I’ve been consulting and presenting on captioning access for over 10 years, written a book (currently it’s undergoing a new edition), and given a TEDx talk about this. I’ve also seen many podcasts, videos, and events that are either not captioned or have captions that are not easy to read due to poor formatting.
Captions are more than just adding text. Captioning is an art and not something that can be learned overnight. Anyone can make captions, but not everyone can do it well. It’s like writing a book – anyone can write a book, but not everyone is a professional writer and they also need to work with professional editors and designers. That’s why expertise is needed to make captions and transcripts look professional and easy to read.
There are captioning guidelines adopted by many organizations, but based on my experience, they are not consistent and some are outdated. Many of them also do not consider our actual experience as deaf and hard of hearing users for whom captioning access is the only way to perceive aural information. It’s important that captions and transcripts are not only accurate, but also formatted in a way that improves reading experience for us deaf and hard of hearing people.
So I did a survey in 2019-2020 on the captioning reading experience and asked deaf and hard of hearing people aged 18 or older for feedback on their experience. The survey had 16 questions, and 14 were multiple choice.
I would like to note that is a very complicated subject. It was challenging for me to draft a list of questions. I had to narrow down to the most common issues that we deaf and hard of hearing experience with caption reading. There are many variables to consider when discussing those issues and more research is needed. So this survey is not meant to represent the comprehensive research on captioning preferences, only to show some common preferences of the majority respondents.
I will explain only selected parts of the survey in this article. For more stats and details, you can check my summary in Google Docs. You can also check images there.
The survey has 305 responses from people living around the world. Most of them are in the USA, followed by people from Australia, Canada, UK, and some other countries.
According to the survey:
- 87.2% use captions all the time;
- 57.4% have used captions for 20+ years;
- 93.4% watch captions in online web videos;
- 64.9% are not familiar with captioning quality standards.
The reason I asked people if they are familiar with captioning quality standards is because not all deaf people are aware of the standards even if they have used captions for a long time. Many deaf people may put up with low quality captions thinking they are better than nothing. For example, I was not familiar with captioning quality standards until after I started doing research on captioning access – even though I have been a long time captioning user who is deaf. I have noticed a lot of inconsistencies in quality and styling standards, even in professional content.
The following are my interpretations of the responses to questions about caption reading preferences.
Text color contrast of captions
To the question about text color contrast of captions (numbered 6 in the summary), 55.7% of respondents said that they prefer letters in a 100% opaque (non-transparent) box.
It’s easier to read captions when white letters are in a fully opaque black box for easier readability. It doesn’t have to be white on black. It could be black on white or yellow on dark blue. It’s important that color contrast between letters and the box makes it easy for us to read and especially for people with color deficiency (aka color blind). The absence of a box makes it hard for us to read captions because letters blend into a background. It’s like trying to listen to speech in a noisy place.
Caption location in a video
To the question about caption location in a video (numbered 7 in the summary), 48.2% of respondents said that they prefer captions below the image.
It would be ideal to place text on a plain background below the image. While you can place letters in a box on the image, they are easier to read when there’s plain color behind letters and no busy background. The main reason is because the image may have other text elements like a score board or news ticker that may be overlapped by captions.
Font style of captions
To the question about font style of captions (numbered 8 in the summary), 85.2% of respondents said they prefer sans serif font.
Captions in sans serif font are generally easier to read on web and TV than serif. It’s also based on UX research.
Letter case of captions
To the question about letter case of captions (numbered 9 in the summary), 74.8% of respondents said they prefer sentence case.
I wrote an article about all caps vs mixed case type for captions. The reason some TV sets show captions in all caps is because in the past the old technology could not render the mixed case well. Many people, including myself, may be used to all caps, but when we are offered a choice, it would be easier for many of us to read captions in mixed case. I would like to see more research on this because I wonder if those who are used to all caps in TV captions may take time to get used to mixed case, among other things.
Caption chunking
To the question about caption chunking (numbered 10 in the summary), 95.7% of respondents said they prefer captions in 1-2 short lines.
Many post-production videos have captions that are chunked randomly and more than 2 long lines and sometimes with small or thin font that makes it hard to read captions. When captions are in long lines with small and thin font, you spend more time trying to read text and may miss what’s going on in the image. Better chunked captions with good font size and weight allows you to spend more time to look at the image than text.
Live captions at events
To the question about live captions at events (numbered 11 in the summary), 39.7% of respondents said they prefer captions on a separate screen full of captions next to a screen with slides.
I had to add that it is not for movie theaters as the question made it sound like it’s for movie theaters. Based on the question numbered 5 in the summary, 50.2% of respondents watch captions at live events, webinars, classes, work meetings, etc – the lowest of all options for where they watch captions. Based on my experience interacting with deaf and hard of hearing people and their experience with accessing live events, many of them are not aware that interpreters are not the only way to access information at live events and that they can request live captions for those settings if they don’t know sign language. Also, live events happen not only on site, but also online (more frequently now due to the pandemic). Caption settings vary from event to event depending on the size, number of audience, AV setup, and so on. For those reasons, I provide customized consulting services to event organizers on optimal accessibility strategy of their live events – online and offline. For this question, I picked 4 general caption settings at an on-site event to see how many prefer which option.
Post-production captions in videos
To the question about post-production captions in videos (numbered 12 in the summary), 71.1% of respondents said they prefer offline captions.
As per quality guidelines, post-production videos (including post-event recordings) are expected to have off-line captions. As I explain in the article, live captions at an event are not the same as post-production captions for videos. While it’s tempting to record live captions for post-production videos, it’s not the best practice because videos are expected to have post-production captions that synch with speech, among other things.
Additional comments
Instead of asking for explanations for each question, I decided it was better to offer one optional text box in the end of the survey. It gave the opportunity to respondents add any thoughts they wanted to add regarding the questions they answered above.
You can see all comments in the last part of the summary.
Many people complained about the lack of captions and asked that captions need to be available everywhere.
Some people complained about the lag and errors in live captions. As explained in the part about the question 12 above, live captions and post-production captions are not the same. So the delay and errors are inevitable and expected in live captions – as long as the delay does not last longer than a few seconds and errors are no higher than 2-3%.
Some people said there needs to be options for users to determine styling preferences. I agree. YouTube offers that, but not many other video players.
Conclusion
My personal preference for videos would be that captions are displayed on a blank background below the image – as shown in the beginning of the article.
Also, all web pages with embedded videos need to include a descriptive text transcript below the video in HTML (not as a separate PDF or DOC to download unless it’s complementary to HTML).
You are free to share the caption survey results and my interpretation, but please credit Svetlana Kouznetsova and Audio Accessibility and link to this article and the survey summary.
I also welcome your questions and comments – you can reach to me via the online form.