• Encodes most of the speech content of audio as an ascii text

• Equivalent to a high accuracy speech recognition system

• Encoded data signal in the unblanked portion of Line 21, Field 1 of NTSC video signal

                Speaker changes
 
                Speech-Music changes

• In-house text processing - include points where the primary speaker or topic changed, locations in the video which discuss specific subject or contain information about a domain specific event