Claims Analysis
Total Claims: 7
Claim 1
Independent Claim
181 words
A speech signal quality evaluator comprising: a logarithmic amplifier responsive to said speech signal including noise level signals occurring during pauses between speech signals for providing a signal proportional to the logarithm of said speech signals and said noise signals, an envelope detector responsive to said signal proportional to the logarithm of said speech signals and said noise signals for providing a detected signal proportional to the envelope of the logarithm of said speech signals and noise signals, a peak detector responsive to said detected signal over a first time period including normal speech for providing a signal proportional to the peaks of said detected signal, a valley detector responsive to said detected signal over a second time period including said speech pauses for providing a signal proportional to the intersyllable noise floor, and subtractor means responsive to the amplitude level of said peak detected signal and to the amplitude level of said valley detected signal for providing a signal quality signal equal to the arithmetic difference between the amplitude levels of said peak detected signal and said valley detected signal.
Claim 2
Dependent Claim
181 words
A speech signal quality evaluator comprising: a logarithmic amplifier responsive to said speech signal including noise level signals occurring during pauses between speech signals for providing a signal proportional to the logarithm of said speech signals and said noise signals, an envelope detector responsive to said signal proportional to the logarithm of said speech signals and said noise signals for providing a detected signal proportional to the envelope of the logarithm of said speech signals and noise signals, a peak detector responsive to said detected signal over a first time period including normal speech for providing a signal proportional to the peaks of said detected signal, a valley detector responsive to said detected signal over a second time period including said speech pauses for providing a signal proportional to the intersyllable noise floor, and subtractor means responsive to the amplitude level of said peak detected signal and to the amplitude level of said valley detected signal for providing a signal quality signal equal to the arithmetic difference between the amplitude levels of said peak detected signal and said valley detected signal.
Claim 3
Dependent Claim
21 words
The combination of claim 1 wherein said first time period is sufficient for both normal speech and speech pauses to occur.
Claim 4
Dependent Claim
14 words
The combination of claim 1 wherein said first time period is about one second.
Claim 5
Dependent Claim
14 words
The combination of claim 3 wherein said second time period is about four seconds.
Claim 6
Independent Claim
164 words
A system for selecting a speech signal from a plurality of speech signals of like information content where said speech signals include normal speech and pauses between said speech comprising: a separate signal quality detector responsive to each of said speech signals over a time period including normal speech and pauses between speech for providing a level signal equal to the logarithm of the ratio of the peak of the speech signal to the valley of the speech signal, a separate selection gate responsive to each of said speech signals with the outputs of said gates connected in common to a utilization circuit, and means connected to each of said quality detectors responsive to said level signals for providing a control signal to that one of said selection gates corresponding to the quality detector responsive to the highest of said peak to valley ratio signal to cause that selection gate to couple the speech signal to which it is responsive to said utilization circuit.
Claim 7
Independent Claim
164 words
A system for selecting a speech signal from a plurality of speech signals of like information content where said speech signals include normal speech and pauses between said speech comprising: a separate signal quality detector responsive to each of said speech signals over a time period including normal speech and pauses between speech for providing a level signal equal to the logarithm of the ratio of the peak of the speech signal to the valley of the speech signal, a separate selection gate responsive to each of said speech signals with the outputs of said gates connected in common to a utilization circuit, and means connected to each of said quality detectors responsive to said level signals for providing a control signal to that one of said selection gates corresponding to the quality detector responsive to the highest of said peak to valley ratio signal to cause that selection gate to couple the speech signal to which it is responsive to said utilization circuit.