The Computer Speaks

Present techniques to output computer speech are offshoots of analog and compressed digital recording. Speech is reproduced from prerecorded text on an analog audio disk or it is produced by using read-only memory integrated circuits. The focus is now turning to speech synthesis from text. One way t...

Full description

Saved in:
Bibliographic Details
Published in:IEEE spectrum 1979-08, Vol.16 (8), p.18
Main Author: Sherwood, Bruce A
Format: Magazinearticle
Language:English
Subjects:
Computers
Conversion
Digital
Magnetic tape
Speaking
Speech
Synthetic
Techniques
Technology
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Present techniques to output computer speech are offshoots of analog and compressed digital recording. Speech is reproduced from prerecorded text on an analog audio disk or it is produced by using read-only memory integrated circuits. The focus is now turning to speech synthesis from text. One way to provide speech output is to play analog recorded messages under computer control by using magnetic tape with recorded audio on one track and digital addressing and timing information on another track. Slow serial access along the tape to the desired message is a major disadvantage of this technique. Human speech can be digitized to about 100 kb/s by an analog-to-digital converter and ouput into a digital storage medium. Also, it is possible to synthesize speech from stored or generated character strings. Many algorithms are available that can be used to drive a phonemic speech synthesizer to produce high-quality speech in several languages from stored or computer-generated text.
ISSN:0018-9235
1939-9340
DOI:10.1109/mspec.1979.6368188