Video accessibility: Captioning, subtitles and transcripts

What is alternative text?

Alternative text takes content from videos, like speech and sound, and presents them as text. This helps people who cannot hear or see the video to understand what's happening. Another example of alternative text is the 'alt' field that's used to describe images on the web or in documents.

Alternative text for video is delivered as subtitles (also known as captions) and transcripts.

Why do we need alternative text for our videos?

Accessibility matters

As a public body, we must provide alternative text for our video content to make them accessible for disabled viewers. This includes people who are deaf or have a hearing impairment, but also those who have other disabilities that make it difficult to navigate video content.

By making our content accessible for disabled people, we make it easier to access for everyone. With captions, not only do we help those with a hearing impairment, but we also make engaging easier for those listening without sound or in noisy environments, as well as those for whom English is not their first language. 

If more people can engage with our content – and engage with it well – the better we can reach the goals intended by that content.

Subtitles and transcripts help everyone consume video content in many different ways:

• Anyone working in a noisy environment can read subtitles
• Non-native speakers can read subtitles in their own language
• Students learning to read can follow along with the speaker
• Anyone can search the video's content for key terms using the transcript

Subtitles

Subtitles are the text on the screen showing what people say in video content. They may also have audio descriptions when people are not speaking. There are different types of subtitles – closed and open. In general, we should be creating closed subtitles where possible.

Closed subtitles 

Closed subtitles are where the subtitles are an overlay on the video platform and can be toggled on and off by the viewer, are currently preferred for social media and digital platforms where the user can choose to use or not. 

Open subtitles

Open subtitles are where the subtitles are part of the video and are “burned-in”, are preferred if your video is to be shown on an internal screen or another platform where users do not have the capability to control subtitles. 

Should I use closed or open subtitles?

Choose whether your subtitles are closed or open based on the medium. For example, if you're using social media to communicate with your audience by video, then closed subtitles are the best option because that audience may want to switch them on or off.

However, if your video will be shown on an internal screen around campus or as part of a presentation, for example, then open subtitles would be the best option.

Captions vs subtitles

In American English, "captions" are text for helping people who cannot hear the audio, whereas "subtitles" are English translations on foreign language films. In British English we use "subtitles" for both concepts. Some software may (incorrectly) use "CC" to refer to open subtitles.

Subtitle tips

Here are some general tips to help ensure your subtitles are as useful as possible to the people who rely on them.

• A subtitle should be a full sentence, spread over two lines (though this may vary slightly in practice)
• Subtitles should be as true to what is said on screen as possible (but omit "ums" and "ahs")
• You should add in audio descriptions when people are not speaking in square brackets – e.g. [Music playing], [Alarm rings] or [No sound]

How to create subtitles

The actual method you'll use for creating subtitles for your video will depend on what tool you have available to you. Please avoid relying completely on subtitles that are generated automatically by any tool as they cannot be entirely accurate due to reasons like these:

• strong accents
• complex terminology (scientific or mathematic terms, for example)
• unusual names of people and places
• other factors that require manual human input

Using YouTube to generate subtitles

YouTube's automatic captioning service provides a solid start to captioning your videos.

• Upload your video to YouTube
• Wait two to six hours
• Log back into YouTube and go to your Video Manager
• Next to the video you want to edit subtitles to, click the drop-down menu next to the Edit button
• Select Subtitles and CC
• Click Automatic Captions
• Edit the existing subtitles

Although YouTube's automatic subtitles are far from perfect, it will get you 80% of the way there. The clearer the speaking voices and diction, the better results you'll get. The best part of the automatic captioning tool is that the time codes are synched to your content. All you need to do is edit the existing text in each frame.

Subtitles should be a maximum of two lines, and each line should be no more than 37 characters (though use discretion if there is a particularly bad break – try to break the line earlier).

Using Microsoft Stream to generate subtitles

Stream can automatically generate subtitles using Automatic Speech Recognition technology. It's worth mentioning again that automatically generated captions are not accurate enough to be used without manually corrections and input.

Automatic subtitles take time to create. It typically takes 1-2x the video's duration, so for example, for a one hour video, expect to wait approximately two hours to finish processing. Here's how to do it:

1. Upload a video for which you want to enable autogenerated subtitles. If the video is already uploaded, go to the Edit video page
2. In the Details section, select a supported language
3. On the Options tab, set 'Autogenerate captions' to On
4. Stream will automatically begin generating subtitles

You can then edit autogenerated and uploaded subtitles using the transcript window, or by downloading the files from the Edit video page.

Using YuJa to generate subtitles

YuJa's Video Platform includes an auto-captioning feature. This feature can create an automatically generated word-by-word searchable closed-caption file of your media that can be edited using the Video Editor. 

1. Hover over a media thumbnail and select Edit to launch the Video Editor
2. Click the CC button to pull up the auto-captions
3. Below is the list of all the available actions users can use:
   • Update the auto-captions timestamps
   • Edit the auto-captions texts
   • Delete a caption line
   • Add a new caption line
4. Once done, click the Save button to save all the new changes

View more information about managing accessibility options in YuJa.

Using Microsoft Word to generate subtitles

If you have a video project you need to provide closed subtitles for, you can use Microsoft Word Online. This method should be used whether you are self-processing or working with colleagues in Media Services, Marketing, or another part of the External Relations directorate.

You will need:

• Your completed video (save a lower quality version or just the soundtrack of this to speed up transcription time)
• Microsoft Word Online
• Time to proofread and edit your subtitle transcript

Step 1

Open Microsoft Word Online (office.com/launch/word) and open a new blank document. Select Transcribe on the Home tab:

Step 2

Click Upload audio and upload your video file.

Step 3

Once the file is transcribed, open it in the document using the dropdown menu at the bottom right of the screen. Select With timestamps.

Step 4

Delete the text at the start of the transcription so you are left with only the content for subtitles. Increase the point size of the text to 20 (if using Calibri font) and add guides (find them on the Home tab).

Go through the audio file, making corrections to the text and splitting lines so that the subtitles display in consistent blocks of text. Subtitles should be a maximum of two lines, and each line should be no more than 37 characters (though use discretion if there is a particularly bad break – try to break the line earlier).

• A soft return (holding shift and pressing return) will move the content to the next line but keep it attached to the subtitle content above
• A hard return (pressing return only) will move the content on to a new line, starting a new subtitle

The text in your finished file should be split into two-line blocks with a timestamp between each block. You will need to add timestamps between some of these blocks, so use the audio player at the top right of the file to work out what these should be. If there is a single line where a sentence ends and the audio stops, add an additional timestamp between these.

Don’t forget to add in any additional audio descriptions where relevant.

Step 5

Once you have edited your document online, download a copy to your desktop. Then save a version from that desktop copy as a .txt document (Word online doesn’t currently allow export as .txt). 

Use this naming convention to help us keep track of files:
[School/Directorate] – [Department] – [Video/project title] – [Your name] - [Date]
e.g. External Relations – Marketing – Interview with Joe Bloggs – Jane Doe - March 2021

Mac users: If you are on a Mac, Word may change apostrophes to an Í when you save as .txt so make sure you change them all back.

Step 6

Send your completed video and .txt file to Media Services (mediaservices@dundee.ac.uk) who will be able to upload them to YouTube for you to review. If Media Services are unavailable, you can also contact Joel Hewett in Marketing (j.hewett@dundee.ac.uk) to help.

The video should remain unlisted until you have checked the subtitles are in time with the video. 
Once you are satisfied that the subtitles are correctly aligned, request a download of the .srt file. This can be used to upload the video to other platforms like Facebook and Twitter.

If you need subtitles to be embedded in the video (open captions) you can send this .srt file to the person producing the video to have them added in. This should only be used for instances where closed subtitles cannot be used (e.g. digital screens, presentations).

You can now use the video for its intended purpose.

Open subtitles

If you are commissioning a video with open subtitles from a third-party supplier, please ensure that these guidelines are followed:

• Open subtitles should be Baxter Sans font, size 48 (assuming your video is 16:9 ratio)
• Closed subtitles will use the template of the platform the video is uploaded to
• A subtitle should be a full sentence, spread over two lines (though this may vary slightly in practice)
• Each subtitle should be no wider than 65% of the screen. Again, this may vary on platforms, but as a rough guide:
• For 16:9 ratio this equates to around 35 characters per line 
• For 1:1 ratio this equates to around 20 characters per line

Transcripts

Every video we display online requires a transcript. A transcript is a pure text translation of a video's audio, without timestamps. Screen readers and other accessibility tools use transcripts to help people who can't see the video understand the video's content.

Some visitors, including non-native English speakers, may prefer to read a transcript rather than watch a video so we should format transcripts for readability by using paragraphs. No other formatting, such as headings or lists, is required.

However, to deliver the best possible experience, a video transcript should communicate more than just speech; it should also clearly identify who is speaking and include essential non-speech sounds using square brackets, for example [LAUGHTER] or [SILENCE].

Here is an example of ideal transcript formatting:

>> ALICE: Hi, my name is Alice Miller and I am a lecturer here at the School of Business

>> JOHN: and my name is John Smith and I'm a recent Accountancy graduate

>> ALICE: Today, we'll be explaining all the benefits of studying a business course here at the University of Dundee 

[Intro music]

Transcripts also make your videos searchable by search engines.

How to create transcripts

Most video tools let you generate both subtitles and a transcript. Again, be wary that without manual corrections, an automated transcript will contain a significant number of errors – enough to give the people relying on them a bad experience.

Transcripts in YouTube

Here is the method to generate a transcript from YouTube. Only do this after first generating and correcting the subtitles for your video.

1. Play the video for which you need a transcript
2. Click on the three horizontal dots below the video
3. Select “Open transcript"
4. The transcript will be visible on the right of the video
5. Copy the transcript

You can then paste your transcript into a document for sharing or paste into the transcript field on the website.

Transcripts in Microsoft Stream

Stream provides an option to display a scrolling transcript adjacent to your video. Here's how to show it:

1. In Stream, go to a video, and click View settings
2. Click Show transcript

You can make edits to the transcript directly in Stream when viewing your video. It is possible to copy and paste the transcript from Stream but it can take a long time to highlight each line of the transcript for longer videos.

Transcripts in YuJa

YuJa's Video Platform includes an auto-captioning feature. This feature can create an automatically generated word-by-word searchable transcript of your media that can be edited using the Video Editor.

View more information about managing accessibility options in YuJa.

Further assistance

For further assistance with creating your subtitles and transcript, please contact Media Services (mediaservices@dundee.ac.uk) in the first instance.

For questions about how to best display your video on a website in an accessible way, please contact Web Services (web-accessibility@dundee.ac.uk) or create a support enquiry on Help4U.