How to Do Closed Captioning Right
Updated: June 3, 2019
You know you need captions, whether to accommodate a student or employee who requests them or to comply with FCC broadcasting rules.
But ultimately, you need to make sure that the captions you end up with are good enough.
To that end, we brought in two closed captioning experts, Jason Stark and Cindy Camp, to review the Described and Captioned Media Program (DCMP) captioning guidelines. They presented in the webinar Understanding Closed Captioning Standards and Guidelines, which you can watch in the video below.
The following are abridged excerpts from the Q&A that will help you produce captions that are accurate, consistent, clear, and readable.
What if the speaker makes a grammatical error? Should we caption based on the speaker or based on incorrect grammar?
- JASON STARK: You basically want to follow what the speaker is saying. Certainly if a speaker is doing false starts, a lot of um’s and uh’s, those can be edited, particularly when you’re faced with rapid dialogue and captions appearing too quickly on screen.
But obviously, hearing individuals are able to hear how a speaker is speaking — whether they have an accent or they’re making grammatical errors. Those details can give insight into the character. We want to make surethat’s conveyed to the viewers of captions as well.
How should you handle foreign languages being spoken in a mostly English video?
- JASON STARK: At DCMP, we caption foreign-translated dialogue in yellow. There’s just a little bit of a difference in color so that it’s very clear that it’s translated. This is possible because we use open captions, which circumvent some of the formatting limitations of closed caption display.
If that’s not an option, then, much like you would for a sound effect, you would add an ID that it was translated dialogue.
What does it really mean to create an “equal” viewing experience with captioning?
- JASON STARK: Equal access requires that the meaning and intention of the material is completely preserved. That’s everything from making sure that you caption the sound effects and the dialogue, accents, the grammatical errors, etc. The goal here is to convey exactly what’s being communicated.
In addition, there are times when you must edit the dialogue. This could be because somebody is speaking rapidly and there’s just not enough time to get all the words on the screen.
We do this at DCMP because we’re focused on educational captioning. If we have a video that is targeted particularly to younger populations–kids who are just learning to read–we realize that 200 words per minute thrown up on the screen is just going to be too much for them to deal with.
So we do have a policy where we edit that dialogue to be shorter and simpler. We don’t want to edit out important vocabulary. We don’t want to change concepts. It needs to be an equal representation.
What font and size are best for closed captions?
- JASON STARK: For closed captioning on most video players like YouTube, the player itself is going to dictate the settings, so you’re not going to have control over that. The caption display is customizable by the viewer, and can be affected by screen resolution and even what browser they’re using to watch your video.
But for standard-definition videos with open captions, we use Arial (a sans-serif font) and a font size of 22. For the high-definition stuff where the resolution is greater, we bump that up to 44.
YouTube doesn’t allow you to add line breaks without copying them from somewhere else. They do this so that word wrapping for larger caption text does not increase the number of lines being used. How should we handle that when considering the 32-character-per-line limit?
- CINDY CAMP: Normally if I’m trying to caption a video on YouTube, I will create the captions in a Word document as opposed to creating them in YouTube. Then I can create the line breaks and upload that. For me, I have not found that it takes any more time to create it in Word and upload that file than it would if I were typing it out in YouTube.
How do you handle profanity or inflammatory language?
- JASON STARK: DCMP is an educational program, so we avoid content that has profanity.
The strict interpretation of equal access is, if hearing folks are going to hear profanity, it should be captioned.
How do you handle a person who speaks with an accent or dialect?
- CINDY CAMP: When it is significant to the character, setting, or story, you can include a descriptor for a character who speaks in an accent. Within reason, you can type out the word phonetically.
For example, you would type “y’all” instead of “you all” if you’re captioning a speaker from the South. You might type “gov’na” instead of “governor” if you’re captioning a speaker with a thick cockney accent.
Should you caption sound effects that are not integral to the plot?
- JASON STARK: That is up to you, the captioner.
If you would have to sacrifice other dialogue in order to fit them in, certainly omit them. But in the spirit of equal access, if you’ve got time, they should be included.
Captioning music lyrics may present a copyright problem. How do we manage that?
- JASON STARK: I guess the best rule of thumb to avoid legal issues down the line is, if you are concerned about that, then don’t caption them. But DCMP’s rule is that we always do. And in 50-some odd years of our program, we’ve not had an issue doing so.
Is there a format for closed captioning that supports caption placement?
- JASON STARK: Yes, several: SCC, STL, WebVTT, DFXP, and SMPTE Timed Text are the caption formats that I’m aware of that support placement.
When we upload files to YouTube, we use SCC files. We’ve had some difficulty getting WebVTT files to be exported in a usable fashion from the software that we utilize. SCCs work beautifully with YouTube.
When music is used to set a mood, would we want to describe the music as ominous or happy to show how the music and visuals relate, or would that be considered subjective?
- JASON STARK: Yes, that would be subjective.
You definitely want to convey the meaning of the scene. You just simply don’t want to necessarily use an adjective that might be your own personal opinion. Objective descriptors of music like “fast” or “soft” would be acceptable.
How should you include captions if there are foreign-language subtitles burned into the video?
- JASON STARK: DCMP guidelines originally stated that you should not duplicate on-screen text.
We’ve since modified our guidelines to be flexible there. Certainly in a situation where you had on-screen text, whether it be the foreign translation or vocabulary words or the new drug that you were watching the video about, you probably want that content indexed. Open captions cannot be indexed by search engines, but if you include a closed caption file, then that will make your video more searchable.
And so we would suggest that you caption that and just make sure that your caption placement is such that you don’t obscure that graphic.
Is there a minimum time that a caption frame should remain on screen?
- JASON STARK: The old analog closed-caption decoders used to have a minimum load time that the captions had to be displayed in order for them to realize the caption needed to be displayed. Nowadays, we keep them up for a second or two.
It’s really going to come down to readability. If you’ve got a very short caption– somebody saying, “Yes”– obviously it needs a much shorter time than a multi-word caption.
How do you handle speakers talking over each other?
- JASON STARK: It’s a challenge. The best option would be to do a hyphen followed by the dialogue on top of each other just to denote that they are two different speakers. But it gets difficult if it’s not clear which of the speakers is saying which of the statements.
Lessons We’ve Learned From Past ACCESS Speakers
We are headed to Washington DC on October 2nd for ACCESS DC. At ACCESS, we dive deep into everything you need to know about video accessibility. Every year, we visit a new city and bring together accessibility professionals and advocates to discuss…
Hey SF! Have You Registered for ACCESS?
Hey San Francisco! Have you heard? We are going to be in town on Tuesday, October 17, 2017, for ACCESS! What is ACCESS? ACCESS is a one-day event hosted by us that celebrates video accessibility and the people who make it happen.…
Hey LA! Have You Registered for ACCESS?
Hey Los Angeles! Have you heard? We are going to be in town on Wednesday, October 18, 2017, for ACCESS! What is ACCESS? ACCESS is a one-day event hosted by us that celebrates video accessibility and the people who make it happen.…