Audio Description the Pixar Way

AI-based, synthetic audio description may have a place in some forms of accessible video content, but the artistry of the entirely human-produced audio descriptions Pixar produces set a creative standard no AI is likely attain, and that's all for the good. Meet members of the Pixar team behind excellence in audio descriptions.

[MUSIC PLAYING]

BRIAN: Audio Description the Pixar Way. Moderator, Tom Wlodkowski. Speakers, Anna Capezzera, Eric Pearson, Laura Post, and Christina Stevens.

THOMAS WLODKOWSKI: Thanks, Brian. Hello, everybody, and it’s great to be here at Sight Tech Global. My name is Tom Wlodkowski and I am the Vice President of Accessibility for Comcast. Welcome to our panel on Audio Description the Pixar Way. And so we got two major studios represented here today with Pixar, and of course, I’m representing Comcast and NBC Universal. So for those who may not be familiar with audio description, it’s narrated descriptions inserted into the natural pauses of program dialogue, and it could be TV or a film or even other types of videos that describes the key visual elements of the program. So action, scene changes, any kind of lower thirds that appear on screen. Visual elements that just bring the show or film to life for people who are blind or visually impaired. And of course, we know that with accessibility, when you make a product more inclusive, you build a better product for everyone. And of course, audio description also makes the experience better for many in the audience who are not just blind or have low vision. So we’re here with an esteemed group of panelists from Pixar and Deluxe. Deluxe does a lot of the audio description for Pixar and has done some description for us as well in the past. And so why don’t we do this to get started. Let’s do some introductions. And Eric, I think I’ll start with you over at Pixar, and we’ll just go round-robin here.

ERIC PEARSON: That sounds great. Thanks so much, Tom, and thanks for having us here at Sight Tech Global, very excited to be here. My name is Eric Pearson. I’m the Home Entertainment Product Supervisor at Pixar, which means that I shepherd all of our films out onto the small screen. And one of the things that I’ve been really passionately involved in since our first home video release where we did audio description on Up have been really involved in the process at Pixar and making sure that this audience has the same home entertainment experience as all of our other audiences. And I’ll kick it to Anna.

ANNA CAPEZZERA: Thanks, Eric. Hi, everyone, it’s great to be here, and thanks for having us. My name is Anna Capezzera and I’m the Director of US Audio Description Operations at Deluxe Media. I actually started in AD as a writer and was one of the writers on Up. So it’s a great full circle here with Eric. I’ve been working with Pixar for a long time. And so yeah, just excited to be here and talk more about this. Christina?

CHRISTINA STEVENS: Hi, everyone, thank you for having me and us. I’m Christina Stevens. I am the Writing Manager at Deluxe, and I oversee our staff writers as well as the stable of freelance writers In the process of writing the AD script and getting it through the process before recording, which brings us to Laura.

LAURA POST: I’m Laura Post and I am a narrator for audio description. I work a lot with Deluxe. I do a number of titles for Disney and Pixar. I’ve done a bunch of the Pixar shorts that have come out recently. And that’s my job. I guess back to you, Tom.

THOMAS WLODKOWSKI: Well welcome all of you, and it’s great to have you joining us here for this panel. Eric, let me start with you and have you talk to us a little bit about why is audio description so important to Pixar?

ERIC PEARSON: Yeah, thanks, Tom. I think the best way to encapsulate that is we’re storytellers. And from the beginning of telling any story, we think about who’s going to receive the story, what is our audience? And we want everyone in our audience to have an exciting, moving, compelling experience accessible to them. There are a number of us who have blind or low-vision family members or experiences and understanding of how to create this type of compelling experience for this audience. And it became really, really important to us when we realized how limited the availability was in the home entertainment market. So right around the time that Up was being released, we really wanted to lean into it. We really wanted to try and make it a priority not only for ourselves creatively, but also for our studio, for Disney. So we really pushed hard to have the audio description included on the disc, on the DVD and Blu-ray. And it was very well received, and it was so exciting for us and so exciting for the community that we really leaned into it 100%. So now 100% of our catalog, our shorts and features have audio description, and we’ve made sure that it’s all available on the Disney+ platform. We also have deep relationships with partner organizations such as LightHouse San Francisco, Guide Dogs for the Blind, and the Blind Babies Foundation. And we feel like we’re doing a great job, but we also want to spend time with those communities and have them tell us how we’re doing. So we’ll often have screen– well, pre-pandemic when we could actually get together, we would have screenings at Pixar, invite over as many folks as could make it, as many folks as could fit, and then spend time before the screening and after talking about what was important to them. What types of aspects of storytelling were missing from other audio description experiences. What could we do? What was special about animated films that needed to be described? Really trying to deeply understand the story telling aspects that we could bring to the table. And so in the beginning, way back when, when we first worked with Anna, it was with WGBH. And since then, the majority of our work has been with Deluxe. And we’ve developed an incredible partnership with them.

THOMAS WLODKOWSKI: That’s excellent. We’re going to get into that in a little bit. You mentioned WGBH. That was my first jumping-off point in an accessibility career. I worked with the descriptive video service, and that feedback you talked about is so important. I actually ran some federal grants where we were, at the time, 30-plus years ago, looking at how to more effectively describe kids programming. And we had a question– or the description was like a mouse-sized fruit bat. And we asked the kids how big a mouse was, and they had their arms way out wide, almost like the wingspan of an eagle. And I’m like, geez, if I ever see a mouse like that, I’m going to run. And so it just really came back to– we really need to be more tangible in the types of objects if we’re going to do any type of comparison. So it’s interesting to see all of this come full circle, and who knew 30 years later I would be moderating a panel and you’re talking about GBH, and now a lot more, including Deluxe, a lot more organizations are into the AD space, and that’s really excellent. That feedback remains at the forefront of how to do this. And I know at NBC, we’ve been doing live audio description around the Olympics since we have the broadcast rights till 2032, around the primetime broadcasts, we had live audio description. And bringing that feedback in from the community is so critical. And whether it’s prerecorded or live AD, you really do want to bring in that community. Nothing About Us Without Us is the mantra that was behind getting the ADA passed, and it really means bringing disability community into whatever you’re doing, whether it’s AD or any type of accessibility effort. So Anna, let’s start with you, dig into a little bit of the process that you go through to get this content accessible.

ANNA CAPEZZERA: Sure. So it all starts, of course, with receiving the videos that we need in order to start working on the AD script. And we’re already usually in touch with Eric and his team in terms of this title is coming in and what type of casting, if there are any notes for that just so that we can get started and have that in mind when we work on the project. So we’ll get everything set up in our cloud-based platform which we can write and record in. And just– what I’ll start thinking about is just who’s the good writer for this project? What is this title about? We know this is a very high-touch, white glove treatment type of AD work. And so we really put our most experienced writers on these Pixar projects and stay in touch with them and in touch with Eric about working on it. So the writer will watch the entire film first or the short or show so that they just the continuity and what it’s about and the tone and everything like that. And then they’ll get started on writing. And it really is an interesting process because something with a lot of dialogue, sometimes you don’t have a lot of pauses to fit the description in, but Pixar movies are pretty visually rich, so figuring out what to prioritize. Sometimes there’s no dialogue at all. Some of the Pixar shorts, they maybe have music or sound effects or sound effects that replace dialogue like animal noises or something like that. And so I think that Christina can speak pretty well to the writer’s process and what are interesting projects to work on and what can be difficult.

THOMAS WLODKOWSKI: All right, Christina. I’ll let you take it from there.

ANNA CAPEZZERA: And you are on mute.

CHRISTINA STEVENS: So sorry. With Pixar projects, they really can run the gamut. We have some that are all about the dialogue and all about characters going back and forth and being really cute and funny and witty with each other. And then there are some that are almost silent. And the great thing with our really experienced writers speaking to these people that I’ve worked with for a really long time, they just love movies. We all have such a passion, and that’s– one really great thing about working with Pixar is I think we all share the same passion for storytelling. And so when we approach it, it’s not the same, this like, this happens, this happens, and this happens. It’s the overall, how does it feel when we watch this? And that’s maybe a woo-woo way of describing it. But it’s a way of getting in the tone of the story. If it’s more melancholy, we’re going to use certain words, we’re going to use a certain sentence structure. And with something with a lot of short montages, we’ll use short sentences to match the visuals there. There’s a lot of thought that goes into it on that end of things. And then once we have the script, we get notes from Eric and his team and possibly from the filmmakers themselves, and then we do our best to incorporate it. We’ll have the writer go through the script since they know it the best. And then they can let the filmmaker or whoever was giving the notes know, this works or this doesn’t work or here’s why I did it the way I did it, and it’s really a great dialogue that we have back and forth. We know they’re not telling us, you have to do it this way, it’s just we’re having a conversation. And it’s a really great partnership.

THOMAS WLODKOWSKI: So does that differ from how you work with other clients?

CHRISTINA STEVENS: Yeah, yeah. I mean, there are some clients that don’t give any notes, and then there’s some clients that give notes and you just have to make them work for– it’s all over the place.

THOMAS WLODKOWSKI: And how do you deal with the audience that you’re writing for? The filmmakers know the film that they’re making, and Eric is obviously between you and them or connecting you with them. How do you maintain your independence around style, et cetera versus maybe working with people who are new to audio description and may not be as familiar with some of the approaches that we’ve all come to know and love that make audio description work?

CHRISTINA STEVENS: We really do our best to– and we’ve done it a lot, especially the experienced writers, where we are basically explaining why we do things the way we do things. There are a lot of standards and guidelines in the audio description world. And there’s a reason we do them this way, and we do our best to explain them to the client and explain why they’re done this way, and it’s sort of up to them at that point what they want to do with it. Of course, we want to advocate for the end product, but it can be a tricky line to walk.

ERIC PEARSON: I can throw a little in there as I’m sort of the liaison between maybe some folks who might not have as much experience, but also have a really clear vision of the story. It’s about education bidirectionally. There are certain things maybe about the tone or the character or the feeling that we’re trying to get across that maybe Christina’s team doesn’t have. But then similarly, there might be some conventions or time-based limitations or other things about the best way to present audio description that our producers and directors might not understand. So I sit in the middle. And I’ve learned enough over the years to be able to at least articulate and broker a compromise if one needs to happen, if there’s any tension between what I really think needs to be said and what should be said. But there’s plenty of other folks on my team. Cynthia Slavens, Jesse Schroeder, Jeremy Sloan, Susan Eggett, a bunch of Pixar folks. And Jonas Rivera actually is an executive at Pixar. And we all take turns being involved reading and facilitating. So I feel like we have a deepening understanding within the studio that goes all the way up the executive ranks of what’s important, why it’s important. So it gets easier the more you teach.

THOMAS WLODKOWSKI: Excellent, excellent. So Laura, we have the film, we have the script. Now you bring it home. And you’re the person I hear when I’m watching that show. So tell us a little bit about your side of the world.

LAURA POST: So the main responsibility as the talent is to just fit in with the movie. You don’t want it to be a noticeable. You want to try and match the tone of the movie or show or short as closely as possible. And audio description is a very subtle art. You’re trying not to editorialize. You want the audience to come to their own conclusions based on what is happening on screen, so you don’t put any of your own emotions or opinions about characters in there as much as you may want to sometimes, ha-ha. But there’s a certain– being able to make those slight adjustments to tone I think is what makes having a human narrator exceptional in the end product. Knowing that a simple line like, he takes the hat, whether it’s a comedic, He takes the hat. Or it’s a very serious scene. He takes the hat. Like, it’s subtle, it’s still not editorialized, but it fits in with whatever is happening around and so it doesn’t jar your audience quite as much as, for example, an AI voice. So that’s– I think it’s– I try to take it as my responsibility to just fit into the movie as seamlessly as possible, and hopefully be ignored after that.

THOMAS WLODKOWSKI: You do a great job at it.

LAURA POST: Thank you.

THOMAS WLODKOWSKI: This is really interesting. I do want to– since we’re at the Sight Tech Global Conference, I probably ought to talk a little bit about tech. I mean, audio description is. Certainly we can get into a lot of things here. But you highlighted one area that that’s a bit controversial in the community, which is human voiceover talent versus text-to-speech. And I guess Eric, I would start with you. How do we balance– obviously we want to keep voiceover talent on these films and shows, but do you see a role for text-to-speech somewhere in the audio description space?

ERIC PEARSON: Well, probably the best way to address that would be to talk about how we use artificial intelligence or AI concepts as part of some of the things we actually do. So we actually do– we have noise reduction technology and increasingly trying to understand how to– we call it super-res, how to create higher-resolution images without the burden of rendering. So in image production, it’s one of the most important things that our artists do, is create the images. We do lean on artificial intelligence for certain things so that we can reach a scale. So that we can tell the best story possible within the constraints– time and space constraints around us. We are in a privileged position at Pixar specifically that we have enough resources to do this high-touch version of audio description for everything that we do. But if we couldn’t, I would still want our audience to have the opportunity to have an audio description experience. So if scale– if the time and space confounded us and we had to lean on AI for some portion of what we were doing, I think we would absolutely do it because I think that– I would rather our audience have some audio description than nothing at all.

THOMAS WLODKOWSKI: When you take AI at scale, where would something come into play like that? Can you give me an example?

ERIC PEARSON: Well, if I go– if I look at beyond Pixar, if I look at the entire Disney catalog, if we didn’t have an individual or groups of individuals like us to be able to personally curate every audio description experience for the entire Disney catalog, I wouldn’t say don’t provide the audio description for it. I would say, let’s look for appropriate reasonable technology solutions so that we can have– so that we can raise the amount of accessibility that’s available. So that’s where scale would come in. If the volume– if the folks who were available to address the accessibility needs and wants that we had weren’t available, where could technology help us get to that scale?

THOMAS WLODKOWSKI: And I think for us, I agree. I think on the text-to-speech side, I don’t believe that anyone is advocating for removing the human voice talent from first-run films and TV shows, things like that. But like anything in the AI space or even in text-to-speech, we do need to find opportunities where we can exercise the technology and see how we can push it forward and how can it make the process even more efficient, more cost-effective, whatever that might mean down the road. But the only way to get there is with trying to do the necessary R&D to get that benefit longer-term. But I think where we are today is still not where it would need to be. Maybe even text-to-speech, certainly not for a first-run film, but for a corporate video, it might be just fine and might be the difference between having a described all-hands video content for a meeting versus not having it described. I guess, Christina and Laura, any perspectives from the creative side to what I just talked about in terms of where we might find opportunities to drive AI forward, but perhaps just not on that first-run film or TV show?

CHRISTINA STEVENS: I’m going to let Anna take this one, actually.

THOMAS WLODKOWSKI: OK.

ANNA CAPEZZERA: Sure. So I think something maybe for documentary content might be a more realistic place to look into using synthetic voice. And as Eric mentioned, to scale. If it’s a docuseries on a rush or a lot of work coming through and you need to do it that way, that may be a place that we would start looking at. Something I did want to mention in terms of just the Pixar way and the human touch is that, as I mentioned in the beginning, I’ll be in touch with Eric when we start a Pixar project, and he and his team will discuss any choices they might have in terms of casting. What type of voice? Would they like it to be culturally relevan? The representation. These are all really important things that go into specifically casting Pixar projects. And so I’ll really work hand-in-hand with Eric and the writer, whether it’s Christina or another one of our staff writers, to talk about what is this movie– what is it about? What’s the tone? Who do you think would be– if casting for representation is very important. And we have voice actors that we know so well like Laura who we really trust to read the content , read the AD, hear the program audio, know the story, and guide themselves with that as sort of a director role. I’ve heard Laura voice kids shows and young adult shows. She has a higher register, she has a lower register that she’s used for Wonder Woman and stuff like that. And so we’ll really think about what actors do we have who are really good at knowing how to read the AD and the content to get the right tone. And then what does Eric and his team, what are they interested in so that we can all work together to send samples so that Pixar can choose what they would like? And I think that would be very hard with AI, especially in terms of obviously representation.

THOMAS WLODKOWSKI: Absolutely. Yeah. I mean, I think that makes sense. Find the right areas to exercise the technology. How has the technology changed the writing and maybe even the voice piece of it? I know at WGBH, back in the day, we had these 3/4-inch tapes and we had software, and it was quite the process. And then we went into a studio and did the mix session. It’s been a while since I’ve been on the production side of audio description, more on the distribution side, and we certainly have our challenges there in terms of keeping audio description with the content as it moves across platforms. But Christina or Laura, anything you want to embellish on here in terms of how technology has improved since AD really came into its own?

CHRISTINA STEVENS: Yeah. We have an excellent cloud-based platform which we write and record in. And even when I started in about 2014, we were not using it. So when I went into the office, you’d have to get the video downloaded, we weren’t using tape then. But you [INAUDIBLE] video downloaded. And it was a process, exporting files and getting them to the record room, and then the actors would have to come in. And we did get onto the cloud platform before the pandemic hit, which was a big blessing, because once we all went home, it allowed us to seamlessly go into working from home and being able to work from basically wherever and be able to work with a wider– people from all over for our freelancer pool and stuff. And then it really changed things for the actors, that they did not have to come in to record in the booth. And we had a lot of scheduling because we had two recording booths, and so there was a lot of shuffling and scheduling. And then during the pandemic, I’m not even sure how that would have worked, but I’ll let Laura go into how it’s been recording.

LAURA POST: Yeah. The cloud-based platform is a total blessing. It’s really nice for recording from home, which we all had– basically the entire voice acting world had to adjust to for the pandemic. I had always been recording from home pre-pandemic for certain things, but I used to go into Deluxe and go into the little studio they had there and I would record things. And it’s always hard because you don’t know– you might know the program length of what you’re going to record, but it doesn’t necessarily tell you how long you’re going to take to record that. It doesn’t tell– so it’s nice to not have to worry about maybe holding other people up if you’re taking longer or wasting studio time if you finish really fast, although it’s always better to finish fast. So it’s been really nice from a talent perspective. It makes my scheduling for my day a lot easier, I’m sure, and it also makes your guys’ scheduling a lot easier when everybody is going from home, too. So yeah, it’s great.

ERIC PEARSON: I can definitely say, we’ve only we’ve got a really short amount of time left, we only about a minute, so I’ll be super quick. But just the speed. The ability to engage our creatives and get approvals on things. I mean, the technology has really made a huge difference.

THOMAS WLODKOWSKI: I want to thank everybody for joining us. I mean, there’s so much more we can dig into here. Eric, I’d love to maybe get Ned to get us back together next year and talk about how we keep content intact when we move it across platforms. I mean, you talked about Disney+, Disney+ is on our Xfinity X1 and Flex and Xumo TV products now. And so obviously people are getting that content. But sometimes it airs on TV– it airs on TV and the content may or may not be there. So plenty more to talk about as we move forward.

ERIC PEARSON: Absolutely.

THOMAS WLODKOWSKI: I know we’re over time here, but does anybody have any final words of wisdom here that we want to impart?

ERIC PEARSON: I’ll just say, some of the themes that have come up, like working together, collaboration, trust, partnership, just huge parts of continuing to move this really exciting, important work forward. So yeah, let’s get everyone back together, Tom, I would love to spend some time thinking with all of us about the future of audio description and how we can make it even better.

THOMAS WLODKOWSKI: Great. Well thanks, everyone. And we will hopefully be back together soon, and enjoy the rest of the conference.

CHRISTINA STEVENS: Thank you.

ERIC PEARSON: Thanks, everybody.

[MUSIC PLAYING]