A New Age of Animation(theatlantic.com)
theatlantic.com
A New Age of Animation
http://www.theatlantic.com/entertainment/archive/2016/05/a-new-age-of-animation/483342/?single_page=true
8 comments
I've noticed that many official k-pop videos have English, Japanese and Chinese subtitles, a clear sign that they are reaching out to fans oversea.
However, most J-pop isn't even available on YouTube.
I'm surprised to hear that daisuki.net is a free site. Have been watching rips from there.
However, most J-pop isn't even available on YouTube.
I'm surprised to hear that daisuki.net is a free site. Have been watching rips from there.
> But one company is finally helping Korean studios go fully digital, and the impact for animation in the U.S. could be significant.
Does this not smack of a PR piece?
http://paulgraham.com/submarine.html
Does this not smack of a PR piece?
http://paulgraham.com/submarine.html
It must have worked even better than expected, because the featured company's website seems to be down.
> Korean animators can draw 240 pages for a single 20-second scene, and around 7,000 drawings per half-hour episode.
I'm embarrassed to say that I thought animation was somehow more automated now. I had no clue each micro-movement was still hand-drawn, and that frankly blows my mind, and gives me even more respect for animators.
I'm embarrassed to say that I thought animation was somehow more automated now. I had no clue each micro-movement was still hand-drawn, and that frankly blows my mind, and gives me even more respect for animators.
It depends on the show. You see more and more CG, especially in cheaper shows, but the AAA shows and movies will still have each frame drawn by a human. And even with CG you'd likely review each frame to ensure it came out correctly.
Most studios doing 2D television animation in the States are using Flash (mostly CS2 or CS3 vintage) and it's still slow, tedious work drawing every frame.
American prime-time cartoon seemed overly focused on non-serious work.
Don't get me wrong, I loved Courage, the Cowardly Dog, but there's a reason why I continue to watch anime.
Don't get me wrong, I loved Courage, the Cowardly Dog, but there's a reason why I continue to watch anime.
Most American cartoons like that have very episodic designs. This presumably is great for business, as you avoid a series of uncomfortable questions about what happens to your cash cow when your series comes to an end or takes a turn that people aren't interested in anymore. It also means that new viewers are missing very little, so you can run a bunch of cartoons and not worry about getting people hooked up front in Ep 1 and you can experience business synergies from broadcasting a variety of cartoons and getting people to try new things (it's very low risk).
With live actors, I guess dealing with your stars is already a series-longevity issue, so maybe people are more willing to countenance finitude? And there are different kinds of hype-train available with real actors, and the audience for live shows has self-segregated.
My dream is that some day someone at Cartoon Network who watched Puella Magi Madoka Magica will get an experimental budget to launch a dozen episodes of a real, serious cartoon series with incredibly good writing, and win all the awards. (Bonus points if they also troll people into thinking it's fluff for the first two-and-a-half episodes and try to break their minds.) In the meantime, I guess there's Steven Universe? which at least deals with serious issues a little (still in episodic ways)
With live actors, I guess dealing with your stars is already a series-longevity issue, so maybe people are more willing to countenance finitude? And there are different kinds of hype-train available with real actors, and the audience for live shows has self-segregated.
My dream is that some day someone at Cartoon Network who watched Puella Magi Madoka Magica will get an experimental budget to launch a dozen episodes of a real, serious cartoon series with incredibly good writing, and win all the awards. (Bonus points if they also troll people into thinking it's fluff for the first two-and-a-half episodes and try to break their minds.) In the meantime, I guess there's Steven Universe? which at least deals with serious issues a little (still in episodic ways)
> Bonus points if they also troll people into thinking it's fluff for the first two-and-a-half episodes and try to break their minds.
You might like Space Dandy. Japanese again, but they really play around with continuity expectations between their episodes. I'm pretty sure most of s1 is meant to occur simultaneously in parallel realities due to the (lack of) continuity between episodes and occasional story-ending events. It's never really explained. And then s2 seems to tie some of the threads back together into a cohesive story arc over the season. Bonus: it's almost always available on Netflix, with their recent push for more anime.
You might like Space Dandy. Japanese again, but they really play around with continuity expectations between their episodes. I'm pretty sure most of s1 is meant to occur simultaneously in parallel realities due to the (lack of) continuity between episodes and occasional story-ending events. It's never really explained. And then s2 seems to tie some of the threads back together into a cohesive story arc over the season. Bonus: it's almost always available on Netflix, with their recent push for more anime.
Holy crap, I can't believe so much stuff is still animated on paper. When I was in the animation industry back around 2000 it was pretty clear that hunching over a light box was on its way out; I even worked on one project that was using traditional ink and paint for stylistic reasons (it was trying to look like an old 1920s cartoon) and the folks overseas thought we were crazy for doing this.
Why do they use Korean animators? I was under the impression that South Korea was a quite developed country and economically on more or less equal footing with the US.
South Korean cost of living is about 2/3 of the cost in the USA, so wages are likely a bit cheaper [0]. I would guess that since, as the article says, the industry has been centred there for 30 years, at this point they probably trump the US in expertise and experience.
[0] http://www.nationmaster.com/country-info/stats/Cost-of-livin...
[0] http://www.nationmaster.com/country-info/stats/Cost-of-livin...
That's the wrong impression. South Korea GDP per capita is less than half the US, $56k vs $27k. (GDP PPP per capita is a bit closer at $56k vs $37k, but nominal GDP per capita is the relevant metric for the cost of producing an export.)
It's interesting to realise how animations are still visualised on paper.
Given that video games exist, thus it is clearly possible to animate highly diverse and dynamic scenary and complex interactions, why the hell are people drawing scenes still? That was never explained well.
> why the hell are people drawing scenes still?
Net everyone has the equipment and skills (though here we are talking about companies that would), and some just prefer the traditional technique, and some prefer the output of the traditional methods (it takes quite some skill and time to convincingly mimic this in digital media, more than you might think).
Older techniques don't just get abandoned when new ones are developed, the newer methods are just new tools and sometimes the older tool is the one you want for the job.
Net everyone has the equipment and skills (though here we are talking about companies that would), and some just prefer the traditional technique, and some prefer the output of the traditional methods (it takes quite some skill and time to convincingly mimic this in digital media, more than you might think).
Older techniques don't just get abandoned when new ones are developed, the newer methods are just new tools and sometimes the older tool is the one you want for the job.
* tradition (we've always done it this way! retraining is hard! - which gets mentioned a lot in the article)
* stylistic choice ('toon shaders' keep getting better but still look kinda weird compared to what an expert animator can do; there is a distinct visual pleasure to looking at A Cartoon versus a 3D model)
* it's fun! There's a certain magic to making a series of still drawings come to life.
* malleability (if you want wild takes, distortions, weird transformations, extreme poses, etc, you have to do a lot more work in a 3d package and explicitly model it, then figure out ways to swap it in versus Just Drawing It; hell, even a simple costume change for a few scenes is a lot more work in 3D)
* much more complicated and expensive tools (the traditional pencil and paper method is gonna cost you maybe a couple hundred bucks to get started, a pro-level 3D package is going to be around a thousand bucks, more if you end up with the common workflow of one package for modelling, and another for animating, plus there's a crazy steep initial learning curve - I'm working on a show pitch intended to be in super-stylized 3D, and I'm really not looking forwards to learning how to wrangle that if it actually goes anywhere.)
We're seeing a lot of lavishly-made 3D feature cartoons, but the 3D tv cartoons we're seeing are a lot rarer. This is because it costs a hell of a lot more to meticulously model a bunch of new stuff for every episode compared to drawing it - you always want to have new places to go, new characters to deal with.
That said the state of the art of making 3D look like a cartoon has come a long way, my favorite current example is Guilty Gear Xrd, which is completely done in Unreal and is nearly indistinguishable from a hand-drawn fighting game. And had a lot of sweat and blood behind the scenes making it look that way: http://www.gdcvault.com/play/1022031/GuiltyGearXrd-s-Art-Sty...
TV studios are starting to experiment with this; a friend of mine at Nickelodeon (who previously ran the all-cg Kung Fu Panda show) is working on a project that's being built and animated in Unreal: https://www.unrealengine.com/blog/gdc-2016-a-glimpse-at-the-... - search on 'Project 85'.
* stylistic choice ('toon shaders' keep getting better but still look kinda weird compared to what an expert animator can do; there is a distinct visual pleasure to looking at A Cartoon versus a 3D model)
* it's fun! There's a certain magic to making a series of still drawings come to life.
* malleability (if you want wild takes, distortions, weird transformations, extreme poses, etc, you have to do a lot more work in a 3d package and explicitly model it, then figure out ways to swap it in versus Just Drawing It; hell, even a simple costume change for a few scenes is a lot more work in 3D)
* much more complicated and expensive tools (the traditional pencil and paper method is gonna cost you maybe a couple hundred bucks to get started, a pro-level 3D package is going to be around a thousand bucks, more if you end up with the common workflow of one package for modelling, and another for animating, plus there's a crazy steep initial learning curve - I'm working on a show pitch intended to be in super-stylized 3D, and I'm really not looking forwards to learning how to wrangle that if it actually goes anywhere.)
We're seeing a lot of lavishly-made 3D feature cartoons, but the 3D tv cartoons we're seeing are a lot rarer. This is because it costs a hell of a lot more to meticulously model a bunch of new stuff for every episode compared to drawing it - you always want to have new places to go, new characters to deal with.
That said the state of the art of making 3D look like a cartoon has come a long way, my favorite current example is Guilty Gear Xrd, which is completely done in Unreal and is nearly indistinguishable from a hand-drawn fighting game. And had a lot of sweat and blood behind the scenes making it look that way: http://www.gdcvault.com/play/1022031/GuiltyGearXrd-s-Art-Sty...
TV studios are starting to experiment with this; a friend of mine at Nickelodeon (who previously ran the all-cg Kung Fu Panda show) is working on a project that's being built and animated in Unreal: https://www.unrealengine.com/blog/gdc-2016-a-glimpse-at-the-... - search on 'Project 85'.
Photography exists and still people insists on painting on canvas. How come?
Presumably to create art/conepts that don't exist in real life and thus can't be photographed...
My point is that, you have a fairly known batch of charachters, obviously you can't (and probably shouldn't) eliminate real artists but it seems like software would be much better than a warehouse of humans. I gave the example of video games because they clearly create a game engine and level editor and thus they re able to internally build maps quickly as well as allow the charachters to interact with both the environment and otehr characthers.
I know nothing about animation, but I rememeber seeing a demo of adobe after affects that lets you create and animate a charachter with just photoshop and your laptop camera. So I don't understand the leverage of having a warehouse of people draw by hand 20-30 images per 1 second of telelevision. The after affects came out pretty awesome, so I suppose having more people and better software would be a huge improvement on what is already pretty awesome.
Maybe that is what they meant by "transition to digital" but they didn't make it clear how or why computers are emloyed and the processes they cant do.
My point is that, you have a fairly known batch of charachters, obviously you can't (and probably shouldn't) eliminate real artists but it seems like software would be much better than a warehouse of humans. I gave the example of video games because they clearly create a game engine and level editor and thus they re able to internally build maps quickly as well as allow the charachters to interact with both the environment and otehr characthers.
I know nothing about animation, but I rememeber seeing a demo of adobe after affects that lets you create and animate a charachter with just photoshop and your laptop camera. So I don't understand the leverage of having a warehouse of people draw by hand 20-30 images per 1 second of telelevision. The after affects came out pretty awesome, so I suppose having more people and better software would be a huge improvement on what is already pretty awesome.
Maybe that is what they meant by "transition to digital" but they didn't make it clear how or why computers are emloyed and the processes they cant do.
The "transition to digital" this article is talking about is to throw out the old method of drawing animation on a huge pile of paper, and move to drawing it in a huge sequence of frames in the computer. It would still very much be a human process; essentially, you use a bunch of well-trained human brains as a very powerful non-photo-realistic renderer.
I am, frankly, kind of amazed these studios are still doing it on paper. I learnt to animate on paper and it was a royal pain in the ass. Working digitally lets you have a very fast feedback loop for developing the motion; I think the best analogy to use here is probably that of a REPL versus a slow compiled language, where you can very quickly make a tiny edit, test it, then make another tiny edit based on how it tests, and another and another.
I am, frankly, kind of amazed these studios are still doing it on paper. I learnt to animate on paper and it was a royal pain in the ass. Working digitally lets you have a very fast feedback loop for developing the motion; I think the best analogy to use here is probably that of a REPL versus a slow compiled language, where you can very quickly make a tiny edit, test it, then make another tiny edit based on how it tests, and another and another.
yeah. I am getting downvoted, but I am totally with you here. The arguments above seem nonsensical, e.g. tradition, resistance to change, not having the resources (except large media companies, i.e. the subject of this discussion who have the resources), and liking the aesthetic hand drawing produces (which I suspect could be duplicated in a computer).
> I think the best analogy to use here is probably that of a REPL versus a slow compiled language
I agree. Something like this analogy, except after it is "compiled" and after it would be "rendered by the REPL" it would be a commodity. I mean this in the sense that, typically working close to the metal allows for amazing optimizations but in terms of a video/animation you are rendering a static piece of content. So I agree, you are just spending large amounts of time and resources for what seems like very little leverage.
I bet that some talented kid at RISD or another artschool who is technical and artistic could probably do a better job with adobe creative suite, so I don't understand why an entire industry would do it like this. Also pixar was started >30 years ago at ths point.
I am not trying to condemn the industry or peoples views here. I am just trying to understand the impediments to this. It seems to be unpopular, but also have no logical reason. The only one that seemed reasonable would be difficulty to produce the aesthetics, which would be especially true for a legacy show like the simpsons where the charachters strongly predate the new technology. However, southpars, the simpsons and family guy have noticeably improved their animation and charachter rendering so I don't see this as fully valid.
edit: I should say that tehri (in a dead comment) does make a compelling argument. Reproduced for those who do not have show dead [x] enabled.
> Because it's really, and I mean REALLY hard to teach the computer to create stylised illustrations from a 3d scene. Even harder, to actually animate the scene geometry in a way that corresponds to what a human animator would draw. Think to one of the core principles of animation - squash and stretch. Every character is more or less a soft body that deforms in a volume-preserving way to add emphasis to movement and intent and that's a really important artistic element of animation. For example, just watch an old Tom&Jerry short and try to imagine how you'd represent these characters as 3d geometry. How would the animator move and deform them? How do you do the effect where Tom is hit with a shovel in the face and turns into a table? With hand-draw animation (either digital or on paper) you are just drawing the final result, but with 3d animation you have to describe the actual physical geometry in the scene.
> I think the best analogy to use here is probably that of a REPL versus a slow compiled language
I agree. Something like this analogy, except after it is "compiled" and after it would be "rendered by the REPL" it would be a commodity. I mean this in the sense that, typically working close to the metal allows for amazing optimizations but in terms of a video/animation you are rendering a static piece of content. So I agree, you are just spending large amounts of time and resources for what seems like very little leverage.
I bet that some talented kid at RISD or another artschool who is technical and artistic could probably do a better job with adobe creative suite, so I don't understand why an entire industry would do it like this. Also pixar was started >30 years ago at ths point.
I am not trying to condemn the industry or peoples views here. I am just trying to understand the impediments to this. It seems to be unpopular, but also have no logical reason. The only one that seemed reasonable would be difficulty to produce the aesthetics, which would be especially true for a legacy show like the simpsons where the charachters strongly predate the new technology. However, southpars, the simpsons and family guy have noticeably improved their animation and charachter rendering so I don't see this as fully valid.
edit: I should say that tehri (in a dead comment) does make a compelling argument. Reproduced for those who do not have show dead [x] enabled.
> Because it's really, and I mean REALLY hard to teach the computer to create stylised illustrations from a 3d scene. Even harder, to actually animate the scene geometry in a way that corresponds to what a human animator would draw. Think to one of the core principles of animation - squash and stretch. Every character is more or less a soft body that deforms in a volume-preserving way to add emphasis to movement and intent and that's a really important artistic element of animation. For example, just watch an old Tom&Jerry short and try to imagine how you'd represent these characters as 3d geometry. How would the animator move and deform them? How do you do the effect where Tom is hit with a shovel in the face and turns into a table? With hand-draw animation (either digital or on paper) you are just drawing the final result, but with 3d animation you have to describe the actual physical geometry in the scene.
South Park is actually fully CG. Done in Maya IIRC. The original short it spawned from was construction paper being moved around on an animation stand, but now it's completely virtual.
The entire industry does not do it like this. I worked in the industry around the turn of the century and saw a lot of new shows starting up on full-digital workflows. I am really surprised that it is taking some overseas shops this long to switch from roughing the animation on paper to roughing the animation in the computer.
Building it in 3D and making the computer "draw" it is another thing. I mean, we still have people doing beautiful stop-motion work because that's what they like the look of, even though CG movies tend to have a similar look. It's very much an aesthetic choice. It's art, and sometimes art involves weird choices.
The entire industry does not do it like this. I worked in the industry around the turn of the century and saw a lot of new shows starting up on full-digital workflows. I am really surprised that it is taking some overseas shops this long to switch from roughing the animation on paper to roughing the animation in the computer.
Building it in 3D and making the computer "draw" it is another thing. I mean, we still have people doing beautiful stop-motion work because that's what they like the look of, even though CG movies tend to have a similar look. It's very much an aesthetic choice. It's art, and sometimes art involves weird choices.
Loads of reasons! Lots of people enjoy the aesthetics of line art and computers still aren't very good at simulating that to the same degree of sophistication as people.
2D animation also degrades more gracefully with budget, and when your end result is as simple and abstract as Bobs Burgers, there's no need to complicate things. These shows are primarily dialogue driven, so don't require and aren't budgeted for dynamic scenery and complex interactions.
2D animation also degrades more gracefully with budget, and when your end result is as simple and abstract as Bobs Burgers, there's no need to complicate things. These shows are primarily dialogue driven, so don't require and aren't budgeted for dynamic scenery and complex interactions.
Because it's really, and I mean REALLY hard to teach the computer to create stylised illustrations from a 3d scene. Even harder, to actually animate the scene geometry in a way that corresponds to what a human animator would draw. Think to one of the core principles of animation - squash and stretch. Every character is more or less a soft body that deforms in a volume-preserving way to add emphasis to movement and intent and that's a really important artistic element of animation.
For example, just watch an old Tom&Jerry short and try to imagine how you'd represent these characters as 3d geometry. How would the animator move and deform them? How do you do the effect where Tom is hit with a shovel in the face and turns into a table? With hand-draw animation (either digital or on paper) you are just drawing the final result, but with 3d animation you have to describe the actual physical geometry in the scene.
For example, just watch an old Tom&Jerry short and try to imagine how you'd represent these characters as 3d geometry. How would the animator move and deform them? How do you do the effect where Tom is hit with a shovel in the face and turns into a table? With hand-draw animation (either digital or on paper) you are just drawing the final result, but with 3d animation you have to describe the actual physical geometry in the scene.
They're looking for profitable projects to fund. See their investment criteria.[2]
[1] https://www.cj-fund.co.jp/en/ [2] https://www.cj-fund.co.jp/en/investment/flow.html