Method and apparatus for time domain compression and synthesis of audible signals

• US 4,433,434 A
• Filed: 12/28/1981
• Issued: 02/21/1984
• Est. Priority Date: 12/28/1981
• Status: Expired due to Term

• Current Assignee: Electronic Speech Systems, ESS Technology Incorporated (Semiconductor Holding Corp.)
• Original Assignee: Electronic Speech Systems
• Inventors: Mozer, Forrest S.
• Primary Examiner(s): Kemeny, E. S. Matt
• Application Number:
  US06/335,312
• Time in Patent Office: 785 Days
• Field of Search: 381/30, 381/29, 381/31-40, 381/51-53, 328/13, 328/14, 328/16, 328/109, 364/717, 375/122, 84/1.01, 84/1.24, 331/78
• US Class Current: 704/211
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
• CPC Class Codes: G10L 19/02 using spectral analysis, e.g. transform vocoders or subband vocoders

Abstract

Compression and synthesis techniques and related apparatus for time domain signals, particularly signals whose information content resides in the power spectrum such as speech. Compression techniques include adjusting the phase of harmonic components of a signal unit to obtain an equivalent power spectrum signal of a minimum number of discrete levels. The invention finds application in speech compression and compact speech synthesis devices.

Claims

1. A method for compressing a time domain information signal, said method comprising the steps of:

receiving said information signal; and

adjusting the phase of harmonic components of said received signal to produce an equivalent signal, said equivalent signal having sampled amplitude values at selected sample times, said amplitude values being limited to a selected maximum number of amplitude levels less than the number of amplitude levels utilized to define said information signal at said selected sample times, said equivalent signal having a power spectrum substantially the same as said information signal.

2. The method according to claim 1 wherein the number of permissible peak non-zero amplitude values is no more than two magnitude levels.

3. The method according to claim 1 or 2 wherein the permissible peak non-zero amplitude values are symmetric with respect to a zero reference level.

4. An apparatus for compressing a time domain information signal comprising:

means operative to receive said information signal; and

means coupled to said receiving means for adjusting the phase of harmonic components of said received information signal to produce an equivalent signal having a power spectrum substantially the same as said information signal, said adjusting means further producing said equivalent signal as a serial sequence of sampled amplitude values at selected sample times which is limited to a selected maximum number of amplitude levels less than the number of amplitude levels utilized to define said information signal at said selected sample times.

5. The apparatus according to claim 4 further including means limiting the number of permissible non-zero amplitudes values at selected sample times to no more than two magnitude levels.

6. The apparatus according to claim 4 or 5 further including means limiting permissible non-zero amplitude values at selected sample times to values which are symmetric with respect to a zero reference level.

7. A method for compressing a time domain information signal whose information content resides mainly in its power spectrum comprising the steps of:

digitizing a finite segment of said time domain signal;

analyzing said digitized waveform to determine amplitude and phase parameters in terms of harmonically related transcendental functions; and

altering the magnitude and sign of selected ones of said phase parameters without modifying said amplitude parameters to obtain an equivalent time domain signal whose amplitude in the time domain may be reconstructed by a selected limited maximum number of finite amplitude values less than the number of amplitude values required to digitize said information signal.

8. The method according to claim 7 wherein said altering step comprises Fourier transforming said time domain information signal into the frequency domain to determine frequency and phase components of said information signal.

9. An apparatus for synthesizing from compressed information an output signal which is substantially equivalent to a source time domain signal whose information content resides mainly in its power spectrum, said apparatus comprising:

memory means for storing digital representations of the amplitude of segments of a compressed time domain signal and for storing instructions correlating said segments to said output signal; and

means responsive to said digital representations and said instruction signals for constructing said output signal from said segments, said segments having a limited maximum number of finite amplitude values at selected sample times and said output signal having a power spectrum substantially equivalent to but having phase components differing from said source signal.

10. The apparatus according to claim 9 further including means for limiting the number of non-zero amplitude values at selected sample times to no more than two magnitude levels.

11. The apparatus according to claim 9 or 10 further including means limiting permissible non-zero amplitude values which are symmetric with respect to a zero reference level.

12. A method for synthesizing from compressed information an output signal which is substantially equivalent to a source time domain signal whose information content resides mainly in its power spectrum, said method comprising:

storing digital representations of the amplitude of segments of a compressed time domain signal with representations of instruction signals correlating said segments to said output signal; and

constructing said output signal from said segments in response to said instruction signals, said segments having a limited maximum number of finite amplitude values at selected sample times and said output signal having a power spectrum substantially equivalent to but having phase components differing from said source signal.