photorealistic rendering and it's implications on relating to characters was not was i planned to write about. Could I conclude from your post that if the actors used for the capture would show more anticipation and acted in a more theatrical manner this would make the "animation" better? Or is the focus on Idealised, making animation with less non essential "noise" preferable. (but leaving the secondary actions)
Also I didnt get the idea a lot of people here are in favor of using mocap as an animation tool. Is this just because you dont consider this to be worthy of the word animation, since you dont get full control over the character? Or is it something else?
There's got got to be some good sides to it seeing how polular it is, especially in games, vfx, and animated series. Or is it just because its fast and therefor cheaper than having a team of animator work for weeks on an episode?
I dont think much people here know this example. It's a dutch TV series called "Cafe de Wereld" which aired every night, commenting on yesterdays news. In order to be able to do so the entire show is based on mocap and rendered out on a daily (?) basis.
You can see an episode here and find some more info on MOTEK 's site.
What do you people think about this serie in particular or mocap series like Roughnecks in general?
I haven't used mo-cap beceause I've never had access to facilities to try it out. But I have had an actor act out an entire script and used the video reference as a template for animating a cg character, though I hand keyed all the animation. I don't think there's anything wrong with that, but of course I had to adjust the animation and get rid of some of the motion noise to make it seem less jittery. Perhaps as you were saying, if mocap actors exagerrated their performances a bit instead being subtle the result would be less jarring? Interesting hypothesis.
Definitely no mo-cap going on at PDI/DreamWorks (that's from experience), and I know for a fact there's no mo-cap going on at Pixar (as should be obvious, even without insider info).
A few reasons. Mo-cap is no good for exaggerated stuff. Mo-cap is cumbersome and in many cases slower than having an experienced, talented animator do the same shot from scratch (i.e., there is no real time or money savings, unless your goal is second-rate, weightless animation). Mo-cap can't really be done for characters that don't have human proportions (unless you don't care how wonky they'll look). Even the very best mo-cap system can only convey the most rudimentary information about what a person is doing with their face (see Polar Express as a recent example). Most animation, even 'realistic-seeming' animation, requires significant exaggeration to look 'right,' and mo-cap is an impediment to that. There's more, but that's off the top of my head.
i am not sure how much of this will be mere regurgitation of the same thing, but my two cents would be that mo-cap is much too restricting. there is no creativity in mo-cap, at least on the side of an animator, or 'clean-up' artist. the actor provides the movement and the tweaking of that movement is extremely difficult without completely deleting keys and key framing what might look or read better. but then what would be the point of using mo-cap? it just seems that more dynamic poses and gestures can be achieved through key framing rather than being restricted by what the actor gives us. i agree with others that animation tends to need exaggeration in order to read clearly... actors can't always give us that.
"i love the graph editor"
[ChrisMagovern.com]
and visit me on myspace...
If we use mo-cap so no need to animators, just go on,bring some actors and do a real movie...right??...an animator is an artist...so where is the art of animation???...i don't agree with mo-cap..but of course u can have references from acting, expressions, moving in general (as pictures,movies,or anything) and u exaggerate and animate them with your own way....THX
If you want to see the fallacy of mo-cap with animation look at how rotoscoping was handled in the past.
Animation is MORE than just moving about on screen, Its timing and using certain tricks of same to hit the audience emotionally, albeit below their perception range.
Live actors and m-capturing cannot do that without skilled animators augmenting the scenes.
Example, a finger point with a bit of flourish on the end of the getsure, where the figure "snaps" as it points at something needs a bit of over-extension added by an animator. The actor's physiogomy and the mo-cap system isn't going to be able to capture such a gesture in its fullness. The animator provides that bit of flourish so the pose will read, and read emotionally because the audience will not see it consciously--unless they've been trained-- and it would be far too restrained a gesture otherwise.
That's the REAL Illusion of Life.
--Ken
"We all grow older, we do not have to grow up"--Archie Goodwin ( 1937-1998)
Hi, it's been a while since my last post but I had to go without my home internet connection so I wasn't able to post any sooner.
Thanks for all your reply's and please have a look at the following.
I did have a look at rotoscoping and while it was first seen as a wonderful new tool, producing always fluid motion, it was then discarted as being non artistic.
When watching Snow White (1937) the human characters like snowwhite, the prince and (in a somewhat lesser extend) the queen all seem so much less vivid then the dwarfs. This difference in performance is ofcourse also because of the limitations on their abillity to portait a living human and the difference in presentation. The dwarfs being charicatures and the humans fairly realistic.
Fairly recent there was a film depending heavily on rotoscoping, Waking Life (2003) While I must admit, I couldnt get myself to watch the entire film, Waking Life is also being claimed to prove rotoscoping is at the least an interesting tool in the animators toolbox. I heard people state it should have gotten the Academy Award nomination instead of Jimmy Neutron (2003).
From the previous reply's I think I may conclude that the areas motion capture seems to fail most are the lack of exaggeration and control of timing.
When I was just reading "Acting for Animators" I came across the idea in there that the difference between an actor and an animator is the way they come at the subject. As Hooks puts it:
"Animators are oriented towards what stage actors call results." An animator is concerned with how to move to indicate a certain emotion (how many blinks occur in an excited moment) while an actor is taught not to play results of his characters inner motivations. Instead he is searching for his intentions which result to actions, played in pursuit of objectives. The appropriate facial/body movement will then occur naturally.
Could it be that this difference in approach is what shows in motion-captured animation in comparison to key-framed animation?
Not showing the required amount of detail to reveal the impulse behind the motion? Or is it just motion and not what we're looking for namely, a performance?
Although virtually all the previous reply's have been very helpfull I havn't yet discovered what makes hand animated sequences so much better.
When looking at the principles of animation as layed-out by the disney studio way back in the golden-age they just about all seem to be in some way applicable to motion captured scenes aswell, as they are mainly about 'making things move right'. Yet, as previously already stated, animation is more then making things move. Is it perhaps so that just following all those principles doesn't (necessarally) lead to good animation?
That Acting for Animators quote you put in is fascinating. It has become the first thing I've ever printed out and put in my reference from this forum. Hmm. I might have to go buy that book...
Ender
I think it's probably pretty obvious to most professional animators. Some of the fundamental "principles of animation," which make it so interesting to look at, are avoided when you take directly from life: squash and stretch, exaggeration, and appeal.
How interesting would an action or comedy film be if they endeavored to make it realistic, rather than exaggerating scenes and characters?
Why is Modernism so much more interesting to view than realistic paintings?
Of course, if you're trying to composite a believable CGI human into a live-action film, motion capture is very useful. If you're making a cartoony feature like Shrek, it's not as useful.
erm...of course Modernism being more interesting to look at than realism is a matter of opinion...
So is motion capture vs. "hand animation."
Being animators, I felt it was pretty safe to assume that most of us prefer the inventive and the imaginative.
I dont seem to completely grasp this yet...
I think we can all agree that for certain situations it is useful and for others it apparently is not. There is for some reason a difference in perfomance for live-action and for animation. The way I look at it, animation is, at least in that area, closer related to theater than to cinema.
I sometimes get the idea people(animators) just want to say mocap is no good, this doesnt help much. Although I'm not so fond of it myself i'm looking for reasons.(I only have limited experience with mocap.)
The things I'm hoping to discover for myself are the reasons that lie at the root of the mocap problem and how to create better animation with (or without) it. Hence my last question in the previous post.
When looking at a lot of productions they hire athletes instead of actors or even worse did it themselves. But what if, in order to get a cartoony animation for example, we would have strapped Charly Chaplin in a mocap suit?
In my expectation he should to be able to deliver what is needed. When you watch those movies he is virtually a cartoon. He's got exaggerated movement, great timing and appeal. What makes that this would still not work, or would it? (besides the fact Chaplin is no longer among us...)
Maureen Furniss, editor of Animation Journal, Savannah, Georgia
http://www.animationjournal.com/
thanks harvey! this look promissing....
I'll read it tommorow. Couldnt find it on animationjournal but found it elsewhere. For others interessted you can read the whole article here:
http://web.mit.edu/comm-forum/papers/furniss.html
P.S. your new avatar looks much better! :)
Thats funny. That's a teacher of mine.
I can get you in contact with her if you want...
Ender
hey everyone,,,i don't agree that mo-cap is cheating because it's a technique that an animator use to finish work more quickly (technology is playing a role here)...Exactly like 3d animation and 2d animation...in 2d u must do inbetweens..but in 3d u must'nt...so...but i'm still saying mo-cap isn't a style...Only a technique that i will'nt like to use it.. :) ..thx
There are also a couple of good articles on AWN.
http://www.awn.com/mag/issue3.11/3.11pages/kenyonrosenthal.php3
and Puppetology: Science or Cult?
http://www.awn.com/mag/issue3.11/3.11pages/degrafmotion.php3
I guess the reason why mo-cap doesn't work sometimes, is because we don't look into the screen looking for the same stuff we see in real life. Look at motion pictures, with real people, and see how it becomes necessary to suspend the actors with wires, use slo-mo or speed up the shot, or create miracles in edition. I assume no one believes tht a Karate actor can jump 3 meters, or that Arnold can fall from a plane onto the roof of a car and just mess his hair, or that the cool fight scenes frim Matrix were made in real time.
I guess the reason why mo-cap doesn't work sometimes, is because we don't look into the screen looking for the same stuff we see in real life. Look at motion pictures, with real people, and see how it becomes necessary to suspend the actors with wires, use slo-mo or speed up the shot, or create miracles in edition. I assume no one believes tht a Karate actor can jump 3 meters, or that Arnold can fall from a plane onto the roof of a car and just mess his hair, or that the cool fight scenes frim Matrix were made in real time.