Audio description – the visual made verbal
CLIQUEZ ICI POUR LIRE LA VERSION FRANCAISE
In the post Come any closer and you’ll be in the movie, I gave an account of an experience seeing a film from a front-row seat which was too close to the screen for comfort. Assaulted by oversized images which were so intense they made the film hard to watch, there were moments when it felt like being temporary blind.
You can read the whole of that particular article for all the details. The experience it describes of trying to construct meaning and understand a film, when I wasn’t always able to keep my eyes on the screen the whole time, made me wonder what it is like for someone who is permananently blind. The system known as audio description was developed to make films accessible to the visually impaired, and I decided to try and find out more.
What is audio description?
Let’s begin with a definition. Audio description or AD is the oral narrative system used in film, and during certain live events such as theatre performances and sporting competitions, to compensate for the absence of images for the visually deficient. ADs of screenworks are pre-recorded, but for live events the description is spoken by the describer simultaneously with the event.
Audio describers act as mediators between the visually impaired and the images in a film which they could not otherwise perceive. They lend their voices in order to make vision possible through an oral description of what is happening. They lend their eyes so other ears can see.
For proof of just how important this work is, play the short video What is Audio Description? where visually impaired users share what audio description changes in their lives.
In this post, beginning from the origins of AD, we’ll visit the various stages involved in creating the audio description for a screen fiction by which an audio describer works to make the visual become verbal.
The visual made verbal
The process which would ultimately become audio description was initiated in 1974 by Gregory Frazier who was a graduate student in broadcasting at San Francisco State University.1 During his graduate research, he had become conscious of the fact that blind people were always excluded from the full pleasure of watching television unless a friend or a family member was present to whisper in their ear and tell them what was happening visually.
Frazier set to work creating the concepts for what he decribes as a system for translating an audiovisual event into an audio event. Lack of financing meant that Frazier had to content himself with working on pre-existing movies and TV shows until 1988 when the appointment of August Coppola as the new Dean of San Francisco State University helped him fund the making of the first audio description for a brand-new feature-length film. The film in question was Tucker, the new movie directed by the Dean’s brother, a certain Francis Ford Coppola.
At the same time, through the 1970s and also in the USA, audio description saw a parallel development using earphones in the world of theatre where it accompanied stage performances. The Metropolitan Washington Ear Project was directed by Dr. Margaret Pfanshtiel in theatres in the Washington DC area. The Ear Project still exists today and is freely available as a listener service for blind, visually impaired, and physically disabled people who cannot effectively read print.
Although working with different media, Frazier and Pfanshtiel shared the same objective : to reveal the contribution of visual elements in a screen production or a live event which would otherwise be inaccessible to the visually impaired. Whether this involved adding an oral description to the original audio track of a film or creating the live commentary for a play as it was performed, Frazier and Pfanshtiel were ticking similar boxes : narrating the characters as they appeared, the locations and characteristics of the various places where the scenes were set and, of course, the actions in the story being performed. On screen or on stage, it was always about the visual made verbal.2
There are laws in the US – often called Section 508 – covering video accessibility which specifically require audio description.3 At the time of writing this post, the American Audio Description Project for the American Council of the Blind lists 12,700 audio described films. New works are systematically required to join this list, but accessibility to older movies has seen audio descriptions added to titles going back to the 1920s with the shorts by George Melies or Murnau‘s Nosferatu.
To grasp the importance of AD in cinema today, just click here to read the results of a search using the keywords “audio description” on the website of Boxoffice Pro, the film industry magazine focusing on the world of movie theatres, showing just how far things have come today. An article on the Boxoffice Pro website from March 2025 also confirms that France is also now actively pursuing the availability of audio descriptions in major cinemas. This is proof of the effective action taken since January 1st 2020 when the state-run CNC announced its commitment to actively supporting AD on all new French films , making them accessible to all.
Audiodescribers at work
So what does an audio describer actually do ? Let’s see what two professionals have to say about how they work, beginning with Ryan Hennessy, a British audio describer who works with anything from animated short films, social media content and episodic TV, through to lengthier crime documentaries and full Hollywood films. 4
When asked to outline the steps to the making of an audio description, Hennessy was clear and concise :
RH – I usually start by watching the media all the way though, marking out where I’ll add in the audio descriptions. The first watch helps me pick out exactly what I need to describe for the listener. Then I watch it again, this time creating the audio descriptions as I go. After I’ve written the script, it gets sent off to be recorded, and I review the final audio before delivering it to the client.
Sounds simple enough, but remember this : each time the describer speaks, what s/he says has to be sized and timed so that it will fit comfortably between whatever dialogue is already in a scene in real time. Fitting into such a tight space means choices have to be made as to what to put in. Then again, not everything needs to be described because the visually impaired listener must have time to let the story resonate just like a fully sighted viewer. So the describers also have to know what to leave out.
Award-winning audio describer Dune Cherville has an impressive CV, including the French AD version of Walt Disney’s Brave. As one of the leading French voices in the profession, she confirms the dilemma of deciding what to put in and what to leave out.
DC – We have to make endless choices.5
And choosing takes time. Cherville says the equation one hour of work to get the final description for one minute of film is pretty accurate. It’s not about camera movements or angles, it’s about saying what is there to be seen as a scene unfolds.
DC – Talking about camera movement would take us out of the story. The image in a film is constructed like a language so the subject of a close-up is the subject of the sentence.
When writing the description, the biggest battle is being able to consider things from the position of someone who has low vision. For Cherville, it’s often simply a case of knowing when to watch a scene again with eyes closed.
As we have said, audio description is about the visual made verbal. The describer has be prepared to fine tune the text in the smallest details. This requires an ability to write fluently and clearly in a non-repetitive way. Dune Cherville calls the audio describer a word machine which runs on synonyms. And that machine also has to be capable of choosing registers appropriate to the mood of the situations shown in the film in question – which can be anything from deadly serious to wackily funny.
Navigating between written and oral
All of the above concerns written text, of course. But the author of an AD narration knows that their text is ultimately destined to be spoken. So not only is it necessary to decide what to describe in an image or a scene, but also to do so clearly and in a number of words which will be sayable and sound good in the time available. So while content and timing are crucial, the text also has to work orally which, as Dune Cherville points out, is no easy task.
DC – Film goes from a scenario to images. Audiodescription goes from images to text. But it’s not a text that is meant to be read ; it’s a text meant to be heard. So, in the process of writing the text, the describer is forever talking to him or herself. We are constantly checking that our text conjures up a visual while also sounding good when spoken. Because there are sentences which look great on the page but are awful when said out loud.
The importance of the oral impact of an AD is confirmed by the fact that when another French audio describer, Morgan Renault, received the very first Marius Award6 for the best audio description in 2018, he insisted on sharing it with Marie-Pierre Warnault, his visually impaired associate. By systematically testing and fine-tuning his proposed audio descriptions orally with Marie-Pierre Warnault, he is able to validate both their appropriateness and their intelligibilty to someone who had no access whatsoever to the images described. This collaboration ensures the correctness and quality of the final product.
MR – For the lasting quality of audio descriptions, authors must involve the blind and the partially sighted in the creative process because their collaboration gives our work credibility. They have a lot to teach us.
The final recording of the audio description which will accompany the film is made either with the describer as narrator or with a professional voice-actor speaking the text. French voice-actor Isabelle Faria often does audio descriptions.
IF – Voicing an audio description is something much more delicate than acting on stage. The performance, if we can call it that, is more subtle and much lighter. There is already acting from the players in the film so it’s important not to add more acting when reading the description.7
Audio descriptions using Artificial Intelligence
If you are familiar with audio descriptions, at some point you will have encountered those made using artificial intelligence with the machine-like voice with odd intonation and quirky ways of saying things. Perhaps it’s just a question of time until AI can be programmed to produce quality AD. According to Dune Cherville, AI’s major contribution to audio description is in helping to identify the gaps in film dialogue and their duration, which is an enormous time-saver for teams working on spotting the spaces available for slotting in audio descriptions. Other than that, she insists, human describers are more efficient when it comes to creating content because AI doesn’t always know what to describe : is the foreground more important than the background in certain scenes or the speaker more important than the listener? Humans are, so far at least, better at answering questions like these.
In researching for this article, I came across something called ViddyScribe, a name which speaks for itself. It is an online audio description website developed to use artificial intelligence to describe YouTube or educational videos which are otherwise inaccessible to those who are blind or have low vision. Any video dropped on the website is sent to Gemini API – the Google AI interface for developers – and it comes back in a minute fully audio described. Gemini has the ability to scan a film and identify an uncanny number of contextual details in any video sequence.
The ViddyScribe developers have also found that this technology can be used to open access to home videos or phone videos which have no dialogue or commentary but which have become lost to people who have lost their eyesight over time. The process can bring these videos back to life as shown in this example.
The Human Touch
I hope this post has underlined the complexity and the richness of audio description. The final result is access to the visual via the verbal. It is another form of storytelling which tells a verbal enhancement of the story of a story already told visually making it accessible to people who are blind or have low vision.
A good story always has a human touch. One of the recurrent themes on this blog is finding ways of making people listen. When it is successful, the listener experiences a teller reaching out and making direct contact with their understanding and their imagination. Creating that through audio description requires commitment and, in its screen version, requires the describer to have an eye for film, an ear for dialogue and a taste for the hard work which articulate all that into a written text which will become an oral commentary.
The time and resources required to produce that human touch for all video supports an impossible task. Remember that notion of an hour of work to audio describe a minute of film. There simply aren’t enough audio describers yet, so artificial intelligence will inevitably be part of the process.
The US government’s guidelines on video accessibility have a series of useful tips for the production of synchronized media which you can find here. Among the suggestions is the need for audio description to be part of the creative process in film-making from the beginning : creators need to be aware that at some point an additional soundtrack will be added with the audio description, so gaps in dialogue and unspoken transitions need to be part of the screenplay.
Next time your favourite TV show or movie is proposed in audio description, try the experience for yourself. It is the visual made verbal – and that is transformative.
Still want more?
If you are interested in getting into audio description as a volunteer or as a professional, here’s a good guide on what’s what from Backstage.com.
For a well-referenced article with access to multiple sources, the Wikipedia page on Audio description is also a mine of information.
Stay informed on new posts by subscribing to the monthly newsletter
Notes
- This post from audiodescriptionsolutions.com gives a brief history of audio description in the US. ↩︎
- The formula comes from Joel Snyder‘s 2014 publication The Visual Made Verbal – A Comprehensive Training Manual and Guide to the History and Applications of Audio Description. ↩︎
- If you’re looking for more details on Section 508, click on this link for a summary of the US Laws for Video Accessibility. ↩︎
- All refererences to Hennessy‘s work come from the article Meet the audio describer from tvbeurope.com. ↩︎
- Dune Cherville‘s comments are all taken from an interview from On aura tout vu, a weekly radio show on film broadcast by France Inter on 24th Februrary 2024. ↩︎
- As the European Blind Union says in its 2018 Newsletter : Blind spectators love movies but can’t always find good quality audio descriptions. Each year, a jury of blind people takes three months to assess the audio descriptions of the nominees before attributing the Marius Award to the best one. ↩︎
- These remarks are taken from an interview with Isabelle Faria during the AD recording for the French film Caprice, Emmanuel Mouret, 2015. ↩︎
One Comment
Pingback: