Episode Transcript

This is a computer-generated transcript. While our team has reviewed it, there may be errors.

Morgan Sung: So if you’ve played any video game released in the last few years, you may have noticed this one hairstyle that keeps showing up on Black characters. It starts with a fade, and then at the top, there are locs that are kind of swooped forward and pushed to one side. Fans have a name for it: the “Killmonger locs”. They’re named after Erik Killmonger, the villain played by Michael B. Jordan in the 2018 Black Panther movie.

Michael B. Jordan: I found my daddy with panther claws in his chest! You ain’t the son of a king, you a son of murderer!

Morgan Sung: The locs might be a natural dark brown, or a shocking white, or feature frosted tips. There are some variations from game to game, but the base hairstyle is relatively the same. It’s taken over animation in video games and in movies. There’s Miles Morales from the PS5 game Spider-Man 2.

Miles Morales: I can’t talk about me without talking about Spiderman.

Morgan Sung: Or Ekko from the Netflix series Arcane.

Ekko: Sometimes taking a leap forward means leaving a few things behind.

Morgan Sung: Or Eddy Gordo, a character from the fighting game series Tekken, whose long-time ponytail was recently swapped out for… yep, Killmonger locs.

Tekken Game 8: NEW CHALLENGER!

Tekken Game 8: EDDY GORDO!

Morgan Sung: All of these characters are Black. And online, people are getting sick of seeing the same style over and over again.

TikTok: You would think that this style is so popular amongst Black men and Black people in general, you would think it’s trending, right? But it’s not.

Morgan Sung: All of this caught the attention of A.M. Darke, an artist and professor at UC Santa Cruz, who specializes in game design and new media.

A.M. Darke: I think it became the default hairstyle because we hadn’t really seen much um interesting sort of contemporary styles um represented in media.

Morgan Sung: She’s a longtime advocate for more diversity in games, particularly when it comes to animating Black hairstyles.

A.M. Darke: And so when Michael B. Jordan appeared on screen as that character, it was sort of a wow moment because he’s very handsome and, you know, he was this like cool, you know, uh figure. And so I feel like people just felt like, “Oh, here’s finally a fourth Black hairstyle that we can use,” which points to us just not having a lot of diverse Black hair representation in media in general.

Morgan Sung: When A.M. calls Killmonger locs “a fourth Black hairstyle,” she’s not exaggerating.

A.M. Darke: The other three are the gigantic disco afro, the always wrong cornrows, like they’ve been shaved into someone’s head instead of actually braided, and then the messy, unstyled locs.

Morgan Sung: And none of these styles involved curls, a feature long overlooked in computer animation. A.M. is a game designer and often relies on pre-made assets to design her games. But the assets for Black characters that she found weren’t just inaccurate — they were straight-up offensive, showing animals and racist caricatures. And so, A. M. launched the Open Source Afro Hair Library in 2019. It’s a database for 3D models, specifically featuring different hair textures and styles submitted by Black artists around the world.

A.M. Darke: One, I wanted to make sure that there was a space where we could find authentic Black representation, but I also didn’t wanna sort of populate these already problematic platforms with sort of beautiful representations of Blackness sitting side by side with derogatory representations. The other half of this work is actually creating new technology that understands Black hair to be the premise and to treat Black hair as the default as opposed to some outlier.

Morgan Sung: For context, the Sims 4, which launched in 2014, literally used cauliflower as a reference for afros in the game’s concept art. And you could tell.

The Sims 4: Skeeba yamoob!

Morgan Sung: When the game launched, that hairstyle looked exactly like the stiff, comically large disco afro that A.M. described. This isn’t just a Sims problem — It’s an industry-wide issue because the existing animation technology was developed with straight hair in mind. So why is very curly, textured hair so hard to animate in the first place? Why hasn’t the field of computer graphics gotten it right yet? And how is new research starting to change that — not just making it easier to animate, but allowing artists to depict more realistic, resonant characters on screen.

Morgan Sung: This is Close All Tabs. I’m Morgan Sung, tech journalist, and your chronically online friend, here to open as many browser tabs as it takes to help you understand how the digital world affects our real lives. Let’s get into it.

Morgan Sung: Let’s open our first tab. The duo behind the research.

Theodore Kim: I saw the first Toy Story movie when I was in high school. And I said, “This is amazing. I didn’t know that computers could make entire movies. How can I get a job in this?”

Morgan Sung: Theodore Kim is a computer science professor at Yale University, and he basically researches the physics involved in animation. He’s also A.M. Darke’s co-author on two research papers focused on how to accurately animate Black hair.

Theodore Kim: I’ve been doing research in it since um 2001, so for almost a quarter century now.

Morgan Sung: Theodore actually has two Academy Awards for his work on animation technology. In 2015, he was recruited as a senior research scientist at Pixar. This was a huge deal for him. It’s the studio behind the movie that inspired him to pursue animation. And before that, Pixar was one of the pioneers of computer graphics.

Theodore Kim: Joining this, this very storied institution was a very enticing thing for me.

Morgan Sung: Theodore is Korean-American and immediately stood out at Pixar.

Theodore Kim: The thing that was really surprising when I got there was it was so white. So over the course of the four years that I was there, uh, it was really difficult to ignore the question of who are the people that made these tools and, and what were they going for? It became very clear that having a highly curly hair is not something that people were aiming for when they designed these algorithms.

Morgan Sung: In 2020, after the murder of George Floyd sparked Black Lives Matter protests around the world, Theodore started thinking more about how discrimination is built into the animation industry. In addition to seeing the racist caricatures that A.M. saw, Theodore realized that Black artists also didn’t have animation tools that could adequately represent them, especially when it came to hair texture.

Theodore Kim: Everyone’s familiar with skin discrimination. People are not familiar, that familiar with hair discrimination.

Morgan Sung: So, Theodore joined Twitter to try to publicize an op-ed about how the industry standard for animation software fails Black creators. While scrolling through animation Twitter, he came across a viral post about an artist who launched a database for Black hairstyles. That’s how he found A.M.’s Open Source Afro Hair Library.

Theodore Kim: And I said, oh my gosh, someone else gets it. So I reached out to her, and I said, “You want to write a grant proposal together?” And thankfully, A.M. said, “Yeah, sure, sure. Sounds interesting.” And on the very first Zoom call that I had with her, sorry, I’m talking about you in the third person. You’re right here.

A.M. Darke: That’s okay, I don’t remember any of this, so I’m excited.

Morgan Sung: Oh yeah, we got them both on a call together to talk about their collaboration.

Theodore Kim: So on the very first Zoom call, you made a reference to Type 4 hair. And I remember that I interrupted you and I said, “Wait, wait, what’s that mean? What is Type 4 hair?” And that is the moment that I learned what the term Type 4 hair was. So the collaboration sort of blossomed from there.

A.M. Darke: In his first message, it was very clear, like, he was very impressive. And I was like, “Wait, you want to do what? You’re where? You’re studying what? Oh, I don’t know anything about computer graphics.” Like, I mean, I do now, but I’m just still thinking from the arts and design side, right? I know I’m working with computer graphics, but I’m not thinking about the technical aspects. When Ted is like, “Oh, wait. I don’t know anything about this. What does that mean? What does that mean, what does that mean?” I’m like, “Oh, wow. Actually, I can tell you so many things. OK, here’s a here’s your reference here. OK.” I remember being like, “Ted, do you have Instagram? I need to share some Instagram accounts with you. Like, it’s it really shows the phenomenon I’m explaining.” I went to visit Yale recently and like, there’s like wigs in the lab. And that’s like so funny.

Morgan Sung: Wigs everywhere.

A.M. Darke: It’s like it’s like, “Oh, here, here is the curly hair sample here. Here’s this curly hair sample.” No, it was interesting to, you know, because we started this work like very early pandemic. And so we’ve been, you know, communicating, collaborating remotely. And so a lot of it has been this back and forth, including being like, “OK, well, how do how can I demonstrate what hair does? Oh, my my curls are doing a thing today. Let me quickly get a photo.” I remember, you know, like asking my partner to like, “Okay, can you take slow motion videos of my hair as I jump around and so I can explain to Ted how the hair should move and fall.” And so it’s been interesting because on one hand, the work is so precise at the mathematical level, but on the other side, I’m just sort of chaotically like, “Oh, okay, so here’s a glint of light, stand in the sun, and my partner’s just constantly getting harassed because he has Type 4 hair. And so I’m like, “Oh, oh your curl is doing the thing. It’s really, it’s really clumped today. Oh, now it’s diffuse. I need to do it before and after. Can I, can I have that hair?” And it’s been a lot of sending photos.

Morgan Sung: You took his hair?

A.M. Darke: To Ted. Oh of course I took his hair. Friends and family are, are, yeah, yes, their, their hair is here. Um, yeah and in my own hair samples. And so it’s really interesting to, um, just see all the, the other touch points that this research, um ends up revealing.

Morgan Sung: What was your experience like, Ted? Were you also taking people’s hair?

Theodore Kim: No, no. So what A.M. says is so interesting because what you take for granted of, you know, if you ask yourself, “Okay, so what does the hair do?” It’s like you can just like look at it. And, you know in our lab, it’s like me and like two other Asians and one white student. And like we look around and we’re like, we do not have access to this data at all. Over the last, you know, 50 years, whenever people said, “I’m going to make a hair algorithm.” What do they do? They like just pull their own hair down. They’re like, “Oh yeah, like, like this, like this, right?” And this is how we ended up where we are. Is because none of those people had had highly coiled hair.

A.M. Darke: Something that’s been so rewarding for me is recognizing how much, just the value of cultural knowledge, especially to be, you know, to be Black and to recognize, like, there’s such depth to, you know my own culture that most people are not asking questions about. I don’t always have the opportunity to explain, like, the phenomena, like that are in the paper. And so all of that back and forth is actually really, has been really fun.

Morgan Sung: So what are the technical limitations artists face when animating textured hair? Let’s open a new tab. Why can’t animators get black hair right?

Morgan Sung: Animating hair isn’t as simple as plopping a hairstyle on top of a character, at least when it comes to computer animation. Straight hair is the default, both on the artistic side and on the technical side. In the fields of computer graphics, the way everything moves and reflects light and interacts with the objects around it, it’s all based on math. Theodore said that the industry software starts out by treating hair as a tube. Animators sculpt the tube, kind of like clay, and that acts as the foundation. Instead of sculpting hundreds of thousands of tubes for each individual strand of hair, animators would sculpt a couple dozen tubes.

Theodore Kim: The assumption here is, is like your hair is a tube, hair is straight, right?

Morgan Sung: To make sure the hair moves naturally, animators at Disney added the elastic rod model. It’s kind of like a rope. It can twist and compress with gravity and then stretch out again. Animators were able to make these rods wavy, but curly hair, especially the textures featured in the afro hair library, is fundamentally different. It isn’t made up of waves at all. It’s more like a three-dimensional coil, like a spring. The math formulas used to animate hair weren’t designed to maintain that shape. So when you try to force these straight rods to coil like that.

Theodore Kim: It’s either like really, really slow or it’s unstable. So it just like explodes.

Morgan Sung: Basically, the rods act unpredictably because they were never designed to function in that way. And then there’s this other problem. It has to do with something called pre-visualization. Animation is notoriously expensive and time-consuming, so artists often use a process called pre-vis to save time. It’s a rough, low-fidelity version of whatever they’re animating, created before the computer spends hours rendering it in detail. Think of it like seeing a sketch before it’s painted. But because straight hair is the default, hair won’t appear curly until it’s fully rendered.

Theodore Kim: So you look at the pre-vis for some of these hairstyles, and it’s just like a few Medusa wires coming out of the head. And then only later do they add the curls. So when you’re actually trying to make your design, you can’t see the curls, you just have to imagine how this Medusa head is going to translate into black curls, which is a terrible way to live. And yet the tools don’t exist to live in any other way.

Morgan Sung: If you’re seeing the pre-visualization and you render it, you go through all the effort rendering, which takes forever, and it doesn’t turn out the way you want, do you just have to go back and do it again and hope that it works?

Theodore Kim: Yeah, you’re driving blind, yeah, you are driving blind. It’s like, “uh, maybe this will fix it and then time to wait again.”

Morgan Sung: Given these technical limitations of the existing software, the existing models, how does that further discourage animating different representations of textured hair?

A.M. Darke: I mean, I don’t know what Ted’s answer is, but my answer is budget. It’s always about budget and resources. So, you know, people want to take the path of least resistance and oftentimes, you know black hair gets left out because the tools were not built to make this process easy. And oftentimes you need someone advocating to say that it’s important to begin with. That it’s necessary.

Morgan Sung: Curly hair research has also faced a lot of resistance in academia. Take the 2012 Disney Pixar movie Brave, set in medieval Scotland. One of the most striking parts of the movie is Princess Merida’s hair, a bright red shock of curls that bounces throughout the story. At the time, that animation was revolutionary. The researchers who worked on Brave wrote a paper about the technical details behind the curly hair algorithm. Initially, scientists in the field panned the research, saying that the straight hair algorithm could do this already, which wasn’t true. With all of these technical issues in mind, and the uphill battle that this field of research faces, A.M. and Theodore set out to fix the root of the problem. Instead of working on a band-aid solution to make the current animation tools work for curly hair, they wanted to create an entirely new foundational model that includes curly and textured hair from the start.

A.M. Darke: What our collaboration is about, it’s like communicating this is the standard to meet because oftentimes people can treat Black hair or Black aesthetics as if it’s just, you know, “Well, anything is fine. You know, just be happy to be represented at all.” And it’s like, well, no, actually we have pretty like high standards about how we present ourselves.

Morgan Sung: So what came out of Theodore and Anne’s collaboration? That’s a new tab, which we’ll open after this break.

Morgan Sung: Okay, we’re back. Let’s dive into this collaboration between art and science in a new tab. Solving the curly hair problem. Instead of trying to force the existing math to work for curls, A.M. and Theodore started from scratch. In 2023, they published Lifted Curls, their first research paper together. After studying hair sample after hair sample, and doing a lot of math, Theodore’s lab developed a model for animating tight curls like A.Ms. They actually recreated the structure of A.M.’s curls, and from there, figured out how each coil might bend or stretch. The final renderings looked like a ball of hair, but with curls that looked just like A.M.’s.

A.M. Darke: Ted is like, “Yeah, I’m not sure about that.” And I’m like, “No, this is it.” And he’s like, “Huh.” And I’m like, “No, this is it, like, like I have like tears in my eyes, like this is like we did. We did something. I don’t know what we did, but we did we, we hit a target.” And so seeing that first, that was the piece that let me know that we were going to be able to like actually capture afro-textured hair. We might have to go smaller, which is taxing, but that particular phenomenon was like the, to me, like the essential component.

Morgan Sung: But in 2023, when they submitted their paper to the top conference in computer graphics, known as SIGGRAPH, it was rejected.

Theodore Kim: I’ve been submitting to the top academic conference on this topic in computer graphics since 2001, so for decades. And when we submitted this paper, it got, I believe, was one of the worst scores that we had ever gotten on a paper submission before. People hated it. And not only did they hate it, but it was clear that their imaginations couldn’t actually quite comprehended it.

Morgan Sung: Getting a research paper accepted at this conference isn’t just about academic clout. It shows the rest of the scientific community that certain topics are worth studying. Without that legitimacy from the conference, rejected fields of research might miss out on grants and other resources.

Theodore Kim: You get lots of anonymous reviews back from other experts in the field. One of the anonymous comments was, “This looks like an agglomeration of tubes rather than hair.” So it actually didn’t even look like hair to this person.

Morgan Sung: I mean, yeah, how did you feel about that?

A.M. Darke: I felt offended because it was my hair. And some of the comments that we have received throughout this period is like, “Did you consult with any Black people?” Or, “Does this even look like hair?” Or, “Is this an improvement over existing models?” And I’m like looking and I’m looking at my hair, I’m at the comments, I’m look at my long history, like 30 years of consuming games and like 15 years of making games and being like, “Yeah, this is better.” So it felt like people were just not acknowledging, you know, this as a breakthrough. So that one was particularly frustrating. And it’s like, you know, obviously, these things are anonymous. So, it’s not like they know our identities. But I also have to ask like “If you’re asking me, have we consulted any Black people, I wonder if there are any Black people in the room because I feel like it would be very obvious to any black person that, yeah, this is what curly hair looks like.”

Theodore Kim: I can’t say with 100% certainty, but I’m pretty sure with about 90% certainty that not a single one of the anonymous reviewers that actually gave all those negative comments, not a single one was Black. There’s maybe three, three or four Black researchers like in our field and the probability that one of them received this in order to provide comments, the probability is very low.

Morgan Sung: When you reject papers that are actually trying to push beyond the status quo, push beyond, you know, “white is default” definition of human features, then you certainly aren’t widening, you certainly are opening the door, you’re closing the door.

Morgan Sung: But A.M. and Theodore didn’t give up. If anything, the rejection spurred them to dive deeper into their research. So they went back to the lab. After figuring out how to recreate the shape and movement of A.M.’s hair, they wanted to tackle a more complicated question. What makes animated hair look real?

Theodore Kim: There’s a test that all these papers try to do where it’s almost like a simulated shampoo hair toss of like, “Oh, my luxurious blonde hair, I will toss my head and it will sway in this certain way.” So there’s all these paper out there where they do this test where it is obviously trying to simulate a shampoo hair-toss. So then we saw all these and said, “Well, OK, so where’s the Black hair shampoo hairtoss?”

Morgan Sung: Curly hair isn’t uniform. Each curl coils differently, clumps together differently, and doesn’t grow out of people’s heads as perfectly-defined spirals. In fact, these imperfections are what makes hair look more realistic. A.M. and Theodore wanted to figure out how to recreate that with math. So over the next year, A.M., Theodore, and his team identified three key components of Black hair textures. First, what they call “phase locking.” It’s the spongy layer at the base of the hair before you see any defined curls.

A.M. Darke: It’s a common thing to look at Black folks who have had straightened hair in Black community and say, “Oh, I’m going to look at the roots, and that’s going to tell me your lineage.” And it’s because at the roots where you can’t get, you know, a straightener that close, it’s like there’s a little bit of growth that, you, know, grows up and out. So you have that spongy matrix, you can brush it down, you do all sorts of things with it. But in especially natural hair, typically you have some layer that’s growing up first before it forms any kind of coil, and then the curls coalesce.

Morgan Sung: And then there’s “period skipping.” The frizz factor.

A.M. Darke: So just in layman’s terms it’s like the more the period skips that means the more hairs are actually going outside of the clumped curl. And so that’s what makes hair look frizzy that’s what makes curly hair expand and have volume

Theodore Kim: If you don’t have that, then everything looks like a super shiny copper coil.

Morgan Sung: And then there are switchbacks.

A.M. Darke: So switchbacks, like any curly haired person will know, you’ve got your curls all going in one direction and it feels great. And you’re like, oh, what a perfect little spiral curl. And then it just, at some point goes wonky. It goes left, it goes the other direction. And so you get a little bump the other way. And I always describe this for those who remember telephone cords. It’s like the telephone cord, if you stretch it out too much, then it would get like a little kink before it might go back into the spiral. Um, but again, with curly hair, you never know. Sometimes it never goes back. And then you gotta, you gotta do something about that. Um, so those are just the everyday features of Black hair that, um, without it, you have hair that looks machine made or like a wig or like a doll.

Morgan Sung: And with all of these components put together, animated afro-textured hair looks real and moves naturally.

A.M. Darke: No matter how much you style, you’re going to have some variation in the hairs that follow the curl. And when you add those features, it feels much more authentic.

Morgan Sung: A.M. and Theodore, and two PhD students who worked with them, compiled their findings into another paper, titled Curly-Cue. Instead of just rendering the hair by itself, growing out of a nondescript sphere, they also included examples of what it would look like on a real person. They demonstrated different hairstyles on an animated model, which was based on a real life author named Carvell Wallace. There was an animated version of Carvell with a buzz cut, Carvell with more grown-out hair. And then Carvell with long, nearly shoulder-length curls. The color shifted with different angles and each curl bounced like real hair, nothing like the stiff disco afro in The Sims 4. And when they submitted this paper at the end of last year to that same conference, it was actually accepted.

Theodore Kim: We got some of the highest scores that I’ve ever received. We just tried to make something new and said, there are all these geometric phenomena that appear in Black hair that it’s meaningless to ask what these are with straight hair, they just don’t appear at all, but there’s just really interesting shapes and stuff that appear on Black hair and let’s name them and let figure out how to compute them. And I think this was a little bit harder for people to resist.

Morgan Sung: Since their paper was published, AM and Theodore’s work has been hailed as landmark research in animation. Videos showing the way the model’s hair moved and sprung back went super viral. People have been really excited to see characters that look more like them in movies and video games, especially if those characters can wear more hairstyles than inaccurate cornrows or Killmonger locs.

Morgan Sung: There’s a rich field of unanswered questions here. And who wouldn’t want to embrace that? I have this hard time thinking of like afro-textured hair as an outlier, because like for me, it’s not about, “Oh my gosh, we need this representation. I just think this is a noticeable gap in our understanding of how human features work. And it just seems obvious that we would want to cover all bases. Like, why is that controversial? And why is something to push back against?” Because it’s going to benefit everyone. So often I have people, you know, as the research blows up or becomes popular on social media, you now, folks are like, “Oh, you, I’m white and I have curly hair.” I’m like, yeah, that’s, and you can’t choose an avatar that looks like you. And that’s the way the anti-blackness impacts you personally. I’m glad that you’ve been able to like identify that. Yeah. You know, it’s gonna benefit everyone, so why would there be any pushback?

Morgan Sung: A.M. and Theodore are just getting started. The examples they rendered for their paper featured unstyled hair. There’s still so much to research. Like, how can they use math to recreate the way hair clumps or stretches as it’s parted? How would each coil behave when that hair is gathered into a scrunchie or taken out from braids? A. M. dreams of getting animation software to the point where artists can animate long, tightly coiled Black hair as it is washed and styled. She envisions a scene of a Black hair salon.

Morgan Sung: This is the scene that I want. It’s like the brush going through the hair or the pick, you know, pulling the tension of the hair and then watching it spring back, right? Or having it taught and being able to see it braided. That is sort of my vision.

Morgan Sung: While they continue to dive into their research, we’re gonna close all these tabs.

Morgan Sung: Close All Tabs is a production of KQED Studios and is reported and hosted by me, Morgan Sung. Our producer is Maya Cueva. Chris Egusa is our Senior Editor. This episode was edited by Chris Hambrick and Chris Egusa. Jen Chien is KQED’s Director of Podcasts and also helps edit the show. Sound design by Maya Cueva and Chris Egusa. Original music by Chris Egusa. Additional music by APM. Mixing and mastering by Brendan Willard. Audience engagement support from Maha Sanad and Alana Walker. Katie Sprenger is our Podcast Operations Manager, and Holly Kernan is our Chief Content Officer. Support for this program comes from Birong Hu and supporters of the KQED Studios Fund. Some members of the KQED Podcast team are represented by the Screen Actors Guild, American Federation of Television and Radio Artists, San Francisco, Northern California Local. Keyboard sounds were recorded on my purple and pink Dustsilver K-84 wired mechanical keyboard with Gateron Red switches. If you have feedback, or a topic you think we should cover, hit us up at CloseAllTabs@KQED.org, follow us on Instagram @CloseAllTabsPod, or drop it on Discord. We’re in the Close All Tabs channel at Discord.gg slash KQED. And if you’re enjoying the show, give us a rating on Apple Podcasts or whatever platform you use. Thanks for listening!

News Daily Newsletter

The Bay

Emma’s Must-Sees

Videos from KQED Live

Donor-Advised Funds

animationanimation

Curls, Coils, and Pixels: Researchers Crack the Code on Black Hair Animation