In part one of this article we examined the first step in making videos accessible: solving the keyboard challenge. In today’s post we’ll cover how to make excellent captions and add a descriptive transcript.
Step 2: Making excellent captions, the easy way
YouTube has the magical ability to automatically generate fairly accurate captions for uploaded videos, saving you much effort and drudgery. However, these automatic captions are never good enough, and can be quite poor especially when there are multiple speakers or background noise.
We’ll start with what YouTube provides, then improve upon it manually.
Upload your finished video (never start captioning until the video is final!) to your YouTube channel. Set it to Private (as we don’t want to share it with the world yet), then wait several hours for YouTube to generate its captions. The longer the video, the longer it takes.
Now, use YouTube’s caption editor to clean up the captions:
- Words: fix spelling, and remove “ummms” that get in the way of understanding.
- Pacing: shift words to the next or preceding caption to ensure complete phrases never bridge two captions.
- Consider deleting any time segments which are left blank, then select the timer on your full-sentence caption and increase the time to run for the duration of the full sentence.
Here are some guidelines to help you get started with the art of improving your captions:
- No caption should appear for less than two seconds.
- Add descriptions of sound in square brackets (such as [music] or [laughter]) to help people understand what is happening.
- If there’s more than one speaker, add tags like ”>>SAAB:”, at the beginning of a new line, to identify speakers or change of speaker.
- If someone is spelling a word, caption it with hyphens, as in e-X-p-l-o-r-a-n-c-e.
Step 3: Adding a descriptive transcript
To achieve Level A compliance, you need to include either a descriptive text transcript or an audio description (WCAG 2.0 1.2.3). Most people will choose the former because audio descriptions take more effort.
YouTube wonderfully provides a transcript button, enabling the viewer to read all the captions in one place. But a transcript of just the captions is not sufficient: we need descriptive text transcripts.
So what you’re going to do is take the complete transcript of your captions and add descriptive text that relates what else is going in the video (“location shot”, actions, body language, scene changes, etc.). Then add a “Transcript” button directly below the video on your web page and have it link to a separate HTML page containing the transcript.
To see an example of what the finished product will look like, visit davidberman.com/accessibility.
How to create the descriptive text transcript:
- Go to the YouTube video’s Transcript (select the Transcript icon) and copy and paste it to a text file.
- Use a text editor to remove the timecode by hand or save time by downloading the .srt file from YouTube’s caption editor, and using the free software Aegisub for Windows, Mac, or Unix (download at Aegisub.org) to remove the timecode.
- To avoid broken sentences, search and replace any extra hard-returns. Also replace all line endings with single spaces.
- Insert any descriptive text (you can conveniently take text from an audio description script if you happen to have one). For example: “The speaker is standing in front of a tiger cage. Throughout the entire video, he addresses the camera directly.”
- Insert the resulting text into the page on your website that is linked to the “Transcript” link below the video.
Keep in mind that you could have, instead, provided an audio-description version (or perhaps even a Level AAA extended audio version) of your video, in which case you don’t need a descriptive text transcript. (For an example of an extended audio description, please see the video at wcag2.com.)
So there you have it: a recipe for accessible video you can do yourself, broadening your reach, complying with regulations, improving search, and broadening your entire audience. See you online!