AbstractThe authors have been involved for 8 years on R&D as well as improvement of the production system for videotapes with digital captions for the deaf, using Japanese word processors. A prototype was developed in July 1982, and a pattern information synthesizer VIP (video information processor)‐2000 was developed in April 1983. Captions which are almost faithful to the speech of the performer were synthesized into the scene of the program, and the results were almost satisfactory. However, a certain time was needed to synthesize the caption with the video scene, which produced a delay between the caption and the mouth movement. Sometimes, the captions were difficult to read depending on the background, since no vertical edge was provided. To solve these problems, the VIP‐4000, which is a coded information synthesizer, was developed. As a result, the time for caption synthesis is reduced drastically and the vertical edge is provided. Furthermore, coloring of the characters and inhibition rules can be applied, which helps the visual sensation. This paper describes the hardware, software, features, and performances of the VIP