LA CROSSE, WIS.—Comprompter announced the release of their newest voice recognition product, Caption Central v2 for live closed captioning. CC2 was specifically redesigned to provide the most accurate and affordable live captioning solution for local broadcasters.
Comprompter CEO, Ralph King said, ”When we brought out our original Voice Recognition system in 2012 the best we could do for accuracy was about 90 to 92 percent. But in the last 6 months we have taken our voice captioning well into the high 95 to 98 percent accuracy range. Caption Central provides a single AD-LIB button click to switch from scripted to non-scripted captioning on the fly.”
Caption Central performs an instant context check to select the most appropriate word when a conflict arises over which spelling of that word occurs. “ Caption Central even does Spell-Checking and capitalizes proper names of ‘people, places and things’ it recognizes.” King said. “While our original intention was to produce Caption Central as an add-on to our own NewsKing Newsroom system, it actually works just as well with any other newsroom system and/or caption encoder box we are aware of.”
While providing for a ‘Shadow Speaker’ like most other systems are limited to, Caption Central stresses live studio input directly from the on-set anchors. Some of the features include: Built-in Voice Training; and 2 forms of learning words simply by reading them … either from the news Script or from a List of Words.
“Caption Central can read a half-hour newscast and automatically learn any unknown words and add them to the dictionary in less than a minute automatically. I have a list of 121 meteorological terms that the weather staff used to have to type in and pronounce. Now CC2 reads and learns all 121 words in 6 seconds,” said King, “ And we fully anticipate that Caption Central will meet and exceed the current and upcoming FFC rules.”
