Live Captioning/Transcription

What is Live Captioning/Transcription?

Live captioning/transcription is the means by which captions are added to live lectures and events. Captions are generated in real time, either by a human or by a computer.

The terminology for live captioning can be a bit confusing. It can be called any or all of the following:

Live captioning
Live transcription
Closed captioning
Real-time stenography
Real-time captioning
Speech-to-text services
Classroom transcription
Text-based accommodations

Types of Live Captioning

There are two types of live captioning: captions created by humans and captions created by automatic speech recognition (ASR). Human-created captions are the only method that meets accommodation standards.

There are two ways human-generated captions are created. The following definitions are adapted from “Communication Considerations A-Z” and “Speech-to-Text Services: An Introduction.”

CART (Communication access real-time translation)

CART is a method of live captioning that provides “word for word” (verbatim) transcription. This is the type of captioning/transcription provided by court stenographers and is also used for broadcast television. CART providers are highly trained and used specialized equipment, software, and techniques. Stenographic equipment is connected to a computer where the words appear in English for the viewer to read at speeds of up to 300 words per minute.

CART transcription includes nearly every word spoken, including false starts, misspeaks, and filler phrases. CART is the standard for live events and broadcast television.

Meaning for meaning

Meaning-for-meaning service providers listen to spoken language and translate it into grammatically correct written language. False starts and misspeaks are typically eliminated, which results in fewer words than CART transcription.

Captionists who create “meaning for meaning” transcription use one of two types of specialized software: C-Print or TypeWell. These programs are run on a standard laptop and captionists are trained in abbreviation standards and text-condensing strategies. As with CART, the viewer will see regular English words on a laptop or other device. Meaning for meaning is often appropriate in classroom settings (or online instruction) but is NOT used for broadcast television.

Automatic Speech Recognition (ASR) is the way in which computers convert audio into text. ASR is used in services such as Siri, Google Home, Alexa, and is also used to generate live captions. While ASR has improved significantly over the past few years, at this time, ASR does NOT consistently meet accessibility standards. ASR can be quite accurate in situations where the speaker is easily heard (and does not have a strong accent), the audio quality is high, and the words are relatively uncomplicated. ASR produces higher error rates when those criteria are not met. Different ASR services offer different levels of quality.

There are a number of tools and programs that offer ASR-generated captions:

Zoom (see information below)
PowerPoint
Microsoft Teams Meetings (currently available in Teams Meetings)
Google Meet
YouTube

Live Captioning as an Accommodation

Human-created captions are provided as an accommodation at IC. There are two primary ways that live captioning is used as a support service at IC:

As a classroom support for students who are deaf or hard of hearing. In this scenario, a transcriptionist types out the content of a lecture or discussion while a student reads the captions from a computer or mobile device in real-time.
Captions are used for live video streams (closed captioning). Zoom meetings and webinars, YouTube livestreams, and Kaltura livestreams can all be captioned in this way.

SAS provides the funding for this accommodation in academic settings for course requirements.

Contact SAS to learn more about this accommodation

Comparison of tools that offer LIVE captioning support

The table below lists the options for live captioning for many of the tools used at IC. Information about adding captions to pre-recorded videos can be found on the Post-Production Captioning page.

Platform	Human-created captions (3rd party)	ASR	Are live captions recorded?
Zoom	Yes	Yes	Yes
Microsoft Teams	No	Yes	No
PowerPoint	No	Yes	No
Kaltura (livestream)	Yes	No	Yes
YouTube Live	Yes	Maybe	Yes

Live Captioning Tech Guides

Below are some primary tools used at IC and guides for faculty/staff when Live Captioning is a needed accommodation. Please note: SAS provides the funding for this accommodation in academic settings for course requirements.

Zoom allows for both human-created and automatically generated captions. Meeting hosts have control over both methods. The ability to implement live captions is set in each users' individual account settings. To view these settings:

Sign in to ithaca.zoom.us
Navigate to Settings -> Meetings -> In Meeting (Advanced) -> Closed Captioning
Toggle the switch to activate or deactivate closed captioning
Check the box to enable live transcription

Once closed captioning is activated, all future meetings will give the host the ability to use this feature. Captions can be enabled within the meeting by clicking the "CC" button at the bottom of the screen. When live transcription is turned on, the words "Live Transcript" will appear on the "CC" button.

To initiate captioning, there are four options:

The meeting host can turn on live, automatic captions by clicking the "CC" button and selecting "Live Transcription."
The meeting host can assign captioning duties to a participant in the meeting. The participant would then type directly into the captioning window.
The meeting host can pass an API token to a meeting participant. This is also done through the CC function. This allows the participant (usually a contracted service) to connect specialized software to the Zoom meeting, which allows for more efficient captioning (software bumps up typing speeds from standard typing speeds up to 300wpm in some extreme examples).
The last option does not use the captioning function directly. It is possible for the captionist/transcriptionist to join the meeting as a participant to hear the audio but use separate software to convey captions to the student (generally through a website). This eliminates the need for the meeting host to assign captioning duties for every meeting. The student just needs to make sure the captionist has the Zoom link.

Microsoft Teams offers live auto-captions in scheduled Teams meetings. This feature can be turned on or off by any individual user. Captions are only visible to the person who enables them.

Use live captions in a Teams meeting

Microsoft PowerPoint has the capacity to show live captions during presentations. When presenting using a computer with a microphone, it is possible to turn live captions on that will be shown below each slide. This feature must be enabled before it can be used for a presentation.

Present with real-time, automatic captions or subtitles in PowerPoint

Live captioning in Kaltura (media.ithaca.edu) is used only during large events that will be streamed live (such as graduation or an all-staff or faculty meeting). If you would like to learn more about live captioning in Kaltura, contact Information Technology at servicedesk@ithaca.edu.

Live Encoding Best Practices Guide

YouTube Live allows for both human-generated and automatically generated captions. However, automatic captions are only available for channels with over 10,000 subscribers at this time. 3rd party, human-generated captions must be set up by the meeting host.