Best Practices for Working with a Communication Access Realtime Translation (CART) Captioner

1. Decide on in-person/on-site CART vs. remote CART. CART services can be provided in person/on site (where both the CART captioner and the audience are in the same location) or remotely (where the CART captioner and the audience are in separate locations). Reasons for using an onsite CART captioner over a remote arrangement include the sound quality (as the location may have poor sound experience), a less than ideal sound system, or a weak internet connection that does not support accurate listening to the event. Good sound quality and experience are vital for the CART captioner to clearly hear what's happening during the event.
   1. There is often a shortage of on-site CART captioners. For this reason, we recommend arranging onsite captioning well in advance of the event (ideally several weeks or months ahead).
   2. If the event sound is clear, free of distractions and has a strong internet connection, using a remote CART captioner can work well and may provide cost savings.
2. Send the captioner any relevant and necessary information to prepare them for the event. The goal of CART is to produce the most accurate message for everyone using the captions. Captioners should be provided with speakers’ names, titles, the agenda for the meeting, any specific acronyms, and the speakers' presentations. If the captioning services are happening remotely, be sure to share the meeting link with the provider in advance.
3. Ask for a transcript (optional). If you would like to have a transcript of the captions, ask the vendor or captioner in advance to send a transcript once the meeting is over. This should be included in the fee for services.

Staff, and especially accessibility personnel in charge of coordinating CART services should familiarize themselves with the event space and setup prior to the event.

In-Person/Onsite CART

1. Prior to the event time, set up a screen and projector for audience viewing. The captioner will provide a link to display the captioning at your event. Captioning can be displayed on a big screen or individual monitor(s). For large events, you can also create a QR code so attendees may access captioning on their own devices should the screen be hard to see or unavailable.
2. Set up a table and chair for the captioner near the presenter and ensure access to an electrical outlet/power source. Their seating and cords should not block the screen, presenters, doors, aisles, or frequently used pathways.
3. Test and display the captions. Ensure the font is big enough, has good color contrast (dark background with white or yellow text is preferred), uses both upper and lowercase (not all caps) and is easy to read. Confirm that the lighting is appropriate and does not produce a glare on the screen.

Remote CART

If both the audience and captioner are remote:

1. Incorporate live captions into your meeting platform (e.g., Zoom, Webex, Microsoft Teams). By including the captions beforehand, the person accessing CART will not have to use a separate device or a second monitor during the meeting.
2. Test and display the captions. You and the captioner should sign into the video conferencing platform 15 minutes prior to the event start time in order to connect and test their services. Turn on the settings for captioning. The captioner will be able to explain the steps to display the captioning, if needed. Although the steps are not complicated, they vary from platform to platform.

If the audience is on site, but only the captioner is remote:

1. Set up a screen and projector for audience viewing. The captioner will send you a link to display the captioning at your event. Captioning can be displayed on a big screen or individual monitor(s). For a large event, you can also create a QR code so that attendees may access captioning on their own devices should the screen be hard to see or unavailable. Ensure the computer used for this does not have a screensaver enabled, as this may cause various challenges during the meeting.
2. The captioner will need to access the event through video conferencing. Set up a conference call for the captioner to listen in. Ensure the captioner has access to high quality audio without interference.
3. Test and display the captions. Ensure the font is big enough and easy to read. Confirm that the lighting is appropriate and does not produce a glare on the screen.

For Both In-Person and Remote CART

1. The captioner must be able to hear clearly, which means it is best to provide a direct feed through the sound system. Often the captioner may have their own headphones. Confirm with them that they do.
2. Ask the speakers to announce their names each time before they speak, to speak slowly and clearly into the microphone, and to pause periodically.
3. Ask for only one person to speak at a time.
4. Monitor the quality of the captions. Watch for “inaudible” or “captioner unable to clearly hear.” You may need to remind the speaker to bring the microphone closer, slow down, or repeat a question from the audience if they are not near the microphone.
5. Take a break every 1.5 - 2 hours if the gathering is scheduled for multiple hours.
6. For all events, provide a URL for CART so that deaf people can watch them separately on their own device at an in-person event or in a separate browser window at a virtual event if they choose. The captioner can provide you with this.

1. You can ask the audience for feedback on the accuracy and timeliness of captioning and for the captioner’s level of professionalism and adherence to the code of ethics.
2. If you requested a transcript, remind the vendor or captioner to send it.

Good captioning should have:

1. A transcription that is done by a trained human and contains appropriate punctuation and speaker identification. Please do NOT use “>>” in any captions, as it is not best practice. Instead, start with the speaker's name.
2. Captioning generated by artificial intelligence or auto captioning may produce some useful captions, but is less accurate than CART; every word and punctuation must be edited and proofread by a trained human captioner.
3. Non-spoken audio descriptions should be used when they are appropriate. Things like [Laughter], [Applause], [Music], etc. are important and should not be skipped.
4. Timing must be accurate. The captions should appear on the screen very close to the time they were spoken so as to not confuse the viewer. Each line of captioning must stay on the screen long enough for the viewer to actually read it. a. When realtime captioners make a caption file without time codes, producing an accurate transcript can be challenging.
5. Captions should not cover up important information on the screen. Avoiding this can be challenging at times. If New York State agencies produce their own videos, they can change placements to avoid placing important textual information in the same location as an important piece of visual information. Examples include covering up an ASL interpreter with a text bubble, obscuring an important object or call to action on the screen, or visual context that makes the text challenging to read. Both the interpreter and the captions should be visible at the same time to be fully accessible for all, especially in times of emergencies.

***Please note that live captions and post-production captions are NOT the same and follow different sets of best practices. In the case of live captions, accuracy is not enough; readability is also an important consideration. Things like proper chunking and font size make this more an art than a science.

Communication Access Realtime Translation (CART): CART is a word-for-word, near-verbatim speech- to-text captioning service for people who have hearing loss and would not be able to fully participate otherwise. Ideally, CART provides a complete translation to produce a transcription of all spoken words and environmental sounds, which empowers consumers to decide for themselves what auditory information is important to them. CART is used primarily for live events, such as educational classes, conferences, business meetings, doctor’s appointments, legal proceedings, religious gatherings, and social situations. Depending on the situation, a CART captioner may be present onsite or delivering the captions to the consumer remotely. When the CART captioner is off site, the service is referred to as “remote CART”.

CART requires a skilled machine stenographer or voice writer with a 96% accuracy rate or higher. CART may be performed at speeds up to and exceeding 225 words per minute. Captioners who often work with certain subject matters develop expertise that is valuable for producing accurate captions and ease of access for people using CART systems.

Closed captioning: A term describing several systems developed to display text on a television or video screen to provide additional or interpretive information to viewers who wish to access it. Closed captions typically display a transcription of the audio portion of a program as it occurs (either verbatim or in edited form) and sometimes include non-speech elements. Closed captions are embedded in vertical line 21 of the broadcast video signal, yet are hidden from the viewer until the captioning is activated or decoded on the TV or Internet video.

Open captioning: Open-captioned text is permanently encoded into the film to integrate it with the image. Open captions do not need to be decoded and are always viewable. Open captions are also embedded in vertical line 21 of the broadcast video signal and are always visible. Open Captions are also used in some settings to provide speech-to-text communication access, but are not comparable to word-for-word CART services.

The National Court Reporters Association (NCRA): An organization that represents a variety of professionals in the field of stenography, especially related to court reporting and captioning.

The National Verbatim Reporters Association (NVRA): An association connecting the Voice Writing Court Reporting, CART and Captioning Communities.

Translation: the process of converting (written text, spoken language or signed language) to another language (spoken or signed language).

Transcription: the process of converting verbal form of information into written form.