Realtime captioning is a translation. A "stenocaptioner" is a highly skilled reporter who is specialized in the field of captioning. A stenocaptioner "writes" the captions with a steno machine, the same machine that you see used in courtrooms today. You may remember seeing a court reporter on TV throughout the trial of O.J. Simpson. The court reporters that were assigned to that case were also writing in realtime. Every word that was spoken in the trial was translated instantaneously onto computer monitors placed in front of the judge, jury and all parties to the trial.
As I mentioned, realtime captioning is a translation, much like Spanish is translated into English. There are 22 keys on a steno machine. Not every letter of the alphabet is represented by a key. Those particular letters are formed by depressing a certain combination of keys. The art of stenography involves the use of a "code," of sorts, for every word in the English language. This code is based on phonetics.
In other words, captioners will stroke onto the machine every word as they hear it. For instance, the name "Vicki", would be written SREUBG/KI. That is the phonetic outline, or the sound of the name in the language of steno. As you've noticed, these letters do not even remotely resemble the English spelling, and you're probably wondering how this language can be a "shortcut" since there are actually more letters involved. Here's the simple answer: Even though there are three more letters in the steno outline, the entire name can be typed with only two "strokes" of the keyboard, unlike the QWERTY keyboard which allows only one key to be hit at a time. This is the case with most words in the English language, and that is what allows the stenocaptioner to write at the high speeds necessary to caption in realtime, often in excess of 240 words per minute. (Here's the secret code to the name, "Vicki": SR are combination keys that represent the "V," the EU represents the short "I" sound, and the BG keys represents the "K" sound. Then KEU represents the second syllable, or sound of KI.)
Attached to a steno machine is a cable that runs to a computer. Loaded into the computer is software that captures the stream of data that is input from the steno machine. Included in the translation software is a personalized "dictionary" that contains tens of thousands, sometimes hundreds of thousands of English words that are matched to steno codes.
In addition to common, everyday words and proper nouns, VITAC's stenocaptioners spend years building these dictionaries to cover a wide variety of captioning topics, including world political names and terms, names in the entertainment industry, and all sports figures. As you can well imagine, proper names are the most difficult for a stenocaptioner to write. When the stenocaptioners' dictionaries are prepared for an event like The Olympics, they will add as many as 5,000-10,000 proper names to their individualized computer dictionaries.
So, all that they need to do for accurate realtime translation is stroke every word that they hear precisely as it is programmed in their dictionaries at speeds of up to 240 words per minute! This is a task that requires very intense training and continual development.
Contained somewhere within this computerized dictionary is the English translation for the steno code SREUBG/KI that tells the computer to translate this stroke as "Vicki." If the stenocaptioner were to "mistroke" this outline by inadvertently pressing a single wrong key or an additional key or by omitting one, the resulting translation will most likely be a word or series of letters that do not even resemble the name.
Last names are far more difficult for us, and I'm sure you can imagine how difficult this whole process becomes when the writing involves technical language or when the speaker picks up speed. The chance for error increases with the length and difficulty of each and every word.
Nonetheless, VITAC stenocaptioners are some of the best in their field. Our goal is to minimize the amount of errors with each and every day of experience. There is approximately a one-hour preparation time for every hour of programming, as well as a one-hour follow-through. Follow-through involves reading the entire file that was written, research of any terminology or names that were not recognized, and making additional entries to the dictionary so that similar errors do not happen in the future. Each of our stenocaptioners must submit work for weekly quality control checks, and ongoing records are kept on the progress of all captioners. On any given show, the VITAC rate of accuracy is usually 99% or above. That average reflects a 1-error to every 100-word ratio.
As you can see, we work with a very unique and sophisticated technology that is itself almost perfected, however, the humans behind the technology, as hard as we try, are not perfect. VITAC is committed to providing quality, verbatim captioning to the caption-viewing audience. Our goal always is to give you as close to 100% perfection as is humanly possible and to ensure that your expectations are met.
Please know that your comments and constructive criticism are always welcome. If you have any questions, please do not hesitate to contact me.
Sincerely,
Laura Doty
Communications Coordinator
.
