Electrical audio signal processing systems and devices – Noise or distortion suppression – Using signal channel and noise channel
Reexamination Certificate
2000-02-04
2003-04-15
Isen, Forester W. (Department: 2747)
Electrical audio signal processing systems and devices
Noise or distortion suppression
Using signal channel and noise channel
C381S092000, C381S106000, C381S107000
Reexamination Certificate
active
06549630
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to the field of communications with possible uses in voice recognition, public address, and recording. The present invention relates more particularly to a signal expander system that can distinguish sounds from acoustic sources placed at different locations relative to a transducer.
BACKGROUND OF THE INVENTION
Conventional voice communication systems typically incorporate some form of voice activated switch or signal expander for suppressing acoustic background noises.
FIG. 1A
illustrates a functional block diagram of a conventional bi-stable signal expander system
100
comprising a microphone
105
, an expander control stage
110
, and a variable gain block (stage)
115
. The microphone
105
is coupled to the variable gain block
115
, and the expander control stage
110
is coupled to the microphone
105
and the variable gain block
115
. The expander control stage
110
includes a detector
120
coupled between the microphone
105
and a first input
125
of a comparator
130
. The comparator
130
has a second input
135
coupled to a reference (threshold) voltage source
140
for generating a reference voltage level Vref.
When an acoustic source
145
becomes active, the emitted sounds from acoustic source
145
will cause changes in the air pressure. The microphone
105
detects the air pressure changes and translates the air pressure changes into corresponding voltage changes (i.e., microphone output signals
150
) that are detected by the detector
120
. The detector
120
outputs the microphone output signal
150
as a detector output signal
155
. The comparator
130
compares the voltage level of the detector output signal
155
with the reference voltage level Vref from reference voltage source
140
. If the voltage level of the detector output signal
155
is below the reference voltage level Vref, then the comparator
130
generates an output signal
160
with a logical state that does not activate the variable gain block
115
. As a result, the variable gain block
115
does not add gain to the microphone output signal
150
. When the acoustic source
145
activates, the voltage level of the microphone output signal
150
rises. The detector
120
detects the higher-level microphone output signal
150
and will, as a result, output a higher-level detector output signal
155
. If the voltage level of detector output signal
155
rises above the reference voltage level Vref, the comparator
130
outputs a control signal
160
with a logic state that causes the variable gain block
115
to add gain to the microphone output signal
150
. The variable gain block
115
then outputs the amplified microphone output signal as an audio output signal
165
.
A disadvantage of the bi-stable signal expander system
100
is that a low-level sound (e.g., soft speech) from the acoustic source
145
may not be amplified if the magnitude of the low level sound does not trigger the comparator
130
. If the threshold of comparator
130
is set too low, then noise signals in the environment will easily trigger the comparator
130
, thereby activating the variable gain block
115
and amplifying the noise signals (“noise pumping”). If the threshold of comparator
130
is set too high, then softer sounds from the acoustic source
145
may not trigger the comparator
130
and not activate the variable gain block
115
, resulting in inadequate gain for the desired signal.
In order to reduce the consequence of low level speech failing to activate the comparator
130
, some systems set the minimum gain of the variable gain block
115
to be only 12 dB to 20 dB below the maximum (fully activated) gain.
A disadvantage of the above conventional bi-stable signal expander systems is that ambient or undesired noises having magnitudes above the threshold level of comparator
130
are amplified. Additionally, if the bi-stable expander system
100
will be used in a noisy environment, the threshold of the comparator
130
must be set appropriately so that the external noises do not trigger the comparator
130
, increasing the severity of noise modulation and increasing the likelihood that the system will not respond properly to voice.
Additionally, if the microphone
105
is oriented away from the acoustic or speech source
145
, the microphone
105
may not be able to properly detect the desired sound waves. As a result, the microphone output voltage level
150
will be low and may not trigger the comparator
130
to permit amplification of the desired detected sound.
The above-mentioned bi-stable signal expander systems have a fast attack and slow decay characteristic that causes the switches for controlling gain to respond quickly to a detected sound of a sufficient voltage level and to maintain the gain for a pre-defined time length (e.g., 150 ms to 200 ms) after the voltage level of the detected sound falls below the comparator
130
threshold. By maintaining the gain for the additional pre-defined time length, the quieter-sounding, trailing ends of the speech envelope are not cut off by the bi-stable signal expander system. These trailing ends are typically below the comparator threshold. However, noise is often amplified during the additional pre-defined time length when gain is maintained.
The above-mentioned bi-stable signal expander systems also encounter problems when a burst of background noise occurs. For example, the noise burst might be a typewriter key impact or other types of noises with impulse waveforms. The noise burst will trigger the comparator
130
in the bi-stable signal expander system, thereby adding gain to the undesired noise burst. In addition, since the gain is maintained for the above-mentioned pre-defined time length after the noise burst occurrence, subsequent undesired noises are also amplified.
FIG. 1B
illustrates a conventional variable gain signal expander system
200
including an expander control stage
205
coupled between the microphone
105
and the variable gain block
115
. The expander control stage
205
includes a detector
210
coupled to an amplifier
215
. The microphone output signal
150
is detected by the detector
210
and amplified by the amplifier
215
. As a result, the amplifier
215
generates a control signal
220
with a magnitude that depends on the initial magnitude of the microphone output signal
150
. The amount of gain provided by the variable gain block
115
to the microphone output signal
150
depends on the magnitude of the control signal
220
.
The variable gain signal expander system
200
can be designed with a shorter time constant for reduced audibility of “noise pumping”. The shorter time constant reduces the amount of time that high gain is applied to the noise signal as the desired signal drops in amplitude. The effect of the combination of variable gain with reduced time constants on the desired signal is to modulate the envelope of the speech signal. This is not generally desirable but may be an acceptable compromise in noisy environments.
A further disadvantage of the variable gain signal expander system
200
is that both the signal from the desired acoustic source and the ambient acoustic noise are detected and used to increase the gain of the variable gain block
115
. Thus as the ambient noise levels increase, the gain of the system for this noise can also increase, resulting in less overall noise reduction.
Accordingly, it is desirable to provide a method and system for signal expansion with improved noise rejection capability.
SUMMARY OF THE INVENTION
The present invention provides a desirable method and system for discriminating between desired sounds from a near-field acoustic source and sounds (noise) from far-field acoustic sources. The invention advantageously prevents gain from being added to the undesired (far-field) loud noises while allowing gain to be added to low-level sounds from the desired (nearby) acoustic source. As a result, the invention can reduce the “noise pumping” problem that is encountered by conventional systems an
Hsieh Peter
Isen Forester W.
Pendleton Brian Tyrone
Plantronics Inc.
LandOfFree
Signal expander with discrimination between close and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Signal expander with discrimination between close and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Signal expander with discrimination between close and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3021504