MCS is a major, nationally recognized provider of realtime closed captioning. The realtime captioning process is without question the most demanding of all types of captioning.      

     By now, most readers of our web page understand what closed captioning is. What many viewers may not understand are the differences in terminology used in the captioning industry, which define differences in the way captions are created.      

Realtime Captioning
     Realtime closed captioning is done by a person trained in stenography, with additional, specialized training in captioning. Using a court-reporting machine modified for realtime captioning, the individual, listening to the audio track of a program, depresses a combination of key strokes corresponding to a phonetic/steno dictionary entry, which is translated into data bits and sent by modem to an encoder at a TV or cable programmer's facility.

     The data is then encoded into the video signal, uplinked to a satellite in geosynchronous orbit, and transmitted to the viewer watching a television set with a captioning decoder. The data may travel as far as 59,000 miles, and, two seconds after the captions are created, they are visible to the viewer.      

     Realtime captioning is extremely difficult, and the accuracy of closed captioning is a function of many variables, such as:    

1. The level of experience of the realtime captioner, including familiarity with the subject matter.
   
   
2. Availability of the proper spelling of names and places which may be used during a broadcast. Often times, despite the best preparation, new names or places, or unusual pronunciations by anchors, may affect the way in which the stenowriter/captioner writes the word.
   
   
3. Anchor speed and clarity of speech. Accuracy is affected by the speaker's rate of speech, the number of people speaking at any one time (particularly relevant in talk shows), and the clarity of their diction.
   
   
4. Clarity of audio signal. A captioner's ability to accurately caption is affected by the clarity of the underlying audio signal. Signal fade, line noise, and poor reception all affect a captioner's ability to hear the audio, particularly if coupled with difficult or unfamiliar accents. Live feeds from remote locations may include each of these factors.

Live Display
     Live Display captioning involves the use of prescripted ASCII files sent to an encoder to produce captions, which are synchronized with the underlying audio track.

     Files that are used in live display are ASCII files created from prepared scripts or anchor copy filed in a newsroom computer. Live display is most commonly used in preproduced programming, where a script file can be created or is available prior to the airing of the program. As the program airs, the caption data is displayed live, in realtime, appearing to viewers as one line of captioned text corresponding to the underlying audio track.

     At Media Captioning Services, we have perfected the technique of using prescripted news stories, when available, to supplement the realtime process. We have highly skilled editors, called caption coordinators, who display such text files of news stories when they are available, and coordinate the sending of such data to station encoders.

     In addition, our coordinators determine whether to transmit such preprepared files or use data created by the realtime writer during each caption session. The use of live display files in a realtime session requires considerable timing and skill by our coordinators, but is a technique which can enhance the overall accuracy of a captioned broadcast.

Off-line Captioning
     This term refers to the process of creating captions prior to the scheduled airing of a program, where captioned data is inserted on a frame-accurate basis, using software which inserts the time-coded caption into different locations on the video screen.

     Captions appear to "pop" onto different screen locations, using one-, two- or three-line formats. This technique or style of captioning is used most often to caption movies and commercials. Since the script of such a production is available to the captioner, there is no immediate time-sensitivity affecting accuracy, except for production deadline pressures.      

     Stenographic techniques are not used to create the caption file. Rather, an individual using a keyboard and software designed to create these time-coded captions are utilized in this process. Typically, the off-line captioning process requires 10 to 14 hours of labor for each hour of off-line captioning.      

The MCS Commitment
     Clearly, the realtime captioning of news and sporting events, which MCS specializes in, is the most demanding form of captioning. Our writers caption consistently at 99.2% to 99.4% accuracy. During an hour newscast, we write, on average, approximately 6,000 words. At 99.5% accuracy, that would leave 30 words to appear as "untranslates," or misspellings, to the viewer.

     The realtime process is demanding, and not error-free. We do strive to maximize accuracy, to produce captions that are verbatim to the greatest extent possible, and at the minimum contextually correct.