Philips MEA8000

Overview

The Philips MEA8000 builds speech from a stream of compact digital speech frames. Each frame sets values such as pitch movement, amplitude, duration, and formant frequencies and bandwidths. Inside the chip, an eighth-order digital filter shapes the sound. It synthesizes the waveform from these parameters instead of playing back recorded words.

Speech arrives one frame at a time

A controller sends the MEA8000 a starting pitch followed by 32-bit speech frames. Each frame describes a short slice of sound. The chip interpolates the changing values as it produces that slice, then requests more data. A sequence of frames can form continuous speech.

A digital vocal tract

The MEA8000 uses an eighth-order digital filter with three programmable formant frequencies, a fourth fixed formant, and programmable bandwidths. The frames also control amplitude, pitch behavior, and whether the sound is voiced or unvoiced. This makes it a digital formant synthesizer, with a different input method from chips that select a stored phoneme from a code.

A different kind of speech chip

The SC-01 and SC-02 accept phoneme codes and handle sound-to-sound transitions internally. The MEA8000 instead expects its controller to supply the speech parameters frame by frame. That gives the host system direct control over the sound, while requiring it to provide the frames that make up the utterance.

MEA8000 at a glance

Continue exploring

• Philips PCF8200

Further references

• Philips/Signetics: MEA8000 Voice Synthesizer data sheet, Philips Radio, Audio and Associated Systems handbook
• Philips: MEA8000 speech synthesizer, Electronic Components and Applications, Vol. 4, No. 2 (1982)
• U.S. Philips patent 4,783,810: device for generating audio information from character sets