STUDY OF SPEECH COMPRESSION SYSTEMS (SYLLABIC COMPRESSION TECHNIQUES)

Abstract

Research concerned a syllable communication system, voice to voice or voice to print, utilizing the phonetic typewriter developed by RCA and a speech synthesizer which proceeds from pre-recorded spoken syllables. Such a system when ultimately developed should allow transmission of the spoken word at an extremely low rate such as 23 bits/sec calculated for a language of 2000 syllables and a normal speaking rate. Apparatus assembled analyzes and synthesizes 50 syllables or words and has the resolving power and memory capacity to print-out an even larger number of words selected for this purpose when set up for a given speaker. For 50 words the information for printout or speech synthesis is transmitted at a channel capacity of 6 bits/word. Tests were made using list of words other than those originally suitable for machine processing. One series of tests was made with a phonetically balanced list of words as used for standardized articulation testing. Other tests were made with the words representing the current phonetic spelling alphabet and with other words chosen for their military significance. Apparatus was assembled to investigate the usefulness of processing envelope information. Studies were made of the rates of growth and decay and the duration of intra-syllable pauses.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 15, 1962
Accession Number
AD0284811

Entities

People

  • E. S. Rogers
  • Herbert Belar
  • R. Desobrino

Organizations

  • Sarnoff Corporation

Tags

Communities of Interest

  • Advanced Electronics
  • Human Systems

DTIC Thesaurus Topics

  • Accuracy
  • Acoustics
  • Air Force
  • Air Force Facilities
  • Analyzers
  • Automated Speech Recognition
  • Communication Equipment
  • Communication Systems
  • Detection
  • Detectors
  • Electronics
  • Frequency
  • Frequency Bands
  • Measurement
  • National Security
  • New Jersey
  • Numbers

Readers

  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Speech Processing/Speech Recognition.

Technology Areas

  • AI & ML