Mathematicians bring ocean to life for Disney's 'Moana' | Hacker News Reader

Mathematicians bring ocean to life for Disney's 'Moana'

phys.org

260 pointsoscarwao91 comments

We watched Monana a couple of weeks ago and afterwards my wife said that she was amazed with how good the water looked. I said, I bet there were a bunch of mathematicians working on that :-)

In another life I'd be doing maths for 3D movies. I went to university in Wellington where Weta are based (Peter Jackson's film companies) and at one point had a look around the digital arm. I seem to recall that they had the SGI machines as desktops and a big unix rendering farm. I just love the way you can use maths to create such magic.

Anyway, if you haven't seen it yet, Moana is fantastic. Incredible animation, catchy pop hooks in the music and a brilliant storyline. It's wonderful to see a Disney film with a strong female lead that needs neither male support nor a love interest.

Take a look at an application of deep learning in accelerating complex 3d rendering. (Two Minute Papers)

Stunning Video Game Graphics With Voxel Cone Tracing - https://www.youtube.com/watch?v=iOWamCtnwTc

The title to the video you linked should be "Neural Network Learns The Physics of Fluids and Smoke"—though he does also have a voxel cone tracing video.

"Stunning Video Game Graphics With Voxel Cone Tracing" is

https://www.youtube.com/watch?v=dQSzmngTbtw

Great Video about a very intruiging subject !

Yep. I mixed up the source when I picked it from my browsing history.

That paper is kind of weird -- linear PDEs are solved by convolutions (yay for Green's functions), so using a CNN seems a bit, well, overkill and obvious?

> a strong female lead that needs neither male support nor a love interest

Now I'm trying to think of other Disney leads of either sex without a love interest... is there one? Other than those for whom it's not really applicable (ones who are too young, mostly).

Brave, with its tomboyish princess Merida is one.

Disney even got some flak for "glamming" her up to be like other Disney princesses [1]

[1] https://www.theguardian.com/film/2013/may/16/disney-princess...

The protagonist in the Force Awakens as well if I'm not mistaken.

Judy (Zootopia) Hiro (Big Hero 6) Elsa (Frozen) Ralph and Velellope (Wreck it Ralph)

There are plenty more. I think it has more been current society that is recognizing it more.

Wreck it Ralph puts the love interest relationship between Felix and Calhoun (and subverts it all over the place).

Those are all quite recent, while the historical Disney films all fall into the traditional template of male and female lead falling for each other. To the extent that Enchanted can assume the audience knows the formula that they're parodying.

There is indeed a recognition and a policy change that stories don't have to be made to fit that formula any more. And it seems that the films are exploring other affectionate relationships, such as the sisters in Frozen. Wreck it Ralph clearly has affection between Ralph and Vanellope, but it's more a parental/big-brother-little-sister thing than a romance. Brave focuses on mother/daughter. And so on.

Come on, Anna is the real hero in Frozen! :)

Lilo of Lilo & Stitch!

Mary Poppins

Home (lead is a teenage girl, other major character is not a love interest, nor actually male).

Up (about a guy escaping being sent to an old people's home .. or something).

Up's love interest is, ah, obfuscated for the sake of trying to avoid spoilers, unavailable at the time of the "main plot", but there definitely is one.

But it's a male lead (two of them) nevertheless.

I saw it and said the same thing to my wife. But also, I'm super glad that the movie is going to lead to more people knowing about Polynesian Dead Reckoning, because it is amazing.

They used wayfinding by reading what was around them, rather than dead reckoning.

You're right. "Polynesian Dead Reckoning" doesn't appear to be a thing at all. Maybe the GP was referring to the "cold reading"-like ingenuity of their navigational techniques.

https://en.wikipedia.org/wiki/Polynesian_navigation

The water was certainly incredible when I saw the movie. The cinematography was pretty nice all around.

To be pedantic, it's not cinematography.

Cinematography is the science or art of motion-picture photography by recording light or other electromagnetic radiation, either electronically by means of an image sensor, or chemically by means of a light-sensitive material such as film stock.

perhaps it was when that was the only way we could create films

I'd call it the art of deciding the movement/placement/focus of the camera, regardless of whether it's a physical camera or a software one. A scene from a different perspective is not as powerful, after all; thus the regard for cinematography.

It's definitely an Art form, but if you look in IMDB, there's no cinematographer listed. That function is "visual effects".

  Rob Dressel      ...  director of cinematography: layout
  Adolph Lusinsky  ...  director of cinematography: lighting

Awesome, go Rob! (Historically cinematography in animation tended to be called simply "layout")

As smhost pointed out, this is simply a false statement.

In case others would like to verify, the full credits on IMDB are here: http://www.imdb.com/title/tt3521164/fullcredits?ref_=tt_ov_s...

There are two cinematographer credits listed for the movie.

Disney uses a difference organizational design to produce animated films, including going all the way back to the days when they were shot on film with real cameras. The lack of a "cinematographer" title in IMDB for the film has nothing to do with the technology and everything to do with the organization of the team.

take Rango (2011) for example.

     Roger Deakins	...	(cinematography consultant)
     Mike Sanders	...	virtual cinematography

Deakins worked with the team to determine camera angles, lighting placement, and framing, among other things. They paired an old school film cinematographer with a digital effects supervisor, and the output of their work is what anyone should call cinematography.

> I seem to recall that they had the SGI machines as desktops and a big unix rendering farm.

Is this during LoTR or even before? I though SGI was out by the time of LoTR

I'd guess before, they've been in the CGI business for quite a while. I saw a presentation from Milton back in 2000 (I'm also a VUW CS grad) with pictures of the render wall.

At one point they bought the building of an ice cream factory in Miramar because it already had a massive electrical feed.

They employ a fair few people in Wellington, my neighbour used to work for Weta as a sysadmin. Everyone was a contractor, doing 10 hour days. Not my scene.

Erm, not sure - it would have been 1999 maybe. I can't even remember what I was doing there, meeting a friend of a friend I think.

Actually, I seem to recall that there were guys around at the time that were making the monsters for Xena or one of those shows.

More trivia - LOTR SGI machines went to the railway afterwards and were used by the engineers for making CAD models of wagons and things.

Pretty sure Irix was the _Heavenly Creatures_ era.

How does Weta compare to Animal Logic?

"Solve, or nearly solve, partial differential equations". That's game physics. Most of the effort is to come up with ways to "nearly solve" without having awful stuff happen. There's a long history of awful stuff, going back to when Seamus Blackley botched Trespasser in 1998.

Game physics still tends to go "boink", because with impulse/constraint collisions, everything, including large, heavy objects, bounces instantaneously. Stuff flying apart, though, is rare now; most systems drain the energy out of a system when they detect that happening. It's physically wrong, but looks less awful.

I used to work on this stuff. I solved the "boink" problem for articulated rigid body physics in the 1990s, but couldn't make it work in real time on 100MHz CPUs.

> Stuff flying apart, though, is rare now; most systems drain the energy out of a system when they detect that happening. It's physically wrong, but looks less awful.

Most games are using position based dynamics now (see Nvidia PhysX) so there isn't even a variational principle to arrive at the system from an energy. Just a constraint satisfaction problem with a weak notion of momentum coupled in.

I loved the physics in Trespasser. Good times.

Also coming into play during Moana is the hyperion rendering engine, which was first showcased in Big Hero 6:

https://www.disneyanimation.com/technology/innovations/hyper...

Also, here's an article containing some stills from the movie:

http://blog.yiningkarlli.com/2016/11/moana.html

Thank for sharing - never been on DA website.

The whole blog of publications is super interesting!

https://www.disneyanimation.com/technology/publications

Glad you like it. I work at Disney Animation. I have been pushing to get more "Innovations" posted but it's not always the easiest thing to get "secret sauce" published. https://www.disneyanimation.com/technology/innovations

I also helped push this forward to get other divisions to think about open sourcing code more often. http://disney.github.io/

I was looking at the animation referenced in the article of water flowing over a rocky terrain and couldn't help but notice almost life-like and pixel-perfect precision of collision detection between water particles and the rocky surface. Yet, in modern gaming, while it's improved a great deal, quite often you'll see surfaces "fusing" together before a collision is detected. I think there is a more technical term for this, but it's not coming to mind. Is this just a matter of usability vs performance?

I would imagine its a performance issue. For Moana they can render the water on a server farm with has much computing power as they need. Where a console or personal computer is quite limited in contrast.

Further to that point they can take as long as they like to render a single frame.

The movie is using a whole compute cluster to do a physics simulation when rendering the movie. Your computer has 1 CPU and 1 GPU.

Your computer is just very limited in how much attention to detail it can have, because it has wildly less processing power and has to get its job done in real time.

By contrast, Disney is using a server farm to spend 3 seconds analyzing the mechanics of a tuft of hair wiggling in the breeze for 1 second.

Additionally, multiplayer games often have to run a low-geometry simulation server side to reduce cheating and account for prediction/latency sync issues.

While certainly not directly linked to poor physics simulations of water in computer games I would guess it has played a big role in holding back advancements in this area because how much locally-only non-game mechanic influencing visuals could cause contrast with need to be duplicated on the server physics.

To put this in perspective, 60fps == ~16.7ms per frame.

That's 16.7ms to run your WHOLE game engine, you get some fraction of that left over to do rendering.

The other thing to remember is that Disney are completely at liberty to reject and re-render scenes when something doesn't look perfect. Arguably, this even applies to examples chosen for the demo video.

Has anyone tried to make a real game in the same way?

Ie a game which was hosted on a remote desktop that had direct access to a GPU server farm on a very fast network.

This was initially attempted by G-Cluster in 2000, Crytek tried it as well, popularized by OnLive (who started working on it in 2003) and blew up in 2010.

The general term for it is cloud gaming https://en.wikipedia.org/wiki/Cloud_gaming

The answer is no. People are bringing up OnLive and Nvidia but they are just running the same game you run locally on remote hardware.

What you are asking is basically has anyone run a game on a supercomputer/cluster so that the graphics are way better than what could be done locally. The answer is no, nobody has. That would be pretty neat though. It would be super expensive to make with no way to recoup the cost.

It certainly could happen - http://www.nvidia.com/object/cloud-gaming.html - how much will someone pay per hour to play one game? It feels like a timewarp to pre-AOL when online games cost more per hour than minimum wage.

Here is an example that is not what you are looking for but is somewhat related. It's a game that is played by rendering high quality screenshots of the game world on a server. I enjoyed it a lot.

https://extrasolar.com/

https://lg.io/2016/10/12/cloudy-gamer-playing-overwatch-on-a...

http://www.nvidia.com/object/cloud-gaming.html

Not a field I work in but I guess it would never work. You'll have the network latency to contend with. Or people would need to self host a server farm, which probably makes your market pretty small :-)

I wouldn't say never. The golden rule of computer games code is that you get to cheat a lot.

https://www.playstation.com/en-us/explore/playstationnow/

The quality of game renderings will lag behind movie renderings for a long time. At 60fps there is about 16.7ms available to render each frame.

For a movie[1]:

> There's a principle I call The Law of Constancy of Pain. Back when we made The Adventures of André & Wally B. in 1983 a typical frame took from several minutes to a few hours to compute. Now computers are literally millions of times more powerful and a typical final render takes from a few minutes to a few hours to compute, averaging somewhere between 2 and 4 hours. Render times have remained flat for over 30 years because that's just how long people are willing to wait for them.

So movie artists are willing to have each frame of a movie take 2-4 hours to render vs 16.7 milliseconds.

[1]: https://www.quora.com/How-long-does-it-take-to-render-a-Pixa...

The thing I noticed was how there seems to be some sort of hydrophobic coating on the rocks. They don't get wet like real rocks. The lava demo is similar.

I've been waiting decades to play a surf sim that utilizes proper fluid dynamics. Maybe before I die it'll happen.

Wave Race 64 had plausible wave physics. Could probably go much further with the hardware of today.

God, same. I would love for the team that worked on the Skate series to approach a surfing game.

Why is it that we can accept expressive/cartoonish looking characters but need to have the most realistic environmental rendering (light, physics effects)?

I challenge the studios to be imaginative and equally expressive on the latter. Give us the miyazaki equivalent in 3d animated films.

Don't take fthe easy way out by emulating reality. Put true art into it

My guess is the uncanny valley[0] is still too big for "realistic" human characters. As for the goofy-looking chicken, it's convenient to make an animal goofy-looking for comic relief. (And in the case of Jiminy Cricket, I think Disney would rather draw, and people would rather look at, a humanoid with a weird head than an actual insect.)

[0] "the uncanny valley is the hypothesis that human replicas that appear almost, but not exactly, like real human beings elicit feelings of eeriness and revulsion among some observers" https://en.wikipedia.org/wiki/Uncanny_valley

for the EU viewers that are confused: in some countries the title and the character's name is Vaiana.

moana pozzi was a famous italian porn actress in the '80, that made also standard tv programs and involved in politics, too. I'm pretty sure she was famous in Europe too.

in Italy the movie is called 'Oceania' for ... reasons.

I wish the article went more into detail about what is novel about the algorithms they worked on.

https://www.math.ucla.edu/~jteran/papers/JSSTS15.pdf

You can even play with the algorithm they used "The Affine Particle-In-Cell Method" using the code discussed here

https://news.ycombinator.com/item?id=13325190

The hair was also great in Moana. It is fun to see nonlinear Cosserst rod theory in use on the big screen!

The Taichi library linked [1] on HN recently has a MIT licensed APIC implementation here :

https://github.com/IteratorAdvance/taichi/blob/master/includ...

[1]: https://news.ycombinator.com/item?id=13325190

This is really cool. I wonder how one goes about getting into work like this - niche field but few qualified to work it, I'd guess.

You can come do a PhD with me at Waterloo. #shamelessplug https://cs.uwaterloo.ca/~c2batty/

Small world! I actually am majoring in comp sci and physics. I'll definitely message you if I end up getting into research next year.

I am a Phd student in physics based simulation for Computer Graphics. I was a summer intern at WDAS and most of my colleagues were two of the following : 1) MS in CS with specialization in COmputer Graphics from places like UPenn 2) Phd in CS with specialization in Computer Graphics/Phd in Applied Mathematics from UCLA/Harvard.

The answer is right there in the article. Applied and numerical mathematicians at UCLA math department have for many years been helping out Hollywood with these problems. I'm sure their students, collaborators in other universities, etc. the academic community provides a gateway to get into this kind of work.

That said, UCLA Math dept is one of the best in the country. Some very talented people work there.

Start with a PhD in a relevant area.

very niche

Scarily realistic: https://www.youtube.com/watch?v=3mULoRzfFcE

A little off-topic, but this site's cert appears to have a revoked intermediate authority. Pic: http://i.imgur.com/lIseT66.png

But no-one else seems to be complaining about this so I wonder if it's just me?

All fine for me in Safari on OSX 10.10.3 http://i.imgur.com/ZwbQ08N.png

In the video, it is mentioned several times "stability" and "noise". Can someone explain what do these two terms refer to in the sense of particle simulation?

"where they used science to animate snow scenes"

Cringes

Well, they did, in a very non-buzzwordy manner.

Especially that they bolded "used science".

Use the science, Luke.

This is nice from an artistic perspective. But from a computational physics perspective, it feels like cheating. How much of this is realistic? And are we now encouraging more and more people to focus on "fake" physics because it makes it somehow easier to produce something cool? I sure hope not.

It's not all about looking cool, it's about coming up with algorithms that are computable in a reasonable amount of time while also behaving as similarly as it is possible to interactions of objects in the real world.

Faking physics is a mastery. Trade-offs are unavoidable.

Reply on news.ycombinator.com