Accessible videos

• Essential audio information: Essential information that is part of the audio track needs to be communicated to people with hearing impairments in another way. This is the idea behind subtitling.
• When are subtitles necessary?
  
  • The short answer is: always. The legal situation (Swiss accessibility standard eCH0059, based on the WCAG) unambiguously requires subtitles for all videos with informative audio content; in other words, all videos that contain more than just background music or sound effects. At ETH, all videos have been released with automatically generated subtitles since September 2023, both on the video portal and on the website (AEM). The quality of these subtitles varies depending on the type of audio in the original video (subject/field, accent, etc.).
    
  • The long answer is: ETH Zurich pursues the pragmatic strategy of focusing on manually edited subtitles for videos with high external impact and reach. Videos of this kind are usually created by communications departments, sometimes with the assistance of external agencies. For example, the videos created by Corporate Communications have long included manually created subtitles.
    
  • Search Engine Optimisation (SEO): Subtitles increase videos' search-ability, especially when the subtitles are closed (not "burnt" into the image). They are text elements in their type, which is why search engines can process them better than other element types. Videos with subtitles profit from this fact and enjoy better rankings and more views.
    
• When are transcripts necessary?
  
  • Transcripts can help people with hearing impairments use audio content such as podcasts or radio interviews.
    
  • Transcripts are used for audio content that is not also visual in nature, such as podcasts. Differences between transcripts and subtitles:
    • Transcripts are used when the media format exclusively contains audio information.
      
    • Subtitles are used when the information is audiovisual in nature.
      
    • Transcripts are usually provided on the original page of the podcast, for example,  ETH Podcast. Transcripts can be displayed directly as text or provided as a separate document. Here it is helpful to format the text so that it matches the flow of the conversation; for instance, by putting space between the questions and answers of an interview.
      
    • Subtitles are directly saved in the video. They are sometimes burned into the image and sometimes can be turned on and off.
  • If you use Audiostudio to record audio, you will be reminded of the requirement to produce a transcript.
    

• Essential visual information: Research videos, explanatory videos, filmed experiments, interviews, films and recorded lectures are videos that typically have both audio and visual content. Essential information that is part of the visual content needs to be communicated to people with visual impairments in another way. This can be done via additional audio descriptions or by structuring the content in such a way that all essential information is also communicated via audio as well. Separate audio descriptions are much more labour-intensive and expensive.
  
• When are audio descriptions necessary?
  
  • The following section is intended to provide information to people who are interested in audio descriptions and who would like to improve their knowledge in this area. Audio descriptions are mostly relevant for movies and television shows.
    
  • Audio descriptions describe purely visual details that are essential to the content and comprehension of the video. These descriptions are in addition to the normal audio track. This helps people with visual impairments process the purely visual details that they would otherwise miss out on. It is advisable to involve external professionals when creating audio descriptions (e.g. SwissTXT, Hörfilm Schweiz).
  • An audio description is not necessary if there is a voice integrated into the normal audio track that narrates all relevant details. This is a cheaper and easier solution.
    
• Preview image (still image): Videos usually include a preview image (still image). These images also have to include alternative text so that people with visual impairments can understand what is in the image. You can use the same alternative text template for each preview; for instance, “video preview for [title of the video]”.

• On websites there are many videos and animated images that are automatically played, such as animated GIFs. These looping autoplay images can lead to problems.
• Autoplay audio content is particularly problematic for the users of screen readers, since this software requires the use of the audio channel. If a newly loaded website brings up autoplay audio content (music, voice, podcasts, etc.), screen reader users will find it incredibly difficult to navigate since the screen reader will be speaking at the same time.
  • Recommendation: You should generally avoid using autoplay audio content.
    
  • If, contrary to this recommendation, you use audio content that autoplays, the audio either needs to stop playing after three seconds or it needs to be manually stoppable; for example, via a button.
    
• Animated images on loop: Looping images (e.g. fades, campaign slogans, banner ads, news tickers, animated GIFs, self-updating elements) demand additional attention and distract from the actual content. People with cognitive challenges such as ADHD can sometimes be heavily distracted by these features to the point where they cannot concentrate on other content.
  
  • Recommendation: You should generally avoid using looping content!
    
  • The user should have the option to pause, stop or hide looping elements, or the animation needs to stop automatically after five seconds maximum.
    
  • Exceptions: The moving element is an animation that shows that content is being loaded.

There are tprotected page hree options in AEM for embedding video content: in ETH webpages. For YouTube and Vimeo, subtitles need to be created and activated in line with those platforms’ requirements (external page YouTube, external page Vimeo). They will be created automatically for the standard video player. The “Video.js” standard player meets the following requirements, which also apply to using video players with subtitles outside of the ETH website:

• Keyboard usability: Is it possible to navigate using only a keyboard (without a mouse)?
  
• Focus visibility: When navigating, is it possible for the user to see which component they are currently focused on (e.g. play button, volume, playback speed)?
• Focus sequence: Does the keyboard focus follow a logical sequence; e.g. going from the upper navigation elements of the player, from left to right, before proceeding to the lower navigation elements?
• Clearly labelled buttons: Are the buttons labelled in a way where the screen reader correctly communicates which function the user is currently on?
• Contrast: Does the play button, which is usually displayed over the preview image, have enough contrast compared to the background image?
  

The player on ETH video portal also supports subtitles, which are also automatically generated.

• Captions (subtitles): In American English, subtitles that are created for audience members with hearing impairments are referred to as “captions”. Captions can either be open (always visible, burned into the video) or closed (able to be toggled on and off). Only in American English is the distinction made between captions (for people with hearing impairments) and subtitles (for translations). In Europe, all subtitles are referred to as such, regardless of the target audience.
  • Captioning a video in this sense is not mandatory according to accessibility standards.
    
• Captions (images): Normally when talking about captions, people are referring to descriptions of an image, table, video or other content. Captions are placed near the items they describe, usually underneath.
  • Captioning a video in this sense is not mandatory according to accessibility standards.
  • Ideally,
    • the caption conveys core information about the content of the video without including any ”spoiler
      
    • Captions can be used to “sell” the video to the audience.
      
    • The following is an example of a caption that teases the content of a video while arousing the audience's interest: “Alkali metals form alkali lyes with water, releasing hydrogen. When performed correctly, this experiment can demonstrate the behaviour of sodium in a way that is both safe and impressive.”
  • If you upload videos to the ETH video portal (Matterhorn), you will be asked for a caption.

Subtitles (automatically generated) can also be used for live streaming content. Livestreamed events run by .

Zoom also offers automatically generated live subtitling that can be activated for meetings on request by .

For live streaming events and Zoom meetings that are recorded, automatically generated subtitles should be manually edited afterwards in order to smooth out any errors for the viewers.