You already knew that you can turn on subtitles for your TV show or movie—handy if you’re hearing impaired, or just want to understand the dialogue better. But there’s a corresponding feature for people with low vision: audio description tracks, where an unseen narrator tells you, in real time, what’s happening on the screen. But who creates them, and how, and when? And how do they describe the action during fast dialogue, fast action, sex scenes, and screens full of scrolling credits? A deep dive into a bizarre art form most people didn’t know exists.
Guests: Lauren Berglund, consumer relations coordinator at the Guide Dog Foundation. Bill Patterson, founder, Audio Description Solutions. Rhys Lloyd, studio head, Descriptive Video Works. Bryan Gould, director of the National Center for Accessible Media at WGBH.
You’re probably familiar with closed captions—subtitles—on TV shows and movies—an essential feature if you have trouble hearing. But what if you have trouble seeing?
Narr: The great white’s massive head pops up above the surface as Brody chucks another scoop of chum into the water.
Brody: You’re gonna need a bigger boat.
Audio description is a mysterious alternate audio track that’s been sitting right under your nose for decades. Today, you’ll meet the people who create those tracks: describing fast action, by multiple characters, without ever intruding on the dialogue. I’m David Pogue, and this is “Unsung Science.”
OK, here we go: Season 1, Episode 8: Subtitles for the Blind.
In the mid-90s, I lived on the Upper West Side of Manhattan. One fine night, I flopped on the couch, turned on the TV—and the weirdest thing happened.
MORGAN: Eileen Saxon was wheeled into the operating room.
NARR: In a re-creation, white sheets cover the tiny baby on a gurney.
There was something clearly wrong with the audio. It sounded like there was some offscreen interloper, saying out loud everything that was happening in the show. But in a super weird way. Like, it was leaving some things out. And saying some things before they’d happened. And the tone was kind of, like, clinical; whoever was talking seemed really neutral about the drama on the screen.
NARR: In a sterile operating room, the surgical team wears white lab coats, caps, and masks. An attendant wheels a tray holding gleaming instruments and a stack of white cloths.
That was my entirely accidental first encounter with audio descriptions for the blind.
Just as closed captions were invented for people who can’t hear well, audio descriptions were developed for people who can’t see well. They’ve been right under your nose for years.
Lauren I’ve had vision loss my whole life, so I don’t know what normal vision is or what I’m missing.
Lauren I see color. Things aren’t blurry. They’re just not as sharp and in focus.
Lauren Berglund is the consumer relations coordinator at the Guide Dog Foundation.
David So before audio description came along, what would you be able to get out of a movie?
Lauren I would get the color and kind of the movement, but I couldn’t recognize the people, I couldn’t necessarily recognize the scenes or what was happening. If there was any text on the screen, I couldn’t read that. So really, I was just reliant on the audio cues of what was happening in the scenes, but really was not able to pick up what was going on visually.
With audio description, I’m really able to get a full picture of what’s going on in the scene. I know who’s doing what, who’s saying what, and what the characters look like and — and where the scene is even taking place.
Let me play you a scene from the classic 1994 Disney movie, “The Lion King.” The original animated one, not the live-action version that nobody asked for.
Pretend you’ve never seen this movie. Pretend you’ve just sat down in the theater with your friends and your popcorn. And now, for the first four minutes of the movie, there’s not a single word of dialogue. Because this is an audio podcast, you’re about to get exactly the same information that a blind person would get. See how much you can follow by listening to this:
“Circle of Life” from Lion King clip
I mean, great song. But what is going on in the story?
A lot, actually. There are characters interacting, there’s a phenomenon of nature, there’s crowd action—but you wouldn’t know any of that. Now, here’s the same scene with the description track turned on:
Lion King clip
NARR: Simba dangles from Rafiki’s arms, looking small and scared. A ray of sparkling sunshine beams down on Simba like a spotlight. Far below, the animals bow down, their heads nearly touching the ground. From far away, we see every animal from the savannah paying respect to their king’s new son.
Makes a big difference!
Shortly after my freak TV night in 1996, I figured out what I’d been hearing. But I had a lot of questions. Like, Who invented audio description? Who does the talking? Who hires them? What do they do about sex scenes? Do they mention people’s race? Can you get the descriptions if you’re streaming movies? How ‘bout if you’re in a movie theater?
Let’s start at the very beginning—with Margaret Pfanstiehl. Kind of a cool spelling: silent P-FAN, S-T-IEHL. Pfanstiehl.
She was born in Virginia in 1932, grew up in Maryland, and went to a music conservatory, intending to become an opera singer. But in her thirties, she developed a degenerative disease called retinitis pigmentosa—and lost almost all of her sight. And with that, she also lost her ability to see plays, go to museums, watch TV, or even read the paper.
So in the early 1970s, Margaret Pfanstiehl became obsessed with finding technology that could restore those cultural options. With some government grants, she founded a nonprofit radio reading service called the Metropolitan Washington Ear. A radio reading service is where volunteers read the news each day over the radio, so that blind people can, in effect, read the paper.
Her next target was the live-theater problem—and here’s where our story really takes off.
Bill Well, audio description actually started in 1981 in Washington, D.C. It was a project between the Metropolitan Washington Ear, and Arena Stage.
Bill Patterson is the founder of a company called Audio Description Solutions, and he was there at the birth. Which took place in Washington, DC, at the Arena Stage theater.
BILL: The House manager at Arena Stage, said, “Hey, we’ve got this wonderful new system of assistive listening devices for people who are hard of hearing: Point the microphone at the stage, and boost the sound for people who are hard of hearing.” And he said, “I don’t know why we can’t point a microphone at a person who describes, for people who are blind, what’s going on.”
So he spoke with Margaret Pfanstiehl. And she went, “Wow, that sounds like a great use of some new technology.” So she enlisted—actually, he was not her husband at the time, but he became her husband within a year or so, Cody Pfanstiehl. And the two of them created audio description.
The idea here was that blind people in the audience would wear headsets, and a professional describer person would quietly speak into a microphone what was happening onstage.
Live theater clip
Narr: Cassandra looks around her, takes out a snow-white-dressed voodoo doll. Takes a hatpin out of her turban—STAB. Looks up the stairs, eyes wide.
Bill They began to try doing it from, like the sound booth, the lighting booth, that kind of thing, because typically those are walled off with a glass window.
So that was the very first audio description. The describers didn’t use scripts; they just wung it, live, just describing what they were seeing. But from that experiment, they established the ground rules for working around the dialogue.
Bill They realized, you know, almost immediately, “No, nobody can listen to two people talking at the same time.” And so, you know, there was one of the original rules of audio description. Don’t speak over the dialogue unless you absolutely, absolutely have to. We’re at the mercy of where the silences are.
In the early 80s, Bill Patterson met the Pfanstiehls. He was a professor at the University of Maryland, and he wanted to get descriptions going for the university’s own theater productions. But he kept running into theatrical situations that hadn’t come up before.
Bill There was an occasional something that had never presented itself. Boy, was it handy to have Margaret at the other end of the phone line to call up and say, “OK, I got a new one for you.” And —and she loved, you know, being part of that problem solving.
Eventually, Bill became a describer himself. He teamed up with the Pfanstiehls, and helped to perfect this strange new dramatic art.
Bill Good audio description is a lot like police reporting. It’s well-observed, it’s factual, it’s accurate, it’s efficient, it’s matter of fact, no-nonsense, it’s useful information.
David So, so I shouldn’t write, “There’s an awesome explosion.” I could write, “There’s a huge explosion.”
Bill Exactly. Exactly. And, you know, what are the colors in the explosion?
Yeah, I wondered about that…the colors.
I gotta you, I really sacrificed for this episode. I watched a lot of movies. All for research. All for you.
And I discovered some super weird things about description tracks. Like, they’re constantly mentioning the colors of things. Here’s just one line from “Legally Blonde.”
Narr: Wearing sparkling pink heels, and sporting a magenta dress with a fuchsia satin sash, she sports a pink bag and sports pink lipstick and pink nails and she strolls down the central aisle and approaches the defense table.
And I wondered: How useful is that going to be for blind people?
Bill: (cont’d) many people who lose their sight have been sighted, and have memories of color. But also people who were blind from birth but learned about color. They, they know that red is hot and tempestuous and so forth, and the blues and greens are serene and restful.
Also, remember that 90% of blind people actually have some vision. Yeah—true. Almost nobody sees complete blackness; it’s far more common that you can detect some light, or maybe blobs of color. In that case, a description track that identifies what the blobs of color are is hugely helpful.
Bill: (con’t) if somebody has limited vision, describing the blazing orange sun at the upper left lets them clue in to—that’s what that is up there! I just thought it was a blur.
OK. So when you’re writing a description track, the first thing you have to decide is—what to describe.
BILL: I find description challenging because there are about 47 things I would like to describe right here. And I have time for five words.
Really important is to describe things that keep people connected to the plot, and its changes, absolutely entrances and exits, who’s here and hearing this, and props.
Yeah, we may as well admit that audio describers have to leave out a lot. Sometimes, a really lot.
David There’s this great movie, animated movie on Netflix called The Mitchells versus the Machines, and—
Rhys We did that one.
David Did you really?! Oh my god!
Rhys Lloyd is the studio head for Descriptive Video Works, one of the busiest audio description companies going. They’ve described TV shows like Breaking Bad, Better Call Saul, House of Cards, Friends, The Office…movies like Crip Camp, Mank, Uncut Gems, and The Beatles Get Back on Disney+…and live events, like the last few Olympics and the Macy’s Thanksgiving Parade.
Thanksgiving Parade clip
Narr: Santa continues to dance in his sleigh. Below him, his elves wave and dance.
David: (continued) And there’s a scene where the dad has set up, basically, an animal trap that’s a lasso, a rope on the ground tied to a tree branch, that will snap upward when someone steps in there.
Dad: To survive in the outdoors, you’ll need to learn how to trap wild game. Now this is very sensitive…
Neighbor: Hey, Mitchell! (commotion)
Dad: Hey, who’s this curious guy?
David (cont’d) And he’s having a conversation with the neighbor who walks in and is about to step into that loop. So the dad, in order to save the day, steps in and gets caught in the loop himself . None of that is in the audio description.
Dad: To survive in the outdoors, you’ll need to learn how to trap wild game. Now this is very sensitive…
Neighbor: Hey, Mitchell! (commotion)
Narr: The dad dangles from a tree.
Dad: Hey, who’s this curious guy?
Narr: Four possums attack him.
Rhys Yeah, you’re —you’re not wrong. I mean, the limitations of what you can get to are significant. And —and so we are often making the choices of like it’s like, you know, what can you describe?
You’ve also got to figure out when to do your describing. Here’s Bill Patterson again.
Bill For instance, in a mystery play, at a critical moment, the cat nudges a vase off the mantel. It crashes to the floor. And behind it are the jewels that have been missing. And so at the moment all that happens, you can’t start talking about, “Oh, there was a cat there,” and so forth. No, you need to get the cat established, just, you know, matter of factly, you know.
David So you can’t say, “Oh, I forgot to mention, there was this cat!”
OK. So audio descriptions were chugging along in live theater performances here and there, and describers were gradually figuring out how to make them work well. But descriptions didn’t make the leap to TV until 1988, when WGBH came a-callin’.
WGBH is Boston’s public-broadcasting station. I’ve actually been there, in that building. Because by pure coincidence, GBH also produces “NOVA,” the science show that I sometimes host.
Bryan Well, part of our mission is to provide programming to the largest audience possible, and that includes people with disabilities. So that accessibility story starts 50 years ago, with WGBH inventing broadcast captions for people who are deaf and hard of hearing.
Bryan Yeah, the first program that aired broadcast captions was Julia Child’s “The French Chef,” 50 years ago.
This is Bryan Gould. He’s the director of the National Center for Accessible Media at WGBH—the department that invented not only closed captions for the deaf, but also TV audio descriptions for the blind.
Bryan GBH also invented descriptive video service.
David I had no idea. I’m a fan of that science show they do.
Bryan Oh, “NOVA?” I’ve heard of it. Yeah.
To develop audio description for TV, WGBH went straight to the source.
Bryan: While we were trying to bring this service to television, worked closely with The Washington Ear and took many of their practices and guidelines and applied it to television and got some grants.
It’s this guy, Bryan Gould, who reopened my own cold case—The Mystery of Pogue’s 1996 TV—and solved it.
David I was really weirded out, like, “who’s talking!?”
Bryan So many people have that happen, where all of a sudden the TV starts talking to them and telling them everything that’s happening. And they say, “But I can already see this!”
You had somehow stumbled on to something called the second audio program channel, or the SAP Channel, originally intended to broadcast usually Spanish translation of a show.
So the decision was made to broadcast descriptions on the second audio program channel. And so you stumbled on that, and that—that started happening more and more as remote controls started to have an SAP button on them.
By the way, most people call this feature audio description. But you also run across the term “video description.” And in Canada, they call it described video, or DV.
David So is it, is the terminology—I’ve heard it referred to as audio description, I’ve heard it video description.
Bryan I think it actually comes down to acronyms, because AD is better than VD.
For those under 40, let me explain that reference: VD was the term for what we now call STDs.
David That’s funny, because—because VD is really what it is. They’re describing the video, they’re not describing the audio.
Bryan That’s right. That’s right. it comes down to, what’s the little bug on your on your website or on your publication going to be? And it’s going to be AD, it’s not going to VD.
David And this is how technology is made.
For many years, WGBH was the only game in town. Of course, they made description tracks for PBS shows. But in the early two thousands, GBH also went into the business of describing Hollywood movies—another first.
Bryan Any movie that was described, yeah, it was described by us. No one else was doing it.
David Really!? Oh, wow.
Bryan We probably did every movie that was described up until, I don’t know, 2005 or something like that.
In 2010, President Obama signed the 21st Century Communications and Video Accessibility Act—or as insiders don’t call it, “2CCVAA.” It mandates that the big TV broadcast and cable networks produce description for at least 348 hours of shows a year, and that movie theaters of a certain size have to offer wireless headsets for the description track.
On broadcast and cable TV these days, most of the popular dramas and kid shows have descriptions. You know: Shark Tank, The Bachelor, Family Guy, The Simpsons, Law and Order, and NCIS: Whatever. And, of course, just about everything on PBS.
Bryan And all your episodes of NOVA, I’m sure, are described .
David Oh my gosh!
Bryan Wait—have you, have you never watched? Of course they are.
Turns out—he’s right! Here’s what one my NOVA shows sounds like with audio description.
DAVID: And I run a few comparisons myself.
NARR: David hammers a banana. It smushes. He hammers a LineX-coated banana; it doesn’t crack. He whacks it with a baseball bat. In slow motion, David wields a sledge hammer.
DAVID: Finally, bringing out the big guns.
NARR: He drives a truck over the coated objects. They hold their shape.
I had no idea! Man, those sound like pretty good shows!
Anyway, yeah. Audio description has become a thing.
Bryan We did a dozen movies a year, and now there’s dozens of movies a week or even a day that are coming out with description. It’s it’s —it’s fantastic.
So how do you hear these audio tracks? On a TV, you use your remote to choose that SAP audio channel, Second Audio Program. It’s probably in the menus.
The streaming services, like Netflix, Disney Plus, HBO Max, and Amazon Prime, commission descriptions for just about everything they produce: The Crown, Tiger King, Succession, Curb Your Enthusiasm, Atlanta, Ted Lasso, Game of Thrones, and on and on. Here’s a little bit of “Game of Thrones:”
NARR: Drogon shoots a second jet of flame directly at the iron throne. The spiked fringe, ablaze, glows red hot and bends backward. The swords begin to melt under the immense heat. The iron throne liquifies into a bubbling mass of molten steel.
On those services, the description track is listed in the same menu with the other language soundtracks. Right there, beside Spanish, French, and whatever, you’ll see English (AD). Audio description.
And I have to mention this incredible smartphone app called Spectrum Access, formerly called ActiView. It plays the description tracks for movies and shows—over earbuds, if you like. But here’s the cool part: It figures out how to sync up with whatever is playing, in your home or in a theater, by listening to the soundtrack and figuring out where you are. Here’s the app finding its place in “Black Panther.”
Black Panther clip
NARR: He activates his mask. Okoye points her spear at Killmonger.
Spectrum Access is free and it’s amazing.
Now, if you’re a sighted person, you might be going, “Oh, how nice that there’s a technology solution for blind people.”
But anyone in the accessibility world can tell you: It’s a very rare accessibility feature that doesn’t also go mainstream.
Like, you know curb cuts? Those ramps at street corners? They were designed for people in wheelchairs. But nowadays, who uses curb cuts mostly? People with bikes. Or strollers. Or rolling luggage. Or carts.
Or take subtitles. Nowadays, I watch everything with closed captions on. You just get so much more out of the writing. You miss a lot less.
Rhys Lloyd is convinced that audio description will break out, too.
Rhys: If you’re anything like me, a lot of my viewing is done distractedly with a laptop in front of me or my phone, and audio description can fill that gap for you if you’re cooking.
I mean, I used to live in Los Angeles. I know the traffic there. And people sometimes want to keep enjoying that show they were watching. They could potentially listen to it on their phone as they’re stuck in traffic and— and with audio description, they can– they can not miss some of the visuals.
So yeah: audio description has been a big success. And the concept seems simple enough in practice.
But in the real world, things can fall apart pretty quickly. After the break: the seven nightmares of highly effective describers. They are as follows: Short deadlines, fast action, subtitled scenes, sex scenes, racial issues, long credits …and synthesized voice software.
OK. We’ve been talking about audio description tracks in TV and movies, which I like to think of as subtitles for the blind. Bryan Gould’s department at WGBH creates descriptions for all kinds of visual culture.
David: Let’s talk about how these get made.
Bryan The typical situation is that we receive a final show, a finished program. [00:16:31] And typically, a describer will watch the program and maybe watch it several times so that they know sort of—if it’s, for example, if it’s a mystery sort of program, you don’t know what the red herrings are until you know the end of the mystery.
David And then what? You fire up Microsoft Word and start typing?
Bryan Well, we use some specialized software that helps doing what we call timing out.
Once the script is written, edited, and approved, it’s time to record the audio. Description companies hire professional voiceover actors, the kind of people you hear on audiobooks. And this is the part that blew me away: they read it stone cold. Without ever having looked at the script first.
Bryan It’s an amazing process. They’re in their recording booth with a microphone and a little monitor and their headphones on, and they’re watching the program or even the movie for the first time and reading the script live. Now, of course, if they flub or they read too fast or too slow, they can edit.
And presto: A description track is born.
Seems like it’s pretty easy, right? Seems like creating a description track for people with low vision shouldn’t be any harder than creating subtitles for people with weak hearing.
Ah, but that’s where you’re wrong. Audio description is a lot trickier to make than subtitles—for seven reasons.
First: You generally have very little time to create the track.
Bryan Iif you’re doing a movie that’s coming out Friday night and we get a copy of the movie on Wednesday, it’s two and a half hours long, you need six people to write the script as fast as humanly possible.
David That happens?
Bryan Oh, sure. Oh, sure. I can tell you way, way back in the day, the second theatrical-release movie that GBH ever described was “Titanic.”
David Oh, no way!
Bryan Yeah, so we had to do it very fast. It was like 24 hours of writing like crazy. But it also was in theaters for like a year and a half. So it had a very long tail.
Jack: I’m the king of the world!
Narr: Jack tilts his head back and closes his eyes as the enormous ship surges across the ocean.
The second problem comes with the territory: How do you describe action that’s happening faster than you can narrate? Well, one way is to pre-narrate.
David In “The Matrix,” they say, “Neo, in slow motion, does three cartwheels, grabs a submachine gun and fires into the chest of the guard.” And that hasn’t happened yet.
Narr: Neo drops the submachine guns and runs directly toward the troopers. He flips a cartwheel and grabs an M16 rifle from the ground. (RAT-A-TAT-TAT!)
Bryan [00:21:21] that’s a great example, because in a fast paced action scene— this is an audio experience. if the sound effects for what you just said are about to happen, and they’re quite loud and pretty frankly awesome to listen to—
Matrix clip continues
Bryan (con’d) you want to create that image and you want to, you know, preload it with those words, and then it happens. Because if you’re just describing, “well, this just happened,” you’re never going to catch up.
Narr: Trinity picks up the duffel bag and drops the shotgun. Together, they walk to the elevator.
Problem number 3 is foreign-language subtitles. Remember, the golden rule for a describer is that you never talk over dialogue. But what about…
David What about when the entire movie is in another language?
Bryan Then you have to talk over, obviously, because you have to provide the information. There’s no other way to watch the movie. We’ve gone to the lengths of actually hiring three different people to do the job, one person just to read descriptions, one person to read male parts that are—or male characters, read that with subtitles, and another to read female character subtitles.
I wanted to hear what that’s like, where the describer just speaks right over the dialog. So I spent 4 bucks to re rent the fantastic Korean movie “Parasite,” which won four Oscars in 2019, including Best Picture. Here’s what it sounds like:
Parasite dialogue clip
It doesn’t have a description track. Nobody bothered to make one! For a Best Picture-Oscar movie! What the heck!
I did a little Googling. Turns out that subtitled foreign movies almost never have description, because they tend to do so pathetically at the U.S. box office.
Apparently there was a described version of “Parasite” in the U.K. But the American distributor, Neon, decided to just blow off Americans who have impaired vision. Seems kinda dumb.…You’re gonna write off 26 million viewers because you don’t want to spend a few grand getting the description track done?
OK, now the fourth challenge: Sex scenes. Like this one, from the 2015 Brangelina movie “By the Sea.” And parents, just a warning here: The following audio clip acknowledges the existence of bathtubs.
By the Sea scene
Narr: Still clothed, Roland climbs into the tub. He holds her close as they share a gentle kiss.
David OK, so how about sex scenes? What happens there?
Bryan Well, when a man and a woman love each other very much, David—
David This interview is over.
Bryan You asked! In all scenes, description writing must take in the context. And if it’s a—adult drama or an R-rated film, it’s going to be described in the context of the film, in the quote-unquote language of the film. And we have used, you know, not X-rated words, but we’ve used racier words depending on what the— the level of passion is, I’ll say.
David Would you use body-part names?
Bryan Of course. Of course. if you’re being, I’ll say, prudish, about a sex scene in an R-rated movie about two people who are being very passionate about what’s going on there—well, why should someone who can’t see a scene but is listening to it, not understand what the director is providing ?
And while we’re on the topic of evolving awareness …Here’s Rhys Lloyd again.
Rhys Historically, the guidelines around race and audio description was to describe the character’s race if it’s relevant to the plot, which, I mean, on the surface, it makes a certain degree of sense.
A case in point: “Get Out,” Jordan Peele’s 2017 social-satire horror movie. It’s a movie about race. Every single character is either black or white, and their race matters to the story in every single case. And so, sure enough, the describer almost always lets you know. Here are some examples:
Get Out clip
NARR: A thin young African-American man walks while on his phone… A brown-eyed African-American man faces the steamy mirror…A Caucasian brunette stares at pastries…Walter greets a white couple.
Except in one kind of surprising case: When the white cop starts harassing the Black main character. They mention that the cop is bald, but not that he’s white.
Get Out clip
NARR: Rose faces a bald police officer.
COP: Sir, can I see your license, please?
Of course, this was back in 2017. Bryan Gould says that the world is always changing.
Bryan When I started writing descriptions way back in the mid-nineties, skin tone or race was only mentioned when someone was not white.
You know, if you’re describing a film where, or a —or even a documentary, where someone is a NASA scientist, for example, or in a position of power, but is a person of color… I always think of the the situation where a young person is seeing someone who looks like them in a position where they don’t expect them, right? Well, if they don’t hear their, the person’s race or skin tone, that’s a missed opportunity there, I think.
I asked Rhys Lloyd about this.
David I mean, it occurs to me that the conversations have already been had by the makers of the movie in their casting. I mean, you couldn’t you just punt and say, for every character, “he’s white, he’s fat and bald, he’s 45 years old,” you know, and so on?
Rhys: 100 percent. You can definitely include that. But what are you– what are you omitting that’s also relevant in those in those in those precious seconds that you have? That nut hasn’t been cracked yet. That— that’s still a challenge, and these are good challenges to have, the things that keep this fluid and interesting and ongoing.
Challenge #6 kind of cracks me up. It’s the closing credits.
The credits for “Avengers: Endgame” are 12 minutes long, and include over 10,000 names—and their job titles. Assistant Depth Supervisor. Choirmaster. Propmaker gang boss.
(And yes, ten thousand. I actually pasted them into a spreadsheet to see how many rows there were. I don’t kid around when I write my podcasts.)
David I’m so confused. Sometimes—they, they seem to name some people and not others.
Bryan So that comes down to what—what does a contract say? There was one movie studio that required every single credit to be read. And so you could sit in the movie theater for, you know, another hour and a half and listen to the credits.
Bryan Because, you know, sometimes it’s three columns of credits going, you know. Animators and everybody else.
Seems like the trend is not to expend quite that much effort anymore. When the first Avengers movie came out in 2012, the describer didn’t read all the credits—
Avengers Travis clip
Narr: Kim Donoyan. Tannoy Read. Troy Robinson. Debbie Rondell…
…but he read as many names as he could during the credits. He left out all the walls of animator names and song credits—and skipped right to the copyright notices and, of course, his own credit.
Narr: This film has been described by Deluxe Digital Studios, with funding by Walt Disney Studios Motion Pictures. Voice by Travis.
Nice job there, Travis.
But by 2019, when “Avengers: Endgame” came along, with its ten thousand names…
Travis was sent home early. In “Endgame,” you don’t hear a single line of the credits.
Now, I promised to outline the seven challenges of describers, and if you’re in the audio-description business, this last one might be the scariest of all. If you remember “Unsung Science” episode 3, you know that synthetic speech is coming on strong. These days, text-to-speech systems sound scarily real. Like this:
Voice: This is a synthetic voice. My mission is not to put human describers out of work. Really!
So the obvious question is: What place is there for human describers, when a text-to-speech app could do the same job for free?
Rhys Our stance on synthetic voiced audio description isn’t that it’s inherently bad. It has its limitations today as a technology. I think some of those limitations are in its ability to subtly carry emotion in the way the human voice can do.
Rhys Lloyd from Descriptive Videoworks is generally not a fan. But he’ll allow text-to-speech in some situations.
Rhys: There’s some content for which synthetic voice might be a reasonable solution, like information-based programming that is less emotional or less about it being evocative.
If your program is a medical show, like literally a documentary about medical procedures, your level of evocative requirement is lower.
Bryan Gould of WGBH would agree.
Bryan I mean, there’s millions of hours of video being uploaded every single day, and having that read by synthetic speech just to have access to it is absolutely appropriate and even welcome. Why put in all of the production time and budget into, you know, a How To video? But if you’re sitting down to watch a movie, you know, there’s a cognitive load that that happens when you’re listening to synthetic speech, and human speech is much more natural and is easier to process.
For now, though, the seven challenges of creating descriptions today shrink to insignificance compared to the biggest problem of all: Getting shows and movies described in the first place.
Lauren I stream 99 percent of everything. so it would be, you know, picking something and then going into the menu and the audio settings and seeing, “Oh, does this have audio descriptions?”—and hoping it does.
Lauren Berglund again.
Lauren (cont’d) Like Netflix has a section of only audio-described titles, so sometimes I’ll start there if, like, you know, I want something that is audio described. Sometimes I’ll find that like certain seasons of a show are described, but not the whole thing. Or sometimes I’ll go to like, ACB has a full list of audio described titles, so sometimes I’ll go there and look through the list and try and pick something out as well.
By ACB, she means the American Council for the Blind. They have a site called the Audio Description Project—ADP. At their website, adp.acb.org, which I’ll link to at unsungscience.com, you’ll find an incredible resource—they have an up-to-date master list of every single TV show, movie, DVD, live show, museum, park, and tour that has audio description.
David So overall, if you had to guess a rough percentage of how often you strike gold these days, what would you say?
Lauren Maybe 40 percent? Like on Hulu, it’ll be described if you watch it on live TV, but when it’s on Hulu, it’s not described. Gray’s Anatomy is described on live TV. But if you watch it on Hulu, it’s not.
Lauren I wish more places would audio describe.
So that’s the thing. Audio description is here, and it works great—we just need more of it. Rhys Lloyd sees no reason we can’t expand our definition of “described video.”
Rhys I think there’s also ways in which we can stretch the service beyond what it exists today in order to achieve more.
And so we’re starting to see interest from broadcasters in including audio description in sporting events. We did some work ourselves on the Olympics and Paralympic Games previously. [00:31:46] we did live audio description of the Macy’s Fourth of July fireworks.
Hold up—Fireworks? How on earth would you narrate—
NARR: Above the lights of Manhattan, a trio of red fireworks illuminates the dark sky. They are overtaken by three golden explosions that expand out like fiery flowers. Their golden streaks cascade toward the Hudson River, dissolving into twinkling sparkles.
Oh—OK! I guess that’s how you do fireworks! Rhys, go on…
Rhys There’s no reason those can’t be described. why wouldn’t you want to connect to everybody? Like, if you’re a creative and you’re passionate about what you produce, why wouldn’t you want everyone to consume it?
That’s a great question. Lauren Berglund is especially baffled by advertisers who leave millions of potential consumers in the dark.
Lauren The thing that gets me is like commercials. Because I’m like, “you’re leaving out a whole audience of possible consumers by not making your commercials accessible. Because like, I have no idea what’s going on in your visual-based commercial with just music in the background.”
And if you’re an advertiser, don’t give me the excuse that, “We’ve only got 30 seconds to get our message out! We don’t have time!” Because there’s a hack for that.
Rhys: There’s also a service called extended audio description where —where the video either gets paused, to allow description to take its course, or the video can be looped over the course of while the descriptions happenin. And so the actual duration of the described version is actually different than the duration of the nondescribed version.
Rhys I honestly think the inevitable convergence of gaming and linear media into some hybrid interactive format is going to allow the viewer more latitude to make some choices about that. Like what if instead of just pressing Pause, there was a way you could press “Pause, tell me more?” And those technologies are possible. I think you’re going to start to see a little bit more of that—that ability to control it.
In all of the time I spent making this episode, that “Parasite” thing really bugged me. Best Picture Oscar, and not a single blind American will ever get so much as a taste.
But wait a minute—I now have some of the world’s most experienced audio description companies on email autocomplete!
What better way to wrap up this episode than asking them to fill just a little void—by creating a description that’s never existed before. A custom track for the best scene in the entire movie “Parasite.”
It happens to be a really challenging scene to describe. There are ten characters in this scene, tearing around three stories of a house, and there’s dialogue in Korean with subtitles.
Here’s the setup: It’s a story of two families: one rich, one poor, each with two children. The rich Park family has cut a trip short, and will be returning to their luxurious home in a matter of minutes. But the lower-class Kim family has made itself comfortable in the house during their absence—and actually, made a mess. The Kims have a matter of minutes to erase all evidence of their presence there.
If you were blind, here’s what you’d get out of the scene:
[all dialogue in Korean]
You can follow that, right? Uh-huh.
…the world premiere of that famous scene from “Parasite”—with a new, custom description track, written and produced by Descriptive Video Works. You’ll hear a lot of character names—I told you, it’s a frantically choreographed scene—but listen in particular for Chong-Sook, the matriarch of the poor Kim family, and Kee-Tek, her husband. Oh, also, at the end—you’ll hear the fate of Mon-Kwong, the housekeeper who is threatening to blow the secret of the Kim family’s real identity.
Parasite clip with English dub and AD
Narr: Chong-Sook: “Then you’re almost here?”
Yeon Kyo: “8 minutes, according to the GPS.” Chung Sook: “You arrive in eight minutes?” Yeon Kyo: “Start boiling the water right away.” Chong-Sook hangs up and turns to them. “What the hell is ramdon?”
Chung Sook races into the kitchen with her cellphone. She grabs a bag of flank steak from the fridge, then slides her arm across the counter and sweeps the mess into drawers.
“Move quickly!” Ki Taek forces Geun Se through the kitchen. “Hurry!” He pins Geun Se’s arms behind his back and takes him downstairs.
Chung Sook fills a pot with water and puts it on the stove next to packaged ramen noodles.
Ki Woo drags Moon Gwang downstairs. “Dad! I can’t drag her any further!”
Ki Jung sweeps broken glass under the sofa and cuts her hand.
Ki Taek drops Geun Se to the concrete floor, then grabs an electrical wire off the wall.
Chung Sook chops the steak. Ki Taek binds Geun Se’s feet. “What are you doing?” Ki Woo drags Moon Gwang by her arms. “Dad!” Ki Taek rushes up the narrow steps and down the hall. “Go quickly! Help upstairs!” Moon Gwang slumps against Ki Taek.
Chung Sook puts the seared meat into the pot.
Ki Woo runs upstairs, then into the living room. Ki Jung tosses Da Hye’s diary. “Take this!” He catches the yellow diary, then races upstairs.
Ki Jung stuffs garbage into a trash bag. She spots the family returning, then stuffs the bag under a large wooden coffee table. She rolls underneath the table, as the family walks upstairs. Da Song and Da Hye wear headphones and walk through the living room.
Chung Sook steps out of the kitchen. “Welcome home!”
Yeon Kyo: “Da Song, look! Let’s eat ramdon!”
Upstairs, Ki Woo puts the diary away. “Da Song!” Da Hye flops onto her bed. Ki Woo locks the case. “Dasong!”
Downstairs, Moon Gwang kicks Ki Taek and he falls backwards. She runs out of the bunker with her hands bound behind her back. “Madam!”
In the kitchen, Chung Sook fills the bowl with ramdon and turns her head sharply. Moon Gwang nears the kitchen, and Chung Sook kicks her backward down the stairs. Her head strikes a concrete wall.
Every now and then in this life, it’s nice to know what’s going on..
UNSUNG SCIENCE with David Pogue is presented by Simon & Schuster and CBS News, and produced by PRX Productions.
Executive Producers for Simon & Schuster are Richard Rhorer and Chris Lynch.
The PRX production team is Jocelyn Gonzales, Morgan Flannery, Pedro Rafael Rosado and the project manager is Ian Fox.
Jesi Nelson composed the Unsung Science theme music, and fact checkerKristina Rebelopositioned herself nobly between my scripts and certain humiliation. I’d like to thank my research assistant Olivia Noble, for spending hours trolling through movies to help me find my examples.
For more on Unsung Science episodes, visit unsungscience.com. Go to my website at David Pogue.com or follow me: @Pogue on your social media platform of choice. Be sure to like and subscribe to Unsung Science wherever you get your podcasts.
Thanks for listening!