Electrical audio signal processing systems and devices – Including amplitude or volume control – With amplitude compression/expansion
Reexamination Certificate
1998-08-25
2003-02-04
Isen, Forester W. (Department: 2644)
Electrical audio signal processing systems and devices
Including amplitude or volume control
With amplitude compression/expansion
C381S107000
Reexamination Certificate
active
06516068
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a microphone expander and, more particularly, to a microphone expander that effectively reduces background noise.
2. Description of Related Art
Microphone-expanders have been used to electronically alter the linear sensitivity of electret microphones so that they mimic the response of carbon microphones, which are less sensitive to low-level signals. This lack of sensitivity to low level signals has the incidental effect of reducing the transmission of background noise, which is typically at a lower level than speech signals. That is, in carbon microphones gain can be applied to the microphone signal to transmit the user's speech more clearly, while still keeping the noise level in the signal at an acceptable level.
FIG. 3
depicts a conventional microphone expander circuit that attempts to mimic that effect in conventional telephones. The signal from a telephone microphone
1
is introduced to an expander logic circuit
2
and an amplifier
3
. The expander logic circuit
2
generates a loss that is applied to the amplifier
3
to adjust the signal from the microphone
1
.
The expander logic circuit
2
makes a short-term measurement of the microphone signal and generates a loss signal according to the function shown in FIG.
4
. The loss function is characterized by an upper threshold T
U
, above which all loss is removed. In
FIG. 4
the upper threshold T
U
is 80 dB SPL, at which the loss becomes zero and the speech is passed unattenuated by the amplifier
3
. There is a lower threshold T
L
below which the maximum loss is applied. In
FIG. 4
the lower threshold T
L
is 71 dB SPL, at which the loss is 15 dB (a gain of −15 dB). In between the two thresholds, the attenuation represented by the loss varies according to the level of the microphone signal. In the example shown in
FIG. 4
, the loss varies linearly between the two thresholds.
Conventional microphone expanders such as that shown in
FIG. 3
can also be used to enhance the signal at the telephone receiver by using the expanded microphone signal output by the expander logic circuit. As shown in
FIG. 3
, a side tone path
4
introduces the expanded microphone signal to an amplifier
5
, which attenuates the signal by 12 dB. This attenuated signal is added to the signal at the telephone receiver
6
, thus resulting in a corresponding noise reduction in the received signal.
However, a difficulty with conventional microphone expanders in a noisy environment is that the difference between speech and noise cannot be defined simply in terms of signal level. As a result, using the microphone signal level to distinguish between the voice signal and unwanted noise is inherently unreliable. In other words, since the conventional approach simply attenuates signals depending on their level, noise at the same level as speech will be passed on as if it were speech.
U.S. Pat. No. 4,847,897 discloses another type of microphone expander, which attempts to measure noise utilizing signal properties that typically distinguish it from speech. The measured noise level is used to determine the amount by which to attenuate the microphone signal in the absence of a strong speech signal. This approach uses the noise level to determine the amount of loss to insert in the microphone signal and as an indication of the presence of speech. However, it does not address the problem of how to adjust the attenuation applied to the microphone signal across the normal range of speech and noise levels.
U.S. Pat. Nos. 3,889,059 and 3,963,868 describe speakerphones in which noise and voice levels are detected and used to alter a microphone signal. However, they too fail to address the problem of how to adjust the attenuation applied to a microphone signal across a range of speech and noise levels, particularly in a cellular telephone environment where noise levels can be especially severe.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a microphone expander that avoids the shortcomings of the prior art.
It is another object of the invention to provide an improved manner to determine how much loss to apply to a microphone signal by using loss functions that exhibit hysteresis, and, in another aspect of the invention, the thresholds defining the loss functions can be changed as the noise level changes so that the final signal is as optimal as possible.
In accordance with a first broad aspect of the invention, a microphone expander for attenuating background noise in a microphone signal comprises a signal averaging circuit for generating an averaged signal including speech components in the microphone signal, and a loss adjusting circuit for determining an amount of attenuation to be added to the microphone signal, wherein the loss adjusting circuit determines the attenuation from a first loss function if the averaged signal increases from below a lower noise threshold and from a second loss function if the averaged signal decreases from above an upper speech threshold, the first loss function providing more attenuation than the second loss function for a given averaged signal level.
In accordance with another broad aspect of the invention, such a microphone expander further comprises a noise level averaging circuit for generating a noise level signal representing background noise in the microphone signal and a threshold adjusting circuit for changing the first and second loss functions if the noise level signal exceeds a predetermined value.
In accordance with a more specific embodiment of the invention, a microphone expander for attenuating background noise in a digitized microphone signal from a wireless telephone comprises:
a signal averaging circuit for generating from an input signal an averaged signal including speech components in the microphone signal by sampling the digitized microphone signal in accordance with the following algorithm:
s
t
=
s
t
-
1
+
p
t
-
s
t
-
1
d1
if
p
t
≥
s
t
-
1
s
t
=
s
t
-
1
+
p
t
-
s
t
-
1
d2
if
p
t
<
s
t
-
1
where:
500/sec<sampling rate<1000/sec,
P
t
=the new sample,
s
t−1
=the old signal average,
s
t
=the new signal average, and
d
1
<d
2
;
a noise level averaging circuit for generating a noise level signal representing background noise in the microphone signal by sampling the averaged signal in accordance with the following algorithm:
n
t
=
n
t
-
1
+
s
t
-
n
t
-
1
d3
if
s
t
≥
n
t
-
1
n
t
=
n
t
-
1
+
s
t
-
n
t
-
1
d4
if
s
t
<
n
t
-
1
where:
500/sec<sampling rate<1000/sec,
s
t
=the new averaged signal sample,
n
t−1
=the old noise average,
n
t
=the new noise average, and
d
3
>>d
4
;
a threshold adjusting circuit for determining parameters range, transition and hysteresis, range being an upper speech threshold minus a quantity determined by adding a predetermined margin to the noise level, transition being the lesser of (i) range multiplied by a factor less than one, and (ii) a default value TRANSITION determined by multiplying a predetermined maximum value for range by the factor, and hysteresis being the lesser of (i) range minus transition, and (ii) a default value HYSTERESIS determined by subtracting TRANSITION from the predetermined maximum value for range;
a loss adjusting circuit for determining a parameter delta that is the lesser of (i) transition, and (ii) the averaged signal minus a quantity determined by subtracting transition from an upper speech threshold, if a first loss function provides the amount of attenuation, or the averaged signal minus a quantity determined by subtracting transition plus hysteresis from the upper speech threshold, if a second loss function provides the amount of attenuation, or zero if delta is negative, wherein the first loss function provides the attenuation if the averaged signal increases from below a lower noise threshold defined as the upper speech threshold minus t
Ciurpita Greg
Pennock Scott
Isen Forester W.
Lucent Technologies - Inc.
Pendleton Brian T.
LandOfFree
Microphone expander does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Microphone expander, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Microphone expander will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3161347