Voice data creation

Epson has developed a natural-sounding voice data creation tool for embedded applications, as well as the S1V3034x series of voice chips and the terminal-compatible S1V3S344 IC chip with built-in flash memory.

The tool combines voice synthesis, provided by VoiceText Misaki, with features that facilitate voice guidance sentence analysis as well as pronunciation, phrase editing and ROM image creation on a single computer.

Although the tool has many functions and capabilities, such as semi-automated phrase segmentation for automatic integration of voice guidance data and automatic creation of voice data, the user interface is simple.

Each product in the series is operated by control commands from the host. Voice data transmitted from the built-in ROM or host is decoded and voice is sent through the built-in 16-bit D/A converter.

The capacity of the ROM can be selected from among a line-up of three models.

As the series is equipped with a wide variety of interfaces, including clock synchronous serial interface, I²C and UART, it is possible to build a system for use in combination with a variety of host devices or microcontrollers.

The voice codec function uses the company’s original voice compression format, facilitating the creation of high-quality voice guidance with a minimal amount of data.