Text Size:   A+ A- A   •   Text Only
Site Image

ODHHS Information and Tech Assistance
Categories: Technology, Captioning 
Online (Live) Captioning
(Source: Gary Robson)
How are realtime captions generated?
Realtime captions are performed by stenocaptioners, who are court reporters with special training. They use a special keyboard (called a "steno keyboard" or "shorthand machine") to write what they hear as they hear it. Unlike a traditional "QWERTY" keyboard, a steno keyboard allows more than one key to be pressed at a time. The basic concept behind machine shorthand is phonetic, where combinations of keys represent sounds, but the actual theory used is much more complex than straight phonics.
Stenocaptioners are capable of writing at speeds of up to 250 words per minute, or even faster in short bursts.
The steno then goes into a computer system, where it is translated into text and commands. The captioning software on the computer formats that stream of text into captions, and sends it to a caption encoder. This can be done either directly, or over the telephone using modems.
Can meetings and other live events be captioned?
Absolutely. This can be done either using captions on projected video, or using CART techniques (see "Can captioners replace sign interpreters at live events?" and "What´s the difference between captioning and CART?").
Captioning of meetings can also be done using the Internet to allow remote users to attend (see "Can captioning be done on the Internet?").
Can plays and movies be captioned?
Absolutely. This can be done in the same was as described in "Can meetings and other live events be captioned?". In the case of a theatre where projected video could be viewed by some as a distraction, captioning can be done using an LED sign.
Also, see the description of movie theater captioning in the overview of this FAQ.
Can realtime captioning be done using voice recognition systems?
Speech recognition systems are becoming increasingly accurate and fast. As of today, however, there is no hint of a technology that would actually be able to capture text from multiple speakers in noisy environments with acceptable accuracy. Additionally, much of speech recognition is based on interpreting words using their context. This requires storing up entire sentences, which would leave the captions lagging far behind the speakers. For more details, see the Speech Recognition Technology article that provides a realtime reporting/captioning perspective on the subject.
There is research going on now into the use of speech recognition technology for captioning, but you shouldn´t hold your breath.
Update: There is now a company in Canada with a speech recognition captioning system. The speech recognition is used only for editing or some data entry, however, and the system cannot be used in a realtime environment.
Can captioners replace sign interpreters at live events?
Certainly not. Captioners and sign interpreters address two different groups of people.
Prelingually deaf (or culturally Deaf) people have sign language (usually ASL) as their first language. English came later. They will be much more comfortable with an interpreter, and their comprehension level will be higher. For someone accustomed to ASL, English is quite limited in its expression, and written English is very "dry." Additionally, a Deaf person who has spent most of their life communicating in ASL may not have developed the reading speed necessary to follow captions in realtime.
Postlingually deaf people (or late-deafened adults) learned English before they learned to sign, if they learned to sign at all. For these people, captions will provide a far greater comprehension level. Of the deaf and hard of hearing population in the United States, roughly 10% actually know sign language. Captions benefit the rest.
Each method of communication has its strengths. For example, in a speech with heavy use of proper names and specialty terminology, it may be easier to follow captions than a frantically fingerspelling sign interpreter.
The maximum flow of information and comprehension will occur when sign interpreters and captioners work together at the event.
What´s the difference between captioning and CART?
CART is an acronym for Computer-Aided Realtime Translation, and it refers to the use of machine steno shorthand skills to produce realtime text on a computer. Realtime captioning can be viewed as a subset of CART, but CART also encompasses use of straight text on computer screens (no video picture), projected on walls, or shown on large monitors.
For providing realtime at a live event, it is often easier to find CART reporters than to find captioners, since captioning requires more equipment (and more expensive equipment, as well).
The most straightforward example of CART consists of a CART reporter with a notebook computer and a steno keyboard, sitting next to a deaf person. The CART reporter writes everything that happens, and the screen on the notebook computer is turned so that the deaf person can read it. This differs from traditional court reporting in that the CART reporter is not just there to create a verbatim record, but to help their client understand the proceedings, which may mean paraphrasing, interpreting, and two-way communication.
A larger-scale CART example would consist of taking the video output from that notebook computer and projecting it onto a screen (using an overhead projector with an LCD palette) or placing it on one or more large television monitors (using a VGA -> NTSC "scan converter").
How do I hire a realtime captioner?
Every captioning job is different, so there´s no one answer to where to find a captioner. The following are some general guidelines that can help you get the best person or firm for the job.
To start with, you can get names of realtime captioners from a number of sources, including captioning equipment vendors, your local yellow pages, organizations that use captioners. Once you have some names, talk to the captioners, and ask questions like these:
Do they have applicable experience?
If you want someone for broadcast TV, don´t pick a captioner that only does meetings. If you will have deaf people speaking, make sure they´ve worked with deaf people before.
Do they have adequate staffing?
Not a problem for a single meeting, this becomes a big issue for an annual news contract. If you´ll be using them on an ongoing basis, make sure there are enough people to cover the hours required, and extra for vacations, sick days, emergencies, and so on.
Do they have appropriate credentials?
The CSR and/or the RPR should be the minimum qualification considered for captioning applicants. The CM or RMR is an indication of accuracy under high speed conditions and is desirable. The CRR most directly relates to the type of writing entailed in captioning, but at relatively slow speeds. The RDR is an indication of a broad range of knowledge, including court reporting, captioning, computers and general business management. The previous question in this FAQ provides a list of what all these credentials mean.
Do they have all required equipment?
Make sure that the captioner has everything required for the job beforehand, so you don´t end up with a last-minute panic.
Check their references.
Just as with any other contractor or employee, check references before making your final decision. Talk to people they´ve worked for in the past. Ask not only about the quality of their work, but their reliability as well. If they have any NCRA credentials, these can be verified with NCRA. Similarly, complaints regarding their court reporting performance can be checked with the state´s CSR Board.
Do they have backup?
Especially important for recurring or emergency coverage, this includes not only backup equipment, but backup personnel as well. If the captioner will be working from their own home or studio, this should include at least a second phone line, second TV, UPS (uninterruptable power supply, for power outages), and backup modem. Even better is a second computer system and steno keyboard, and a backup generator for long-term emergencies. There should be enough people to cover everything you need even with somebody out sick. Make sure, also, that their captioning equipment came from a reputable vendor that can help them if problems arise.
See their work!
If you are accepting bids for a large contract (e.g. a year of captioning work), it is quite reasonable to ask to see the applicant´s work. Make sure you see it on something comparable to what you do (if you need a sports captioner, watch their sports captions). Be aware, however, that it is a lot of work for a captioner to build a job-specific steno dictionary with all of the proper names and special terminology for your specific needs. Look for speed, consistency, and accuracy with standard English vocabulary. If that is handled well, they´ll be able to adapt to your needs.
What are the credentials for a realtime captioner?
There is no governing body for captioners, so you must look for credentials assigned either by the state board overseeing court reporters, or by the National Court Reporters Association (NCRA). The skills and knowledge required for these credentials are similar to those required for captioning, and anyone capable of doing broadcast-quality captioning work can easily attain at least the CSR and/or RPR.
For work that doesn´t require broadcast quality and experience (such as captioning an informal meeting), a less-expensive alternative is to use court reporting students that have invested in realtime CAT equipment.
  • CSR-Certified Shorthand Reporter, a state certification, differing from state to state. In California, it consists of writing and transcribing four-voice question and answer testimony at speeds up to 225 words per minute (WPM), and includes a written examination relating to court reporting duties.  
  • RPR-Registered Professional Reporter, given by NCRA. The skills portion consists of writing and transcribing Q&A testimony at 225 WPM, jury instructions at 200 WPM, and literary dictation at 180 WPM. It includes a written knowledge portion relating to court reporting duties.  
  • CM or RMR-Certificate of Merit or Registered Merit Reporter, given by NCRA. The skills portion consists of writing and transcribing Q&A testimony at 260 WPM, jury instructions at 240 WPM, and literary dictation at 220 WPM. It includes a more difficult written test relating to court reporting duties.  
  • CRR-Certified Realtime Reporter, given by NCRA, designed to test writing and technical skills relating to realtime court reporting and captioning. The test consists of setting up realtime computer equipment and writing literary dictation at speeds ranging from 180 to 200 WPM. The examinee is graded on the accuracy of the test as written in realtime. The final portion of the test involves copying the unedited file to a 3.5" floppy in ASCII format.  
  • RDR-Registered Diplomate Reporter, given by NCRA, designed to test a broad range of knowledge associated with court reporting, captioning and related subjects. The highest certification currently attainable, the examinee must first qualify in one of three ways:  
    1. Have passed and held the RMR for a minimum of 5 years; or  
    2. Have passed the RMR and also hold a Bachelor´s Degree; or  
    3. Have passed the RMR, in addition to any two other NCRA tests.
How do I find an online captioning company?
There are several ways to approach this. The best way is through referrals. Ask people who have had captioning done. See what they thought of the people they used. There is also a list of captioning companies put out by Gallaudet University.
In realtime captioning, it is very important to pick people that have done your kind of work before. If you need someone to caption a television newscast in Missoula, pick someone who is going to spell all the names of the towns and people in the area correctly. Pick someone who´s done news work, so that the names of government officials and people in the news will be in their dictionary. If, on the other hand, you need someone to caption a seminar on medical x-ray and laser equipment, pick someone with knowledge of the medical industry.
Ways to find captioners in cyberspace are also beginning to appear. On the World Wide Web, you can check the Verbatim Reporters Center.
Do I need to use someone local for realtime work?
It depends on the realtime work. Most of the network news and sports in the United States is done by people who aren´t even in the same state where the broadcast originates. Generally speaking, if you´re captioning something that´s available on a satellite, you can use anybody with a dish. If you´re captioning something that´s being broadcast, you can use anybody in your broadcast area. If it´s not being broadcast, you´ll want the captioner on-site.
There are exceptions to this rule. If you don´t mind your captioner not being able to see the TV screen, there are companies that specialize in providing captions remotely, using only an audio feed.
See also "If I use a remote captioner, what equipment do I need?"
How much does online captioning cost?
Rates range all over the map, from tens of dollars per hour to hundreds of dollars per hour. Remember that these factors affect rates:
Time: A captioning job requires unpaid preparation time. If you´re paying for a one-hour broadcast or seminar, there´s at least three hours of work involved. Also, remember that captioners usually work alone, unlike sign interpreters who work in pairs and trade off for long presentations.
Equipment: Obviously, someone who is providing $25,000 worth of equipment will charge more than someone who is requiring you to rent equipment for them.
Experience: Just as with any other profession, you pay more for the seasoned professionals.
Notice: If you wait until the last possible minute to call, you will almost certainly end up paying more.
All in all, hiring a captioner is no different than hiring any other professional on contract. Check them out beforehand, pay them what they´re worth, and things will work out well.
See also "How much does offline captioning cost?"
What equipment do I need for online captioning?
Your equipment for online captioning is more straightforward than the equipment for offline captioning. You need a computer, captioning software, and an encoder (plus, of course, the video source). Your captions will then go to a VTR to be recorded, or out to a broadcaster. If you will be providing realtime, then you´ll need a steno keyboard as well.
What skills do I need for realtime captioning?
(This subject is also addressed in my article "How to Get a Realtime Captioning Job", which appeared in the January 1998 issue of The Journal of Court Reporting)
You need to be able to write realtime at speeds well in excess of 225 words per minute, with a total error rate (TER) of under 1.5% to get started. You should have extensive training, on your own and through seminars, to be able to write steno in ways that will differentiate homophones, synonyms, and unfamiliar words, at high speeds and with precision. The National Court Reporters Association CRR (Certified Realtime Reporter) exam is a good start to see if you have the necessary speed and accuracy, even though it tests to a 4.0% TER rather than a 1.5%.
Vocabulary is critical. You must have a well-developed court reporting dictionary, containing all of the specialty terminology for whatever you will be captioning, plus general terms that might come up. Focus heavily on geopolitical terminology. The following is an abbreviated (yes, abbreviated) copy of what should be in an American news captioner´s dictionary, courtesy of Patty White and Kevin Daniel:
  • U.S. Presidents (past and present and potential)  
  • First Ladies  
  • Cabinet Members  
  • U.S. Senate  
  • U.S. House of Representatives  
  • National figures (government, religious, entertainment, criminal...)  
  • All the countries in the world and their capitals  
  • All the states in the U.S. and their capitals  
  • All other major U.S. cities and not so major U.S. Cities  
  • Geographical information, like mountain ranges, mountain peaks, oceans, rivers, lakes, local creeks  
  • Meteorological terms, weather terms  
  • World leaders, including United Nations leaders  
  • World history terminology, like Tet Offensive, Bolsheviks, Chairman Mao  
  • Major wars, domestic and international  
  • Terms related to world organizations, like NATO, OPEC, etc.  
  • Nationalities, and the languages of foreign nations  
  • Supreme Court Justices and important SCOTUS decisions  
  • Military leaders, Joint Chiefs of Staff, military bases, weapons of war, like A-10 tank killers, F-15s, AWACS, Tomahawk missiles, Scud, etc.  
  • National Parks  
  • Colleges and Universities and their team names  
  • Local stuff for the area you plan to caption in, landmarks, hospitals, junior colleges, prominent people...  
  • All the professional sports teams and their nicknames, players and coaches  
  • Sports terminology, awards, and organizations, both amateur and professional  
  • Makes and models of automobiles  
  • Major businesses, brand names, stock terms and trading organizations  
  • Currencies and major banks around the world  
  • All the colors and their shades and hues  
  • Dog and cat breeds  
  • Foods and their measurements and spices that go in them  
  • Holidays, Christian, Jewish and all other faiths  
  • Special events, like Kwanzaa, Cinco de Mayo, etc.  
  • Religious structures, like mosque, temple, etc.  
  • Books of the Bible and religions around the world  
  • Basic chemicals  
  • Drug names and manufacturers, both legal and illegal  
  • Common first and last names  
  • World literature, philosophy and religious terms, like Aesop, orthodox, Buddhism, Eucharist, the Brothers Grimm, etc.  
  • Political terms, like Glasnost, apartheid, anarchy, propaganda, caucus, expatriate, Kremlin, Parliament, etc.  
  • Computer terms  
  • Idioms, like adieu, aloha, de rigeur, fait accompli, modus operandi...
How does a realtime captioner handle obscenities?
Another issue facing the captioner is swear words. As a general rule of thumb, if the word is in the sound track, it should be in the captions as well. If the producer chooses to "bleep" the sound track, then "bleep" the captions, too. Deaf and Hard-of-Hearing viewers should get the same content as hearing viewers.
These obscene words must be in the dictionary, because they will come up sooner or later, but they must be carefully defined so that it is virtually impossible to stroke one by accident. One common approach used by a number of captioners is to place the words in the dictionary with an asterisk in the middle of the steno stroke. Beware, though: if you remove the word from the dictionary entirely, today´s sophisticated phonetic translation systems may produce it anyway if you write it by accident. Put the phonetic way of writing the word in your dictionary as a NULL translation so that if you hit it by accident nothing comes out.
You´ll need a "bleep" stroke as well, for when the audio is actually censored.
How do I get started in the realtime captioning business?
(This subject is also addressed in my article "How to Get a Realtime Captioning Job", which appeared in the January 1998 issue of The Journal of Court Reporting)
The best way to get started is just like any other business: Start calling the leading companies in the field (and the local companies) and see who´s hiring. You can also start by working with local chapters of organizations like the Association of Late-Deafened Adults (ALDA), Self-Help for Hard of Hearing People (SHHH), the National Association for the Deaf (NAD), the Alexander Graham Bell Society, and others that may need your services. They pay won´t be as good, but the satisfaction level is high.
Sometimes the major companies hire people with little or no training under internships or on-the-job-training programs, but there´s no substitute for having a little bit of experience behind you. Practice on the kind of material you want to caption, and offer to demonstrate your skills.
If you are planning to do work for the television news industry, get to know the business and the terminology before you approach anybody. Make sure your equipment is in place and that you know how to use it. Certification is not required, but is a good thing to have. Many captioning companies will only hire CRRs or RMRs, and very few will hire someone without a CSR or RPR. If you´re not familiar with these acronyms, see "What are the credentials for a realtime captioner?"
Can I do realtime captioning from my home?
Some kinds of captioning lend themselves well to working remotely. Certainly, you can do a news broadcast for a local television station if you have a computer, modem, steno keyboard, two televisions, headphones, and the appropriate software. As you get started in the business, though, plan on working on-site. The people you work for will feel better if they can see you. It is also easier to caption things like city council meetings when you can be there to see what´s happening. Obviously, live events that aren´t broadcast will need you to be on-site as well.
Why would a captioner need two televisions rather than one?
There are two very good reasons:
  • Redundancy: If one goes out, you´ve still got the other. It is even better if they have different signal sources, like one connected to cable and the other to an antenna.  
  • One with captions, one without: This lets you see your captions, and also see what your captions might be covering. If an important graphic or picture comes up, you´ll want to blank or move the captions. It also makes it possible to get spellings of proper names from the character generator text you otherwise wouldn´t see.
This document is Copyright ©1995-98 by Gary D. Robson. It may be freely duplicated and distributed only in its entirety. No modifications may be made to this document or any of the files that comprise it, including removal of this paragraph, and no excerpts may be taken, without prior written permission from the author. If you distribute this document, you may not charge for it, or include it with anything else that you charge for without prior written permission of the author.