Automatic background noise suppressor

An automatic signal enhancement system for processing an audio input signal containing unwanted sustained components and desired transient components and providing a system output signal which is present only when said desired transient components are present in said input signal, said system comprising:

• first means responsive to application of first and second signals thereto for providing said system output signal only when both said first and second signals are present and said first signal exceeds a predetermined amplitude level;
• second means connected to receive said input signal for providing a control signal at an amplitude above said predetermined amplitude level only when said desired transient components are present in said input signal;
• means for connecting said control signal to said first means as said first signal;
• and
• means for connecting said second signal to said first means;
• wherein said second signal is said audio input signal;
• wherein said second means comprises:
• automatic gain control means responsive to said input signal for providing a further signal at a preselected amplitude when only said unwanted sustained component and not said desired transient components are present in said input signal, and for providing said further signal at an amplitude which exceeds said pre-selected amplitude and as a function of the amplitude of said desired transient components when both said unwanted sustained components and desired transient components are present in said input signal;
• envelope detector means connected to receive said further signal for providing an envelope-detected version of said further signal only when said further signal amplitude exceeds said pre-established amplitude;
• and
• filter means connected to receive said detected signal and to provide said control signal;
• wherein said automatic gain control means comprises:
• variable gain amplifier means connected to receive said input signal and to provide said further signal;
• storage capacitor means;
• a first charging circuit connected to receive said further signal and including a relatively low-resistance charge path connected to pass charging current from said further signal to said storage capacitor means;
• a second charging circuit connected to receive said further signal and including a relatively high-resistance charge path connected to pass charging current from said further signal to said storage capacitor means;
• a resistive discharge path;
• switching means responsive to charging current being passed through said second charging circuit for discharging charge above a predetermined charge level from said capacitor through said resistive discharge path;
• and
• means for varying the gain of said variable gain amplifier means as a function of the charge stored in said storage capacitor means.

An automatic signal enhancement system for processing an audio input signal containing unwanted sustained components and desired transient components and providing a system output signal which is present only when said desired transient components are present in said input signal, said system comprising:

• first means responsive to application of first and second signals thereto for providing said system output signal only when both said first and second signals are present and said first signal exceeds a predetermined amplitude level;
• second means connected to receive said input signal for providing a control signal at an amplitude above said predetermined amplitude level only when said desired transient components are present in said input signal;
• means for connecting said control signal to said first means as said first signal;
• and
• means for connecting said second signal to said first means;
• wherein said second signal is said audio input signal;
• wherein said desired transient components are speech components, or components having the temporal properties of speech, and wherein said unwanted sustained components are continuous background noise components of substantially constant amplitude in the audio frequency range, said system being further characterized in that said first means is a signal multiplier circuit which provides said system output signal at an amplitude which is the product of the amplitudes of said input and control signals;
• wherein said second means comprises:
• first envelope detector means for detecting the envelope of said audio input signal;
• band pass filter means for filtering the detected input signal envelope, said band pass filter means having a low frequency cut-off above 0 Hz and high frequency cut-off above 50 Hz;
• second envelope detector means connected to receive signal passed by said band pass filter means;
• and
• further filter means connected to receive detected signal from said second envelope detector means to provide said control signal.

An automatic signal enhancement system for processing an audio input signal containing desired transient speech-like components residing in a characteristic frequency range and unwanted sustained noise components residing both inside and outside said characteristic frequency range, and providing a system output signal which is present only when said desired transient speech-like components are present in said audio input signal, whether or not said unwanted sustained noise components are present, said system comprising:

• signal multiplier means responsive to first and second signals applied thereto for providing said system output signal at an amplitude which corresponds to the amplitude of said second signal multiplied by the amplitude of said first signal;
• control means responsive to said audio input signal for providing said first signal:
• (a) at zero amplitude when the desired transient speech-like components are absent from said audio input signal;
• and (b) at an amplitude which varies as a proportional function of the amplitude of said audio input signal only when said desired transient speech-like components are present in said audio input signal, irrespective of the presence of said unwanted sustained noise components in said audio input signal;
• means for applying said first signal from said control means to said signal multiplier means;
• and
• further means for applying said second signal to signal multiplier means.

An automatic gain control system responsive to an applied input signal for providing a gain controlled signal at a pre-selected amplitude when only unwanted sustained component and not desired transient components are present in said input signal, and for providing said gain controlled signal at an amplitude which exceeds said pre-selected amplitude and which varies over a continuous range as a function of the amplitude of said desired transient components when both said unwanted sustained components and desired transient components are present in said input signal, said automatic gain control system comprising:

• variable gain amplifier means connected to receive said input signal and to provide said gain controlled signal;
• storage capacitor means;
• a first charging circuit connected to receive said gain controlled signal and including a relatively low-resistance charge path connected to pass charging current from said gain controlled signal to said storage capacitor means;
• a second charging circuit connected to receive said gain controlled signal and including a relatively high-resistance charge path connected to pass charging current from said gain controlled signal to said storage capacitor means;
• a resistive discharge path;
• switching means responsive to charging current being passed through said second charging circuit for discharging charge above a predetermined charge level from said capacitor through said resistive discharge path;
• and
• means for varying the gain of said variable gain amplifier means as a function of the charge stored in said storage capacitor means.

An automatic signal enhancement system for processing an audio input signal containing desired transient speech-like components and unwanted sustained noise components, and providing a system output signal which is enhanced in amplitude only when said desired transient speech-like components are present in said input signal, irrespective of the presence and absence of said unwanted sustained noise components in said input signal, said system comprising:

• adjustable gain amplifier means, connected to receive said audio input signal, for providing a gain controlled signal;
• slow-attack, fast-release control circuit means responsive to said gain controlled signal for automatically adjusting the gain of said amplifier means to provide said gain controlled signal at a known constant amplitude in response to the presence of only said sustained noise components without said transient speech-like components in said gain controlled signal, regardless of the amplitude level of said sustained noise components in said audio input signal;
• biased detector means, connected to receive said gain controlled signal, for providing a further signal at:
• sustantially zero amplitude when the amplitude of said gain controlled signal is said known constant amplitude;
• and at an amplitude which is a function of the amplitude of said audio input signal when the amplitude of said gain controlled signal exceeds said known constant amplitude;
• filter means connected to receive said further signal and provide a processed signal as a predetermined function of the further signal;
• and
• signal multiplier means, connected to receive said processed signal and a second signal, for providing said output signal with an amplitude corresponding to the product of the amplitudes of said processed and second signals.

An automatic method of processing an audio input signal containing desired transient speech-like components residing in a characteristic frequency range and unwanted sustained noise components residing both inside and outside said characteristic frequency range, and providing a system output signal which is present only when said desired transient speech-like components are present in said audio input signal, whether or not said unwanted sustained noise components are present, said method comprising the steps of:

• multiplying first and second signals to provide said system output signal at an amplitude which corresponds to the amplitude of said second signal multiplied by the amplitude of said first signal;
• and
• in response to said audio input signal, providing said first signal:
• (a) at zero amplitude when the desired transient speech-like components are absent from said audio input signal;
• and (b) at an amplitude which varies as a proportional function of the amplitude of said audio input signal only when said desired transient speech-like components are present in said audio input signal, irrespective of the presence of said unwanted sustained noise components in said audio input signal.