Latest / What's Your Problem? / Creative Technology at Pixar

Transcript

Creative Technology at Pixar · What's Your Problem?

0:15
Speaker 1: Pushkin. There's a bunch of stuff I want to ask you.
0:22
Speaker 1: A lot of it is about the relationship between technology
0:26
Speaker 1: and creativity, and especially given the fact that you've been
0:28
Speaker 1: working at Pixar for a long time, respectfully, how that
0:33
Speaker 1: changes over time?
0:34
Speaker 2: Okay, so how creativity changes as technology changes?
0:37
Speaker 1: Yeah? So, actually, can you just say your name and
0:39
Speaker 1: your job?
0:40
Speaker 2: Sure? My name is Danielle Feinberg, and I work at
0:43
Speaker 2: Pixar Animation Studios and I am a visual effects supervisor.
0:48
Speaker 1: So you're a visual effects supervisor. Now, if I understand correctly,
0:50
Speaker 1: you spent a lot of your career there. You've been
0:52
Speaker 1: at Pixar for more than what twenty five years at
0:55
Speaker 1: this point?
0:56
Speaker 2: So right, yeah, I hit twenty six.
0:58
Speaker 1: Musseltov, you were director of photography for lighting for a
1:04
Speaker 1: lot of your time there, right, And that's striking to
1:06
Speaker 1: me because, if I understand correctly, there are in fact,
1:10
Speaker 1: like no photographs and no actual lights in Pixar movies. Right,
1:14
Speaker 1: So what does it mean to be the director of
1:17
Speaker 1: photography for lighting on you know, movies like say Coco.
1:21
Speaker 2: So picky of you, Jacob, I'm sorry, so curious, I'm curious.
1:30
Speaker 1: I'm Jacob Goldstein. And this is What's Your Problem, the
1:32
Speaker 1: show where I talk to people who are trying to
1:35
Speaker 1: make technological progress. My guest today, as you've already heard,
1:39
Speaker 1: is Danielle Feinberg, and as you will hear, we did
1:42
Speaker 1: talk about creativity and technological change. Also we talked about
1:46
Speaker 1: the problem she's trying to solve now, and we talked
1:49
Speaker 1: about how we can often wind up following rules that
1:52
Speaker 1: are not particularly useful and that in a lot of
1:55
Speaker 1: cases we don't even know we are following.
2:01
Speaker 2: So director of photography is in a live action movie,
2:05
Speaker 2: is the same as a cinematographer, and so they're doing
2:07
Speaker 2: camera and lighting together essentially. And so in our world
2:12
Speaker 2: and animation, we split that what is traditionally one role
2:15
Speaker 2: in live action to two roles, and so we have
2:18
Speaker 2: a director of photography for camera and a director of
2:20
Speaker 2: photography for lighting.
2:21
Speaker 1: And just to be clear, I mean camera is a
2:23
Speaker 1: metaphor in this instance, right, like just so I'm thinking
2:27
Speaker 1: about it, right, there is a camera, right.
2:30
Speaker 2: There is a virtual camera, and so we build a
2:33
Speaker 2: virtual three dimensional world, and we actually have a chunk
2:36
Speaker 2: of code that is our camera that mimics a lot
2:38
Speaker 2: of things that a real life camera does. And as
2:44
Speaker 2: time has gone by, in fact that that camera has
2:47
Speaker 2: gotten more and more like real life. And we even
2:50
Speaker 2: use specific sets of lenses, sometimes modeled after certain companies
2:55
Speaker 2: lenses and stuff, and.
2:56
Speaker 1: So the lenses are our code basically that generate images
2:59
Speaker 1: that look like a picture you take with a certain.
3:02
Speaker 2: Actual yes, exactly exactly.
3:05
Speaker 1: So similarly than presumably the lighting is sort of similarly
3:09
Speaker 1: virtual but modeled on actual lights.
3:12
Speaker 2: Yeah, you know it's originally it was modeled on a
3:17
Speaker 2: version of a light. It wasn't like, oh, this is
3:21
Speaker 2: tungsten light, or this is a light bulb that I
3:24
Speaker 2: would have in my lamp at home. It was like,
3:27
Speaker 2: here's a light. You can do whatever you want with it.
3:29
Speaker 2: You can make it as bright as you want, you
3:31
Speaker 2: can make it as big as you want, you can
3:33
Speaker 2: all kinds of things. And so in those sort of
3:35
Speaker 2: early days, it was easy to do things that looked
3:39
Speaker 2: sort of wrong visually because we weren't beholden too some
3:42
Speaker 2: of the same physics in real life, huh.
3:46
Speaker 1: And so it was almost like too much. You had
3:49
Speaker 1: too much power, and you could just get crazy with
3:51
Speaker 1: the light and was going on, Yeah.
3:54
Speaker 2: We had all the power and you had to use
3:55
Speaker 2: it wisely.
3:56
Speaker 1: Yes, So, I mean, so there's different ways to talk
4:01
Speaker 1: about to talk about this, and I want to try
4:04
Speaker 1: a few. But one way that I thought might be
4:06
Speaker 1: interesting to frame this is to talk to talk about
4:10
Speaker 1: the way that the craft really of what you do
4:13
Speaker 1: has changed. I know you worked on Turning Red, the
4:16
Speaker 1: recent Pixar movie, which, like millions of other people, I
4:19
Speaker 1: watched at home during the pandemic with my daughters. So
4:22
Speaker 1: thank you for that. Excellent And then you also you
4:25
Speaker 1: worked on A Bugs Life in the late nineties, right,
4:28
Speaker 1: And so I thought it might be interesting to compare
4:32
Speaker 1: A Bugs Life to Turning Red in terms of the
4:35
Speaker 1: sort of technology and creativity, right, how you created in
4:38
Speaker 1: those different really kind of technological erars. I mean, that's
4:41
Speaker 1: a long time in the development of computer generated animation, right.
4:46
Speaker 1: I mean one question in that context is what is
4:50
Speaker 1: something that you wanted to do when you were making
4:53
Speaker 1: A Bugs Life but that you couldn't that now would
4:56
Speaker 1: be really easy to do.
4:58
Speaker 2: Yeah, probably so many things. When you think about A
5:03
Speaker 2: Bugs Life. That was my first film I worked on here,
5:07
Speaker 2: and at that time in the world, there there was
5:09
Speaker 2: one feature length computer animated film that had ever been released,
5:13
Speaker 2: which was Toy Story. A lot of what was happening
5:16
Speaker 2: was just being made up of how to make these
5:18
Speaker 2: movies and so and it was a very tiny group
5:21
Speaker 2: of people just kind of figuring out how to do
5:24
Speaker 2: this computer animation thing as a movie. And so, you know,
5:26
Speaker 2: back then Pixar was just this tiny little company that
5:30
Speaker 2: like nobody knew Pixar made Toy Story, like they saw
5:34
Speaker 2: the Disney on it, and like nobody. I used to
5:36
Speaker 2: have to say, oh, I do computer animation. I work
5:38
Speaker 2: out a place called Pixar. And then around the time
5:41
Speaker 2: of maybe finding Nemo, I said that and someone was like,
5:43
Speaker 2: we all know what Pixar is. You can stop saying,
5:48
Speaker 2: and so it was very cool. But like on A
5:50
Speaker 2: Bugs Life, part of the thing is it's this colony
5:53
Speaker 2: of ants, and you know what we know of the
5:55
Speaker 2: ants is there's a lot of them once you find
5:57
Speaker 2: them and where they live. And doing a scene with
6:01
Speaker 2: one hundred ants in it back on A Bug's Life
6:03
Speaker 2: was a huge, huge deal. We had to write a
6:05
Speaker 2: whole chunk of crowd software and figure out really how
6:09
Speaker 2: to jam it into the system and get it to
6:11
Speaker 2: render and meaning generate the final images, which is when
6:15
Speaker 2: the computer is crunching all of the work you've done
6:19
Speaker 2: and doing a million calculations to generate just one image,
6:25
Speaker 2: and you need twenty four of those images to make
6:27
Speaker 2: one second of film. So for a ninety minute film
6:29
Speaker 2: you're doing, you can imagine the sort of scope of
6:31
Speaker 2: the calculations you're doing. And doing a crowd of one
6:34
Speaker 2: hundred was a really big deal. You know, we can
6:37
Speaker 2: do crowds much bigger than that. Now we can give
6:40
Speaker 2: them all kinds of sort of acting properties. Where on
6:43
Speaker 2: a bug's life it really was someone was animating an
6:47
Speaker 2: ant and then you could sprinkle that animation across things,
6:51
Speaker 2: and then they'd animate another bit and you sprinkle in
6:53
Speaker 2: some of the other ones. But like I was giving
6:57
Speaker 2: a talk for I do a lot of talks for
6:58
Speaker 2: girls to try and keep them interested in math and
7:01
Speaker 2: science to get some better numbers in STEM, and I
7:04
Speaker 2: pulled some clips from a bug's life, and I was
7:07
Speaker 2: going through this whole thing of like, you know, what
7:09
Speaker 2: you don't want is for some two ants to be
7:12
Speaker 2: doing the same movement, and then you can see in
7:14
Speaker 2: the clip I pulled, like three ants are moving their
7:16
Speaker 2: arms like this or something. And it was very funny
7:18
Speaker 2: because Back then it was really what we were doing
7:20
Speaker 2: was very hard, and now it's much easier, and it
7:23
Speaker 2: would have.
7:23
Speaker 1: Just been too computationally intensive to have all the ants
7:28
Speaker 1: doing different things basically, so you sort of cheated by
7:30
Speaker 1: like having some of them do the same thing.
7:32
Speaker 2: Yeah, and like you did ten little animated bits and
7:34
Speaker 2: sprinkled at a crossing, and you thought that's probably enough,
7:37
Speaker 2: and we were probably out of time, honestly, you know,
7:40
Speaker 2: also and the next.
7:42
Speaker 1: Certain level it was enough, right, Like I don't know
7:44
Speaker 1: that people watched a Bug's Life and like, oh, that's
7:47
Speaker 1: so cheesy. Those two ants are doing the same thing.
7:50
Speaker 2: Right, And if we're doing our job right, you're actually
7:52
Speaker 2: looking at the characters who are doing the acting, and
7:55
Speaker 2: the crowd is the background in a lot of ways.
7:58
Speaker 2: But you know, there's this whole thing of back then
8:00
Speaker 2: you were you were much closer to some of the
8:02
Speaker 2: sort of the guts and the code, and you'd write
8:04
Speaker 2: a little code and you'd generate a picture and you'd
8:07
Speaker 2: kind of a lot of back and forth. I think
8:09
Speaker 2: a lot of the development that's happened over time has
8:11
Speaker 2: been to abstract away from that code so that people
8:13
Speaker 2: that really are coming from maybe a pure art background,
8:17
Speaker 2: or they're somewhat technical but they're really artistically based, could
8:21
Speaker 2: still thrive in this area. Now, Pixar is this really
8:24
Speaker 2: wonderful place where we have all of these people that
8:26
Speaker 2: are really like this wonderful combo of art and technology,
8:29
Speaker 2: and some are very technical and some are very artistic,
8:32
Speaker 2: and a lot of people are somewhere in between. So
8:35
Speaker 2: it makes it kind of accessible to more people in
8:39
Speaker 2: a great way, I think. But back in the old days,
8:42
Speaker 2: it was definitely it was just a different thing, and
8:46
Speaker 2: you were spending a lot of your time kind of
8:48
Speaker 2: not in a creative space. Part of it was in
8:52
Speaker 2: a just trying to get the computer to do what
8:54
Speaker 2: you want to do in a kind of rudimentary way.
8:57
Speaker 1: Yeah, I mean, clearly Pixar movies have gotten more beautiful, right,
9:05
Speaker 1: and certainly complex, and so the benefits of technological improvements,
9:15
Speaker 1: I mean both. I'm sure a lot of it is
9:17
Speaker 1: just processing speed and more's law, but also I'm sure
9:19
Speaker 1: also on the sort of software side that you all
9:22
Speaker 1: have developed. So it's clear the ways in which it's
9:24
Speaker 1: gotten better, right. I'm curious though, In fact, I heard
9:28
Speaker 1: in one of the interviews I listened to in preparing
9:31
Speaker 1: for this interview, one of your interviews, you mentioned this
9:34
Speaker 1: idea that when you get new tools, obviously you gain things,
9:37
Speaker 1: but sometimes there are things that you lose as well.
9:41
Speaker 1: And so I'm curious when you look back, you know
9:44
Speaker 1: too much earlier in your career, when you had fewer
9:47
Speaker 1: tools and frankly worse technology, were there things that I
9:52
Speaker 1: don't know that you could do better or differently? Are
9:54
Speaker 1: there things you miss about that time from a creativity standpoint.
9:59
Speaker 2: Yeah, definitely. So as an example of things, I'll start
10:03
Speaker 2: with the things that we couldn't do and now can.
10:06
Speaker 2: So bugs life, I mentioned the crowd, and then when
10:10
Speaker 2: we got to Monsters Incorporated, we needed a monster that
10:13
Speaker 2: had hair. We couldn't We had no idea how to
10:15
Speaker 2: do hair. That was a huge R and D project.
10:19
Speaker 2: The next was finding Nemo. We didn't know how to
10:20
Speaker 2: do underwater, and so we had to figure out how
10:22
Speaker 2: to do underwater and what the elements were, and then
10:25
Speaker 2: sort of developed the technology so that instead of just
10:28
Speaker 2: in a test shot, all the people across the show
10:31
Speaker 2: could could do underwater in this sort of easy way,
10:35
Speaker 2: from the most technical person down to the most junior
10:37
Speaker 2: that they could still sort of API, so you.
10:39
Speaker 1: Didn't just have to sort of render water, you had
10:41
Speaker 1: to build kind of a toolkit so that people could
10:44
Speaker 1: create in water.
10:45
Speaker 2: Yeah, and so that's one of the really fun things
10:47
Speaker 2: is you take some we took some live action footage
10:50
Speaker 2: of water and tried to recreate it, and then you
10:53
Speaker 2: start breaking down all of that messy looks development work
10:56
Speaker 2: you did, and it turned out there were five or
10:58
Speaker 2: six elements that when you put them together, we can
11:00
Speaker 2: make it look like underwater. And then you take each
11:02
Speaker 2: of those elements and turn it into a more pathge
11:04
Speaker 2: that's right. Yeah, yeah, Yeah, it's really fun. It's kind
11:07
Speaker 2: of scary while you're doing it because you don't know
11:08
Speaker 2: if you're ever going to get anywhere good. But and
11:12
Speaker 2: so then you go to a movie like The Incredibles
11:14
Speaker 2: and the daughter Violet has long hair, and that was
11:17
Speaker 2: a completely unsolved problem in computer graphics, and it was
11:20
Speaker 2: all the way up kind of to the last minute
11:23
Speaker 2: that it wasn't it got solved.
11:25
Speaker 1: Uh, she might have had short hair if if somebody
11:27
Speaker 1: had something out.
11:29
Speaker 2: Yeah, but this is one of the fun things about
11:30
Speaker 2: Pixar is sort of if a director says, well, that's
11:33
Speaker 2: a story point which is what Bradbird, the director said.
11:35
Speaker 2: He's said, well, this is how we're telling the audience
11:39
Speaker 2: who Violet is. So she hides behind her long hair
11:42
Speaker 2: and it's this big sort of story moment, and he says,
11:44
Speaker 2: it's a story point and that is sort of good
11:47
Speaker 2: as gold here, and it doesn't mean you spend a
11:48
Speaker 2: lot of time talking about whether you're going to do it.
11:50
Speaker 2: You just go figure out how to do it. And
11:52
Speaker 2: so it's this creativity driving technology and I think this
11:55
Speaker 2: really wonderful way and a very fun and scary way.
11:59
Speaker 2: But it's like we're doing this. There was quite a
12:02
Speaker 2: while where Violet's hair looked like kind of March Simpson's beehive,
12:07
Speaker 2: because when the computer doesn't do it right, it does
12:09
Speaker 2: something pretty spectacular with hair and cloth, so.
12:12
Speaker 1: It goes way wrong.
12:13
Speaker 2: Yeah yeah, yeah. So so those are all things. Those
12:16
Speaker 2: are examples of sort of things we literally couldn't do
12:19
Speaker 2: that then the movie and the story spirit so that
12:21
Speaker 2: we then know how to do it. One of the
12:24
Speaker 2: things is there was kind of a push to get
12:26
Speaker 2: more and more real life mimicking real life more and more.
12:30
Speaker 2: So as we push, say lighting and materials more towards
12:33
Speaker 2: what sort of real life light responses. Say, suddenly, what
12:37
Speaker 2: started to happen is that we could get to something
12:39
Speaker 2: that looked better much faster. So you can start to
12:42
Speaker 2: put the sun in. In the olden days, we'd put
12:45
Speaker 2: the sun in and then you'd put another light that
12:47
Speaker 2: was the light bouncing off the ground, and you put
12:48
Speaker 2: another light in that the light coming from the sky,
12:50
Speaker 2: a bounce light, a fill light, a key light. And
12:53
Speaker 2: now what we could do is you put the sun
12:55
Speaker 2: in and the computer is calculating how that light is
12:57
Speaker 2: bouncing around, and then you can start aug.
12:59
Speaker 1: The computer is just figuring out what the world would
13:02
Speaker 1: look like if the sun were in that.
13:03
Speaker 2: Spot in the stit if the sun were in that
13:04
Speaker 2: spot and the light was bouncing off of things. And
13:07
Speaker 2: these materials are more real life than they were before.
13:10
Speaker 2: They have a better material response, and like, oh, doesn't
13:13
Speaker 2: that look nice? It used to take me longer to
13:15
Speaker 2: get to something that looked nice like that, And there's
13:17
Speaker 2: complexity to it that we just couldn't get before. So
13:19
Speaker 2: isn't that lovely? And so all of that is fantastic
13:25
Speaker 2: until you want to do something that is maybe not
13:29
Speaker 2: real life, say you want to stylize it, or even
13:32
Speaker 2: say the director wants you to get the shadow off
13:35
Speaker 2: the face. Suddenly that you can't do that anymore because
13:38
Speaker 2: as you push more and more towards real life, those
13:41
Speaker 2: are things you can't do very well in real life.
13:45
Speaker 1: And so and you can't and because you've built essentially
13:50
Speaker 1: a piece of software where you've said make it real life,
13:53
Speaker 1: the software is like, no, you can't take the light
13:55
Speaker 1: off the face because I'm making it real life and
13:57
Speaker 1: there would be light on their face in real life.
13:59
Speaker 2: Yeah. And so suddenly it was really funny. I'm talking
14:03
Speaker 2: to one of the sort of scientific engineering people who's
14:06
Speaker 2: who's engineering the system, and he's really brilliant guy, and
14:11
Speaker 2: I said, well, what if the director asked me to
14:13
Speaker 2: take a reflection off of something or move a shadow
14:15
Speaker 2: like that stuff we do all the time for the
14:17
Speaker 2: storytelling of it. And he said, oh, you just tell
14:20
Speaker 2: the director they can't do that. And you know it
14:23
Speaker 2: is I don't know what it's like anywhere else, but
14:26
Speaker 2: you don't really usually tell the director they can't do something,
14:28
Speaker 2: and especially not because we took that capability away, you know,
14:32
Speaker 2: like it was really hard, and he was coming from
14:34
Speaker 2: this purist he doesn't ever talk to the director. He's
14:36
Speaker 2: coming from this purists, like this is what the real
14:38
Speaker 2: world does. So there was sort of this time period
14:41
Speaker 2: where as they're getting the system up and running, the
14:45
Speaker 2: thing is they'd say, well, you could do that, but
14:47
Speaker 2: like it's going to slow everything down, and there were
14:49
Speaker 2: so many more calculations happening that slowing it down felt
14:51
Speaker 2: really like the ideal world.
14:53
Speaker 1: And people are like spending all days solving optimization problems, right,
14:57
Speaker 1: trying to speed it up so that the ants aren't
14:59
Speaker 1: all doing the same thing in the background or whatever.
15:01
Speaker 2: Right, yeah, and they're like, why do you idiots need
15:03
Speaker 2: to move a shadow that doesn't happen in real life,
15:05
Speaker 2: you know sort of thing, and so yah. So the
15:07
Speaker 2: interesting thing that's had that sort of an era, And
15:10
Speaker 2: then the interesting thing that happened is now we're in
15:13
Speaker 2: an era where we're getting some of both I think
15:15
Speaker 2: where you get this sort of harness the power of
15:19
Speaker 2: the computer in a lot of ways where it's mimicking
15:21
Speaker 2: real life, but now we have more ability to kind
15:24
Speaker 2: of break out into that. And I think that's.
15:27
Speaker 1: Because in terms of realism versus kind of.
15:30
Speaker 2: Just scilistic flavor and stylization and what a director wants
15:34
Speaker 2: and sort of guiding the audience in lighting. We spent
15:36
Speaker 2: a lot of time guiding the audience's eyes so that
15:39
Speaker 2: they're looking at the right thing. You imagine a film
15:42
Speaker 2: like Wally where there's no dialogue for the first thirty
15:44
Speaker 2: minutes if you're distracted bold.
15:47
Speaker 1: Old I couldn't believe it what I thought.
15:49
Speaker 2: It was very scary. When we were working on it.
15:51
Speaker 2: We're like, oh, it's going to go see this movie.
15:53
Speaker 2: But you know, if we leave some shiny thing off
15:55
Speaker 2: in the corner and you're looking over there, you might
15:57
Speaker 2: miss an entire chunk of the setup of the movie
16:00
Speaker 2: because that was all visual storytelling for thirty minutes and
16:02
Speaker 2: so so so there's a lot of just compositional things
16:07
Speaker 2: that we need to be able to do that aren't
16:08
Speaker 2: necessary early real world things, but we still need to
16:12
Speaker 2: do them. And you know, it also kind of invites
16:15
Speaker 2: the like we have these conversations sometimes that are like, well,
16:19
Speaker 2: why are we making this movie as an animated film,
16:22
Speaker 2: Like you could just go make this a live action movie,
16:23
Speaker 2: and like why we're not taking advantage of the medium,
16:26
Speaker 2: Like they're just different art forms, and like how do
16:28
Speaker 2: you make animations sing in an animation like you can
16:31
Speaker 2: do anything and so take advantage of it sort of.
16:33
Speaker 2: And so so I think we're in a better place
16:36
Speaker 2: now in terms of like there's all this like visual
16:39
Speaker 2: gain and speed gains by mimicking some of real life,
16:42
Speaker 2: but we've started to roll back in a lot of
16:44
Speaker 2: the ability to sort of maneuver it artistically too. We're
16:48
Speaker 2: not I don't know that we're totally there. Sometimes it's
16:50
Speaker 2: still really hard, but I think we're getting to a
16:53
Speaker 2: place that at least makes me happy. It might not
16:55
Speaker 2: make that engineer happy at all, but it makes me happy.
16:58
Speaker 1: So well, well that's the like you're telling the sort
17:01
Speaker 1: of Goldilocks story where like at first just you were
17:05
Speaker 1: limited by computing power and the lack of sophistication of
17:08
Speaker 1: the suffware because we're just figuring ou how to do this,
17:10
Speaker 1: And then there was a moment when you had sort
17:11
Speaker 1: of gone too far into just mimicking reality. And now
17:16
Speaker 1: you're in this sort of just right phase.
17:18
Speaker 2: Although you know it's never going to be just right, yeah,
17:20
Speaker 2: because you're always I mean, I'm sitting on a movie
17:22
Speaker 2: right now where it's like, how on earth are we
17:25
Speaker 2: going to do that? Like it's not clear, and our
17:27
Speaker 2: system isn't made to do that. So what are we
17:29
Speaker 2: going to do? You know? And so well, let's talk
17:31
Speaker 2: about this like this.
17:33
Speaker 1: So the sort of frontier questions that you're bringing up,
17:36
Speaker 1: First of all, can you talk about the movie that
17:37
Speaker 1: you're working on?
17:38
Speaker 2: I can't. It's totally secret.
17:40
Speaker 1: I love secret. So you're working on a secret Pixar movie. Yeah, Like,
17:43
Speaker 1: what's the thing that you're working on right now that
17:46
Speaker 1: you haven't figured out.
17:48
Speaker 2: In a general way? So there's this great thing happening
17:51
Speaker 2: within computer animation right now, I would say, which is
17:54
Speaker 2: the original Spider Verse came out.
17:58
Speaker 1: The Spider Verse that came out a couple of years ago.
17:59
Speaker 2: Yeah, and it was loved it, respectfully.
18:01
Speaker 1: I know it's not a Pixar movie. Blew my mind.
18:04
Speaker 1: I loved the New one. Went to the New One
18:06
Speaker 1: a couple of years, a couple of weeks ago at
18:08
Speaker 1: the theater in the old school style.
18:10
Speaker 2: Awesome, like high art right, totally Yeah, incredible, incredible movie.
18:17
Speaker 2: And I think both of those movies, the first one
18:19
Speaker 2: and now the second one again really inspired almost the
18:24
Speaker 2: entire industry.
18:25
Speaker 1: Huh.
18:26
Speaker 2: I mean, I don't even know if it's almost I
18:27
Speaker 2: think inspired the entire industry because as you look at
18:29
Speaker 2: the films that have come out since then, everybody, there's
18:34
Speaker 2: so many instances of stylization that were not there before.
18:37
Speaker 1: And yeah, I mean, let's be clear when you say
18:39
Speaker 1: stylization for people who haven't seen the Spider Verse movies. Yeah,
18:42
Speaker 1: like they don't look anything like real life. They're not
18:46
Speaker 1: trying to look like real life. They look really interesting.
18:48
Speaker 1: They look kind of like comic books. In the case
18:51
Speaker 1: of the new movie, it's like pastiche. Sometimes there's like
18:53
Speaker 1: six different styles on the screen and it sort of
18:56
Speaker 1: goes to this thing you're talking about about moving away
18:58
Speaker 1: from just trying to represent reality, so go on, sorry.
19:02
Speaker 2: Yeah, no, no, no, So it's that first one, I think,
19:07
Speaker 2: because it was like anything anyone had seen in the
19:12
Speaker 2: same way. Maybe that Toy Story was when I went
19:15
Speaker 2: and saw it when I was in college of just
19:17
Speaker 2: like what is this and this is amazing? And then
19:19
Speaker 2: Spider Verse did that, and I'm not sure any of
19:22
Speaker 2: us at that point realized that we could do something
19:26
Speaker 2: that was that different and that new, and so it
19:28
Speaker 2: was really exciting because we're working in this art form
19:32
Speaker 2: and suddenly it's like, uh, Oh, that's amazing, someone did that.
19:37
Speaker 2: We want to do that, you know, and so and
19:39
Speaker 2: was part.
19:39
Speaker 1: Of it that like, oh, we're Pixar, and there's this
19:42
Speaker 1: thing that a Pixar movie is, and it's a particular thing,
19:45
Speaker 1: you know. I feel like, in a much smaller way,
19:50
Speaker 1: I have felt that on smaller projects I've worked on,
19:52
Speaker 1: where you sort of have this kind of internal grammar
19:55
Speaker 1: you're like, oh, this is a whatever, this is a
19:57
Speaker 1: plant money story, but nobody who's listening actually cares or
20:01
Speaker 1: even knows. Right, It's like you're just creating these rules
20:03
Speaker 1: that are binding yourself for no reason, right.
20:07
Speaker 2: I think the thing is that when I got onto
20:10
Speaker 2: Turning Read, the director was like the word in the
20:14
Speaker 2: hallway before I got on, It was the director wanted
20:16
Speaker 2: to do a different look, which was exciting to everyone
20:18
Speaker 2: here because it's just doing different things is always exciting.
20:20
Speaker 2: And she's at the time was twenty nine maybe, and
20:25
Speaker 2: so younger than some of the sort of old school
20:27
Speaker 2: directors that have been here, and so coming in with
20:29
Speaker 2: just sort of different references, and she grew up with
20:32
Speaker 2: anime and so was really inspired by anime and wanted
20:35
Speaker 2: it to have some reference to that and so there's okay,
20:38
Speaker 2: it's going to have a different look, and she said,
20:42
Speaker 2: I don't want it to have the Pixar look. And
20:43
Speaker 2: I had to ad this moment of like, what do
20:47
Speaker 2: you mean the Pixar look. We try so hard on
20:49
Speaker 2: every film to do whatever's right for the film, and
20:51
Speaker 2: then you sort of go, well, yeah, I mean, I
20:53
Speaker 2: guess we do sort of have a Pixar look, but
20:55
Speaker 2: it's not intentional. It's sort of like we all just
20:57
Speaker 2: grew up here figuring out how to make movies together
21:01
Speaker 2: a lot in this sort of core group of folks,
21:04
Speaker 2: and we use our own rendering system and half our
21:07
Speaker 2: software is our own software, and so like there's something
21:10
Speaker 2: that comes from that. And I think we were just
21:12
Speaker 2: always pushing to be able to do sort of bigger
21:15
Speaker 2: and better things depending on whatever the story was, and
21:18
Speaker 2: not so much thinking about stylizing stuff. And so here's
21:23
Speaker 2: this director saying I want to, I want to, I
21:25
Speaker 2: don't want to do the Pixar look, and I'm like,
21:26
Speaker 2: oh my god, I don't even know what the Pixar
21:28
Speaker 2: look is. Let me think about this. I've been in
21:30
Speaker 2: the middle of it for so long that I didn't
21:31
Speaker 2: even realize it sort of and so kind of looking
21:35
Speaker 2: at how that look comes about and then what do
21:39
Speaker 2: you need to do to break out of that and
21:41
Speaker 2: what It was this really interesting thing for me since
21:44
Speaker 2: I've been here so long, of like what are the
21:46
Speaker 2: rules that you're following that you don't even realize are
21:50
Speaker 2: out of date because some of them were based on
21:53
Speaker 2: old technology, Like, oh, we never do that, and then.
21:58
Speaker 1: We never do that because you had because previously the
22:02
Speaker 1: technology had not allowed that.
22:03
Speaker 2: Yeah, it looked terrible, but maybe now with the technology
22:07
Speaker 2: it wouldn't look terrible. Or I'll give example. So I
22:09
Speaker 2: was doing some looks tests early on before a lot
22:12
Speaker 2: of people were on, and we were trying to get
22:15
Speaker 2: this this really this the lighting look in there, the
22:17
Speaker 2: sort of if you think about light to dark, there's
22:21
Speaker 2: not a lot of dark stuff in that movie. In
22:23
Speaker 2: the first act of it or something. It's very the
22:26
Speaker 2: mids are very lifted kind of. So I'm trying to
22:29
Speaker 2: figure out how to do that without doing a lot
22:31
Speaker 2: of like you know, photoshop type work on it.
22:36
Speaker 1: And that was like a particular creative sort of storytelling
22:39
Speaker 1: desire is to have the color palette be in a
22:41
Speaker 1: certain way.
22:42
Speaker 2: Yeah, to mimic sort of the main character May's sort
22:46
Speaker 2: of world, and then it changing to be much more saturated,
22:50
Speaker 2: much more contrasting, and getting kind of crazy when kind of.
22:53
Speaker 1: Like Kansas being black and white, but in a more
22:56
Speaker 1: subtle way. Yeah.
22:57
Speaker 2: Yeah, And so I I did some stuff, but the
23:01
Speaker 2: way I was doing it was like not working. I'm
23:03
Speaker 2: doing ten other things while I'm doing it. And then
23:05
Speaker 2: we got our director of photography for lighting. He comes
23:09
Speaker 2: on and he sits down, he's doing some tests and
23:12
Speaker 2: he shows something. I was like, that's awesome. What did
23:14
Speaker 2: you do. He's like, oh, I put a light from
23:16
Speaker 2: camera on everything. So a light blasting straight into camera,
23:20
Speaker 2: which we would never do because it looks flat, it
23:24
Speaker 2: flattens everything out, looks kind of stupid. But in this
23:26
Speaker 2: world and in this way, he totally made it work.
23:29
Speaker 2: And it isn't how we ended up doing it for
23:31
Speaker 2: the movie, but it was that was the moment where
23:33
Speaker 2: I was like, oh, I think I'm going to have
23:35
Speaker 2: to let go much more than I thought I did
23:38
Speaker 2: in order to like really find the ways that we
23:40
Speaker 2: can push stuff to find a new look.
23:42
Speaker 1: As you realize that in fact, there were a bunch
23:44
Speaker 1: of things that you did the same without thinking about
23:46
Speaker 1: and there were other ways to do it that didn't
23:48
Speaker 1: even cross one hundred.
23:49
Speaker 2: Percent, And so then it was that really was the
23:51
Speaker 2: sort of pivotal moment where I then was like, Okay,
23:55
Speaker 2: anytime your reaction is like, yeah, we don't do it
23:57
Speaker 2: that way, or there's good reason, you have to reconsider
24:00
Speaker 2: that and decide whether that's actually still relevant. Ory just
24:05
Speaker 2: it's just sort of this long held thing.
24:09
Speaker 1: In a minute, with video games getting better and movies
24:13
Speaker 1: using more and more CGI, are we heading to a
24:15
Speaker 1: world where more or less everything is computer animation? Now
24:27
Speaker 1: back to the show. You know, you've talked about all
24:29
Speaker 1: these ways over time in which you, Slash Pixar have
24:35
Speaker 1: solved technological problems to allow you to do more things creatively.
24:40
Speaker 1: And I'm curious now about that particular dynamic, whether it
24:44
Speaker 1: is for purposes of stylization or for purposes of just
24:48
Speaker 1: going into a world where computer animated films have not gone.
24:52
Speaker 1: Are there things where you all you Pixar are trying
24:57
Speaker 1: to figure out to solve technological problems so that you
25:00
Speaker 1: can do new creative things.
25:04
Speaker 2: You know, right now we have conversations about like, well,
25:07
Speaker 2: that's a lot of hair, that's going to be hard
25:09
Speaker 2: general final image, or oh glass is expensive, or that
25:12
Speaker 2: translucent thing could be hard, or that animation is going
25:17
Speaker 2: to be really tricky, like Hank the octopus in Finding
25:21
Speaker 2: Dory very hard. Eight Tentacles turns out very hard doing
25:25
Speaker 2: a room challenge. Yeah, and so so there's these things
25:29
Speaker 2: where they're like, it's not like you couldn't do it,
25:32
Speaker 2: but they're annoying to do, so you don't see them
25:34
Speaker 2: very often, and so at a.
25:35
Speaker 1: Certain and expensive basically.
25:38
Speaker 2: And kind of tedious and like. And so at some
25:41
Speaker 2: point a lot of that hopefully is not tedious anymore,
25:45
Speaker 2: and so that opens up different stories in different worlds,
25:49
Speaker 2: and maybe it opens up that, you know, like now
25:52
Speaker 2: everyone has a camera, they could shoot anyone can shoot videos.
25:55
Speaker 2: I mean TikTok is proof of that, right, maybe anyone
25:58
Speaker 2: can make you know, animated films in a totally different way.
26:03
Speaker 2: I don't know. You still have to tell great stories
26:06
Speaker 2: and so that's that's still going to always be the
26:09
Speaker 2: main building block of it. But faster and easier is
26:15
Speaker 2: always the thing we are always sort of going after that.
26:19
Speaker 1: So that's this sort of there's this sort of bottom
26:22
Speaker 1: up convergence, right, I mean, there's another convergence that's interesting
26:27
Speaker 1: to me when you're talking about, like, you know, this
26:31
Speaker 1: sort of arc of first trying to make films that
26:34
Speaker 1: look more like reality and saying, oh wait, wait, that's
26:36
Speaker 1: not actually the business we're in. Let's make things that
26:39
Speaker 1: are you know, beautiful and interesting and look like nothing
26:42
Speaker 1: you've ever seen. I feel like, to some significant degree,
26:48
Speaker 1: over the decades you've been working at Pixar, live action
26:50
Speaker 1: films have started to look more like animated films, right Like,
26:54
Speaker 1: I mean, if you think of Marvel, which is probably
26:56
Speaker 1: the most important thing to happen in movies, at least
26:59
Speaker 1: on a commercial level and in some ways on a
27:01
Speaker 1: creative level over the last whatever ten years, Like you know,
27:05
Speaker 1: there's like whatever, a talking raccoon and a person, right Like,
27:10
Speaker 1: they are not going for reality and they are live
27:13
Speaker 1: action films, but they are you know, CGI is basically
27:18
Speaker 1: computer animation, and so right like, maybe there is some
27:21
Speaker 1: convergence where it's everything's made with a computer. So whether
27:25
Speaker 1: we're calling it animation or like live action, like Marvel
27:28
Speaker 1: movies aren't exactly live action, right It's like a dude
27:31
Speaker 1: doing motion capture in front of a green screen.
27:33
Speaker 2: Yeah, yeah, it's everything is overlapping and that, you know,
27:38
Speaker 2: and a lot of the live action stuff they're using
27:42
Speaker 2: all of the same stuff we do. There is no
27:44
Speaker 2: difference except that they're shooting some live action plates that
27:46
Speaker 2: they're matching it to, you know, in a lot of ways,
27:48
Speaker 2: or they're doing virtual production.
27:51
Speaker 1: Avatar maybe even more so, right, Yeah, Avatar is kind
27:56
Speaker 1: of a computer animation movie. Yeah.
27:58
Speaker 2: Yeah. And so you think about things like the Academy Awards,
28:02
Speaker 2: the animation category that's getting hazier and hazier, like what
28:07
Speaker 2: is animation and what is live action? And you know,
28:11
Speaker 2: especially with some of the things like live action Lion
28:15
Speaker 2: King or Jungle Book or different things like that where
28:19
Speaker 2: there is really a huge mix. And you talk about
28:22
Speaker 2: games and that same thing and those the visual level
28:26
Speaker 2: of games has increased coming towards us. It used to
28:30
Speaker 2: always be that we were we would spend a lot
28:33
Speaker 2: of time on our renders to make them look beautiful.
28:35
Speaker 2: Games had to be so fast that there's no way
28:37
Speaker 2: that they could approach sort of the look we had.
28:39
Speaker 2: And that's coming together and live actions coming together, and
28:42
Speaker 2: so everything's becoming much more possible, I think, on all
28:46
Speaker 2: those fronts. And so maybe at some point it all
28:49
Speaker 2: is just one one pie that everyone's pulling from.
28:54
Speaker 1: Does it make you sad? Is that a sad ending
28:57
Speaker 1: for you?
28:59
Speaker 2: It's not sad to be able to think about being
29:02
Speaker 2: able to work, being able to try anything and have
29:06
Speaker 2: no cost to it, so that you could try this,
29:08
Speaker 2: you could try that, and try this and throw it
29:09
Speaker 2: out and do a different thing and not feel precious
29:13
Speaker 2: or worried about it. That's very exciting.
29:17
Speaker 1: Uh huh. Yeah, the cost constraint. I mean it's funny.
29:22
Speaker 1: The editor of this show was like, yeah, ask about
29:24
Speaker 1: cost constraints. I was like, no, let's not do cost coinstints.
29:26
Speaker 1: So then you've brought it up, so points to her.
29:29
Speaker 1: I mean it's interesting, right because there is one of
29:32
Speaker 1: the simple and I think underrated things technology does. Maybe
29:37
Speaker 1: the fundamental thing is make things cheaper, right, And you're
29:42
Speaker 1: talking about that from the creative point of view, Right,
29:44
Speaker 1: If it's cheaper to just try seven different things, or
29:47
Speaker 1: cheaper to make an octopus or whatever, then you can
29:49
Speaker 1: try more stuff and make more stuff.
29:51
Speaker 2: Yeah. So on every movie you look at a thing
29:54
Speaker 2: and you go, okay, am I going to build a
29:57
Speaker 2: system for that that makes it easier to do in
29:59
Speaker 2: a bunch of shots or are we just going to
30:01
Speaker 2: like kind of force it through, right, and we'll just
30:04
Speaker 2: hodgepodge it in every shot to get it through. And
30:06
Speaker 2: so in something like if it's a little bit tricky
30:08
Speaker 2: and it's intense shots, you're just going to kind of
30:11
Speaker 2: make it happen and force it through, you know, kind of.
30:14
Speaker 1: If it's water for finding Nemo, you're going to build
30:16
Speaker 1: a water.
30:17
Speaker 2: You're going to build something that is really usable that
30:20
Speaker 2: people can make it look great. And so there's there's
30:23
Speaker 2: sort of always that trade off and consideration of how
30:27
Speaker 2: you're investing your sort of time and money.
30:29
Speaker 1: And Pixar started as a tool company, right, basically a
30:33
Speaker 1: hard company, and there's this software render Man that is
30:36
Speaker 1: still like there's like a website and you can like
30:38
Speaker 1: pay and use the software that picks our users to
30:42
Speaker 1: make movies. Right, it is at some level a tools company.
30:44
Speaker 2: It is. Yeah, it's we're a funny place in that
30:47
Speaker 2: where where we have our own studio tools group that
30:51
Speaker 2: write are a bunch of software that we use. We
30:53
Speaker 2: have the RenderMan group that writes rendering software that's used
30:55
Speaker 2: throughout the industry. We make movies, but we're not totally Hollywood.
31:01
Speaker 2: We're not totally a tech company. We're some weird combination
31:04
Speaker 2: in between. And then we're in the Bay Area, so
31:06
Speaker 2: we're not even like down in Hollywood or down in
31:08
Speaker 2: Silicon Valley. And so it's an interesting place that I
31:12
Speaker 2: think has bread a lot of interesting things in terms
31:14
Speaker 2: of the combination of sort of art and technology here,
31:17
Speaker 2: and the interest in driving computer graphics to make our
31:20
Speaker 2: movies to suit the story, and then being inspired by
31:24
Speaker 2: technology to then change the story to take advantage of it.
31:27
Speaker 2: And so there's this really cool kind of feeding back
31:30
Speaker 2: and forth and advancement of technology because we don't put
31:34
Speaker 2: limits on story in the way you might otherwise. So
31:38
Speaker 2: when you know they storyboard a character with long hair
31:40
Speaker 2: and we don't know how to do that, we still
31:43
Speaker 2: go after it and do it. And that's that's an
31:47
Speaker 2: amazing place to be, I think.
31:49
Speaker 1: And then you not only have that character presumably in
31:52
Speaker 1: that instance, you have now a way that, oh, great,
31:54
Speaker 1: people in these worlds we create can now have long hair.
31:57
Speaker 1: That's another thing we can do.
31:58
Speaker 2: Yeah, totally, until we got to Brave, and then long
32:01
Speaker 2: curly hair was its whole own problem. We have to
32:03
Speaker 2: rewrite the simulator, So.
32:07
Speaker 1: Does it have to be curly?
32:09
Speaker 2: It's always something.
32:11
Speaker 1: Yeah, we'll be back in a minute with a lighting
32:17
Speaker 1: focused lightning ground. Back to the show. Okay, I'm gonna
32:31
Speaker 1: ask you a bunch of questions. Edith said, you listened
32:33
Speaker 1: to some episodes of the show, so I don't even
32:35
Speaker 1: need to explain that this is the lightning round.
32:40
Speaker 2: But I didn't prepare it.
32:41
Speaker 1: Also, Okay, there's no way to prepare the best part
32:44
Speaker 1: of the show. For that reason, I'll start with an
32:47
Speaker 1: easy one that you were totally prepared for. I know
32:51
Speaker 1: from listening to some other interviews you've done that you
32:55
Speaker 1: have often been the only woman in a room full
32:57
Speaker 1: of men at work. What is one tip you'd give
33:02
Speaker 1: to somebody who is in that setting, you know, the
33:05
Speaker 1: only woman in a room full of men at work.
33:12
Speaker 2: I found when I became a director of photography for
33:15
Speaker 2: lighting for Wally that I was I was generating my
33:22
Speaker 2: own imposter syndrome. So I was getting quieter and quieter
33:25
Speaker 2: because I was suddenly in a room with sort of
33:27
Speaker 2: a new level of people, and I kept thinking, well,
33:30
Speaker 2: would I have known that? Would I have been able
33:32
Speaker 2: to answer that? And there was this moment where I
33:35
Speaker 2: realized that my boss was particularly outspoken, and he suddenly
33:40
Speaker 2: was talking about something about lighting, and I actually knew
33:43
Speaker 2: he was wrong, but he was saying it as if
33:44
Speaker 2: it was fact, and I sort of suddenly realized that
33:48
Speaker 2: just because someone is presenting something as fact doesn't necessarily
33:52
Speaker 2: mean that that's what that is, and that I was
33:55
Speaker 2: sort of like sidelining myself over something that was completely false.
33:59
Speaker 1: So what did you do? What did you do?
34:02
Speaker 2: My jaw dropped, and then I spoke up because I
34:05
Speaker 2: knew he was wrong, right, And so I was emboldened
34:08
Speaker 2: in that way to say, wait, wait, wait, that's that's
34:11
Speaker 2: not true, and or I'm sure I probably found some
34:14
Speaker 2: nice way to say it, so I wasn't poking the bear.
34:16
Speaker 1: But yes, and yeah, yeah, yeah yes. And what's your
34:20
Speaker 1: favorite non Pixar movie? Mm hmm, I'm gonna do another one.
34:26
Speaker 1: I got a better one. Okay, what's the best lighting
34:28
Speaker 1: you've ever seen in a non Pixar movie?
34:30
Speaker 2: Oh? Uh, you know what? The original Blade Runner I
34:40
Speaker 2: love so so much for many reasons. Lighting is wonderful.
34:45
Speaker 2: The newer Blade Runner is fabulous.
34:49
Speaker 1: Yeah, it's interesting when you mentioned Blade Runner, I think
34:51
Speaker 1: of the that relationship between live action and animation. And
34:57
Speaker 1: I mean, if you want to like be warm and
34:59
Speaker 1: fuzzy about it. This is the relationship between like what's
35:01
Speaker 1: real and what's not real, what's made by a computer
35:03
Speaker 1: and what's yeah, yeah, thematically nice. What's one tip for
35:09
Speaker 1: my living room?
35:11
Speaker 2: Oh? I like that? No bear bulbs?
35:16
Speaker 1: Thank you?
35:17
Speaker 2: Yeah?
35:19
Speaker 1: Am I getting interrogated?
35:20
Speaker 2: Yeah, you know, warm light to me always?
35:26
Speaker 1: Yeah. You think you'll ever leave Pixar?
35:30
Speaker 2: It's hard to imagine. It feels like it's my other family. Uh.
35:34
Speaker 2: And I have so much sort of knowledge invested here,
35:37
Speaker 2: and I love this place, and I love getting to
35:41
Speaker 2: work on our films and I love what we do.
35:48
Speaker 1: Danielle Feinberg is a visual effects supervisor at Pixar. Today's
35:53
Speaker 1: show was produced by Gabriel Hunter Chang and Edith Russolo.
35:56
Speaker 1: It was edited by Sarah Nix and engineered by Amanda K. Wong.
36:01
Speaker 1: You can email us at problem at Pushkin dot fm.
36:05
Speaker 1: You can find me on Twitter at Jacob Goldstein. I'm
36:08
Speaker 1: Jacob Goldstein and we'll be back next week with another
36:10
Speaker 1: episode of What's Your Problem? MHM.