PROCESSING OF COMMUNICATIONS SIGNALS TO REDUCE EFFECTS OF NOISE

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-00120R000100030010-3
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 6
• Document Creation Date: December 20, 2016
• Document Release Date: February 22, 2001
• Sequence Number: 10
• Publication Date: September 24, 1968
• Content Type: CONT

Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
Sept. 24, 1968 M. R. SCHROEDER 3,403,224
PROCESSING OF COMMUNICATIONS SIGNALS TO REDUCE EFFECTS OF NOISE
Filed May 28, 1965 2 Sheets-Sheet 1
N N
1111111 I
I ~' IIII'Ilt I O
/4
I I ~i
I I
I I I
i I
?~'i tv
0.~
I
IIIIIII
lilllil 4 Q
1
Oi N O
INVENTOR
Al R. SCHROEDER
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
 Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
Sept. 24, 1968 M. R. SCHR?EDER 3,403,224
PROCESSING OF COMMUNICATIONS SIGNALS TO REDUCE EFFECTS OF NOISE
Filed May 28, 1965 2 Sheets-Sheet 2
NOISE LEVEL
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
 Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
United States Patent Office
3,403,224
PROCESSING OF COMMUNICATIONS SIGNALS TO
REDUCE EFFECTS OF NOISE
Manfred R. Schroeder, Gillette, NJ., assignor to Bell
Telephone Laboratories, Incorporated, New York, N.Y.,
a corporation of New York
Filed May 28, 1965, Ser. No. 459,655
11 Claims. (Cl. 179-1)
ABSTRACT OF THE DISCLOSURE
To improve quality of a communications signal, noise
components which accompany the signal are removed by
analyzing a signal to obtain a measure of the energy in
each of a number of frequency sub-bands, by developing
a signal representative of the average contribution of
noise to the signal in each sub-band, and by selectively
subtracting the noise representative signals from the sub-
band signals.
This invention relates to the processing of communica-
tions signals impaired by noise components and, more
particularly, to the method of and apparatus for elimi-
nating audible noise from a communications signal. It
has for its principal object, an improvement in the quality
of the signal through the elimination of the noise con-
tribution to the signal.
A variety of different techniques have been employed
in the past to cleanse a communications signal, e.g., a
speech signal, of accompanying noise components. Thus,
noise may be "stripped" from the envelope of a speech
signal by clipping, squelching, gating, or by a combina-
tion of these operations. Such techniques are quite effec-
tive for impulse noise and the like which "rides" the sig-
nal envelope, but are not altogether successful in elimi-
nating other continuous noise such as quantizing noise,
thermal noise, or the like. A somewhat more satisfactory
approach to the problem of reducing the impairment of
a signal by noise, involves the suppression of spurious
frequency components in the interharmonic regions of
the "fine" structure of the spectrum of the signal, rela-
tive to harmonic components. Such a system is described
in M. R. Schroeder, Patent 3,180,936, granted Apr. 27,
1965. This system recognizes that continuous noise is
distributed throughout the spectrum of a signal and is
relatively immune to ordinary processing.
The present invention approaches the problem from a
somewhat different though related avenue, and relies, at
least in part, on the ability of the ear to mask low energy
signals which occur with or near a large energy signal
such as a formant, but to respond to the same low energy
signals when they occur at or near relatively low energy
signal components, such as those between formants. Ac-
cordingly, the average contribution of noise to each of a
number of channel spectrum signals, developed for con-
tiguous frequency sub-bands of a communications signal,
is, in accordance with the invention, subtracted from each
of the channel signals. During periods of high energy in
each channel, the subtraction of the average value has
little effect on the quality or intelligibility of the speech.
During periods of low energy in any one of the sub-band
signals, when the noise contribution is most easily per-
ceived, the subtraction effectively removes the noise con-
tribution so that a cleaner signal is passed on to the out-
put of the channel.
In the event that the contribution of noise is predict-
able, for example, is due to certain kinds of random fad-
ing or the like, the appropriate average contribution to
be subtracted from the signal is easily determined. How-
ever, in most situations the noise contribution is unpre-
dictable and thus must be continuously determined for
3,403,224
Patented Sept. 24, 1968
2
each channel of the system. It has been found through
experience that this determination may be made satis-
factorily by examining the variations of speech energy
in each sub-band. In particular, the contribution of noise
to the level of each of a number of channel signals which
define the voice energy level in each sub-band may be
estimated from the relative minima that the channel
signals reach during time intervals that comprise at least
several speech sounds or during a silent interval. It is
assumed that noise, even though of an unpredictable
nature, is reasonably stationary over several speech
sounds, or generally, for one or two seconds. Accord-
ingly, a signal proportional to the average contribution of
noise to each channel signal is developed during periods
of low speech signal energy, preferably during silent
speech intervals, and is averaged over an interval of ap-
proximately one or two seconds. The averaged noise sig-
nals are then subtracted individually or in selected com-
binations from the channel signals. It is not always possi-
ble to estimate the noise contribution to a particular
channel accurately, i.e., during intervals of high speech
energy. This is primarily because of the finite averaging
time in the spectrum analyzer, i.e., the development of
the channel signals. The estimate will at times, there-
fore, be too high and at other times will be too low.
Some spectral distortion may thus result. However, on
the average, the distorting is slight and may not occur in
all channels at once. Experience has shown, moreover,
that the slight distortion which occasionally is produced
is exchanged for an appreciable improvement in speech
quality; the noise level is considerably reduced with
virtually no deterioration of intelligibility. Subjectively,
the exchange is most desirable.
The principles of the invention may be turned to ac-
count in a number of ways. Essentially, a signal is divid-
ed, on the frequency scale, into a plurality of sub-hands,
and the speech energy in each sub-band, or channel, is
measured. A number of frequency domain systems, gen-
erally referred to as a vocoder analyzers, or the like,
10 may be used for this purpose. The contribution of noise
to each channel may similarly be established in a number
of ways. If the channel signal is characterized by periodic
silent intervals, ordinary noise measuring techniques may
be used. If the spectrum of the noise is known, only its
overall intensity need be measured. For unpredictable
noise in relatively continuous speech energy signals, an
average value is estimated on the basis of minimum chan-
nel signal excursions, together with a maximum expected
noise signal level.
After subtraction, the modified channel signals are used
to reconstitute the applied signal, for example, using
conventional vocoder synthesizer apparatus. A pitch de-
fining signal, derived from the input signal, is generally
employed to control the synthesis. As required, the syn-
thesis operation may take place at any physical location.
Thus, the synthesizer may be located at a distant recep-
tion location.
It is apparent that the operation of a vocoder com-
munications system employed, for example, for narrow
band signaling between distant locations, may be ap-
preciably enhanced through the use of the present inven-
tion. The contribution of noise to each control channel
of such a system may be individually reduced. This is
particularly desirable since the control channels of such
a system contribute to the output noise of the system.
Extremely low-noise pitch analyzer circuits are known,
for example, the one described in a copending application
of A. M. Noll and M. R. Schroeder, Ser. No. 420.362,
filed Dec. 22, 1964, but a low-noise pitch circuit is of
limited value in lowering the noise level of the synthesized
speed signal if the channel control signals are themselves
noisy. By thus employing a low-noise pitch circuit, and
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
 Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
3,403,224
reducing the noise of the individual energy control signals, noise fluctuations are apparently still present in the channel
the overall noise of the system may be reduced to a very signal.
low level. The average noise contribution for each channel is
The invention will be more fully understood from the preferably evaluated, in accordance with the invention,
following detailed description of illustrative embodiments s from the relative minima that the different channel signals
thereof taken in connection with the appended drawings, exhibit during the course of several speech sounds. The
in which: evaluation is subject to the restriction that only the
FIG. I is a schematic block diagram of a speech trans- minima below an expected maximum noise level are ex-
mission system which embodies the present invention; and amined. It is recognized that the actual noise contribution
FIG. 2 is a set of curves which are referred to in the 10 will be under-estimated, since minima are used for the
explanation of the invention which follows. estimate. Hence, in order to get a more accurate estimate,
Communications signals, for example speech signals, the running estimate is increased by a factor, greater than
are supplied to the speech transmission system of FIG. one, which represents the ratio r of the average Poise
I by way of speech circuit 10. As indicated, these signals contribution and a typical minimum noise contribution.
may originate in a telephone instrument or microphone 1,; This factor may typically range from one to three; a factor
11. The signals are delivered to spectrum analyzer ap- r of 1.5 has been found to he satisfactory.
paratus by which the frequency spectrum of the signal FIG. 2D illustrates the method. An expected maximum
is subdivided into contiguous segments. Any analyzer that noise level is indicated; it is established on the basis of
produces speech control signals representing the short experience with the system. Whenever the channel signal
time spectrum or a similar description of a speech signal. 2-'0 Ct(t), falls below the running estimated noise level C?5t,
may be used. One suitable one, conventionally used in so- the value of C,,,t, is changed in accordance with the extent
called channel vocoder systems, is shown in the figure. of the channel signal excursion below the running level.
Signals from circuit 10 are supplied in parallel, to each C,,,t. follows C1(t) toward each minimum; it tends to-
one of a bank of bandpass filters 12, each of which is ward the maximum expected level between signal minima,
proportioned to pass a preassigned, sub-band of the voice 25 but is carried back to each minimum encountered. It is
frequency band of interest, and which together pass the never allowed to exceed the maximum expected noise
entire hand. For sake of illustration, ten such filters, each level.
with a 300 cycles per second pass band are indicated, the In practice, these operations are carried out in each
first two and the last one only being shown. Each filter 12 channel by noise signal evaluator 15. Control signals is-
is followed by a detector 13 and, in turn, by a low pass 30 suing from filter 14, whose amplitudes are less than the
filter 14. The control currentoutput of each of the several maximum expected noise level of the system, are passed
low pass filters 14 is thus a measure of the voice energy by way of a gate, such as biased diode 16 and resistor 17.
in that sub-band to which such low pass filter is connected. to adjustable gain amplifier 18. Diode 16 is biased, in
Since each of bandpass filters 12 passes one segment of the absence of a channel signal, to the maximum ex-
the band only, for example, a 0-300 cycles per second 35 pected noise level by a positive potential supplied from
segment of a 0-3000 cycles per second band, it is ap- source 19 by way of adjustable resistor 20. Adjustable ca-
parent that each of the energy control signals will gen- pacitor 21, shunting the input of amplifier 18, evidently is
erally differ from all others, in dependence on the char- normally (absence of channel signal) charged to the bias
acter of the input signal. Since frequency averaging is level, i.e., the maximum expected noise level. The time
helpful in eliminating noise, it is desirable to employ 4u constant of the system should be on the order of several
relatively wide band channel filters. Concomitantly, time seconds with resistor 17 being selected to have a rela-
averaging is also an aid in reducing the effects of noise. tively low resistance, so that capacitor 21 may be dis-
Accordingly, it is desirable to employ relatively narrow charged quickly when diode 16 conducts.
hand control signal filters. Filters 14 thus typically pass So long as the control signal Cu(t) exceeds the run-
a band of frequencies in the range of 0-10 cycles per ,t.) ning estimated noise level, diode 16 is held in its cut-off
second. condition and, with suitable adjustment of the operating
FIG. 2A illustrates an ideal energy control signal devoid level of amplifier 18, no net signal is developed at the
of noise. FIG. 2B illustrates a typical noise signal which output of evaluator 15. In this condition, control signals
may accompany a channel signal. Although the noise is from filter 14 are unaffected as they pass through sub-
relatively unpredictable, its average value is essentially 50 tractor 22 and rectifier 23.
constant, or is assumed to be constant at least over inter- If the control signal level falls below the running esti-
vals of several seconds duration. The effect of such a mate of noise level, the cathode of diode 16 becomes
noise signal on the channel signal is illustrated in the negative as compared with its anode, the diode conducts,
diagram of FIG. 2C. A signal of this sort is, during noisy and the anode voltage tends to follow the cathode volt-
signal conditions, delivered at the output of each one of 5:3 age. Capacitor 21 thereupon discharges through the di-
filters 14. It is apparent that the noise-free channel signal ode to the actual signal level and follows it so long as
of FIG. 2A has been increased throughout by the ampli- it continues to decrease. As soon as the signal increases,
tude of the noise signal contribution. The noise fluctua- the capacitor begins to charge, but at a rate determined
tions themselves are effective, however, to -alter the over- by the time constant of the circuit. It continues to do so
all character. of the control signal, primarily during inter- 60 until either the control signal once again drops below the
vals of low control signal energy. Thisnonuniform con- capacitor potential or until the capacitor is once again
trihution results from the masking faculty discussed above. charged essentially to the maximum expected noise level.
Designating the channel i control signal as Cs(t), and FIG. 2D illustrates the manner in which the running
the noise contribution to channel i as Cun(t), the resulting estimated noise level, namely, the charge Ce,t. on capaci-
composite signal (FIG. 2C) may be defined as: 65 tor 21, follows the negative-going control signals (those
Ct(t)=C>H(t) } s?(t) (1) which occur below the maximum expected noise level),
and eventually charge toward the noise level. It is appar-
During silent intervals, or intervals during which the ent that the charge on the capacitor tends to follow, on
speech energy level in a channel is low, a reasonably good the average, the control signal minima. By suitably select-
evaluation of the noise component may be obtained, 7tt ing the time constant of network 15, a relatively constant
i.e., its average amplitude may be measured. In accord- potential Cavs, may be produced,
once with the invention. such an average value is sub- The evaluation of noise contribution to the channel is
traded from the channel signal so long as the channel sig- completed by suitably increasing the level of the running
nal is greater than the average noise level. Since only estimate Ceet. by the factor r. In practice, the signal level
the average noise contribution is subtracted, the individual 7 is increased by a factor of r=1.5 in amplifier 18. The re-
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
 Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
5
3,403,224
suiting nearly constant voltage closely approximates the
average noise contribution to the control signal due to
embedded noise.
In accordance with the invention, this signal is supplied
to the negative input of subtractor 22 in which unit it is
subtracted from the composite control signal applied to
the input terminal of the subtractor from filter 14. The
resulting modified signal may be designated C1(t), where
6
the invention. Numerous other arrangements mya be de-
vised by those skilled in the art without departing from
the spirit and scope of the invention.
What is claimed is:
1. Signal processing apparatus which comprises, means
for analyzing a message wave, means for deriving from
said analysis a group of control signals each one of which
is representative of the energy within a selected frequency
band of said wave, means for evaluating the contribution
of noise to each of said frequency bands, means respon-
sive to said evaluations for developing for each of said
bands a signal representative of the contribution of noise,
and means for selectively subtracting said noise repre-
sentative signals from said control signals.
2. Signal processing apparatus which comprises, means
for analyzing a message wave, means for deriving from
said analysis a group of control signals each one of which
is representative of the energy within a selected frequency
band of said wave, means for evaluating the contribution
of noise to each of said frequency hands, said evaluating
means including means for examining amplitude varia-
tions of said control signals, means responsive to said
examinations for developing for each of said bands a
slowly varying signal representative of amplitude varia-
tions during intervals of low energy, and means for sub-
tracting said slowly varying signals from said control
signals.
3. Apparatus for suppressing noise in a message signal
which comprises, a source of a message signal, means for
deriving from said message signal a plurality of control
signals representative of the short time spectrum of said
message signal, means for developing a signal representa-
tive of the contribution of noise to said control signals,
and means for selectively subtracting said noise contribu-
tion signal from said control signals.
4. Apparatus for suppressing noise in a communica-
tions signal which comprises, a source of a communica-
tions signal, means for deriving from said signal a plu-
rality of spectrum control signals representative of the
energy failing within a plurality of frequency sub-hands
of said communications signal, means for developing a
group of average level signals representative of the con-
tribution of noise to each of said control signals, and
means for subtracting said average level signals from said
control signals.
5. Apparatus for processing speech waves to remove
relatively continuous noise which comprises, in combina-
tion with a source of a speech wave, means for analyzing
Cu(t).-C)(t)-Cest.(t)?r (2) to
it constitutes a speech energy signal virtually undegraded
by noise.
If the noise spectrum changes relatively slowly with
frequency, noise estimates for adjacent channels are sim-
ilar. In such cases, it has been found advantageous to 15
average selectively estimates from adjacent channels. It is
therefore in accordance with the invention to utilize the
average noise levels estimated individually for all chan-
nels together to form, when desired, an averaged signal
representative of the noise level in all of the channels. 20
Such a signal is subtracted from each channel signal after
suitable weighting for different channels, as required. The
required average may be formed by applying the individ-
ual average noise signals developed by evaluators 15 to
an averaging network 25, for example, a resistive matrix. 25
Several such networks may be employed; only one is
shown. It is assumed that averaging network 25 contains
effective means, such as a plurality of channel amplifiers,
for isolating the evaluators. The averaged signal devel-
oped by network 25 is made available to one pole of 30
switch 24 connected between evaluator 15 and subtractor
22. Ordinarily, switch 24 is thrown to the pole connected
to evaluator 15. However, in dependence on the signal
being processed, one, two or all of the switches may be
positioned to select the averaged signal for delivery to 35
subtractor 22,
In the event that the average noise signal developed
by evaluator 15 is greater than the momentary composite
channel signal, as a result of the averaging process, for
example, it is necessary to limit the channel signal output 40
to zero since channel control signals may never be nega-
tive. Any form of decision network may be used for this
purpose. One extremely simple one employs a rectifier 23
connected in the output of subtractor 22 biased to pass
only positive signals. Thus, if the modified channel signal 45
falls to zero or momentarily falls below zero, the out-
put of the channel is zero, i.e., no signal is passed by
rectifier 23.
signals, means for evaluating the contribution of noise
of the system are then used to develop, in a speech syn- ,5o to each of said spectrum control signals, means for de-
thesizer 26, a replica of the signals applied to the system veloping in response to each evaluation a signal repre-
by way of speech circuit 10. The reconstituted signals are sentative of the noise contribution to the respective spec-
delivered by way of circuit 27 to their ultimate destina- trum control signal, means for selectively subtracting said
lion, for example, to loudspeaker 28. noise representative signals from said', spectrum control
Any form of speech synthesizer may be used; so-called 55 signals, and means for delivering the difference signal re-
vocoder synthesizers are entirely satisfactory. Typically, sulting from said subtraction to an output circuit.
such a synthesizer requires, in addition to spectrum con- 6. Apparatus as defined in claim 5 wherein said means
trol signals, a relatively low noise indication of the for evaluating the contribution of noise to each of said
momentary pitch of the signals. Accordingly, pitch ana- spectrum control signals comprises, means for pc
lyzer 29 may be supplied with input signals from channel 00 trum control signals gating es c-
10. It delivers to synthesizer 26 the necessary indication below a preselected low level to a
of the pitch of the applied signal. A suitable low noise relatively long time-constant storage network, and means
pitch analyzer is described in the above-cited application for increasing the level of signals stored in said network
of Noll and Schroeder. by a preselected small factor greater than one.
Although it is often desirable to process speech signals 65 7. In combination with means for developing a group
or the like to remove embedded noise at one location of spectrum control signals representative of the speech
only, it is also frequently desirable to remove noise from energy falling within a plurality of frequency sub-hands
signals before transmission to a distant location or, after which collectively embrace the frequency band of a
reception, to remove noise added during transmission. speech signal, means for developing a group of signals
Accordingly, it will be understood that speech synthesizer 70 representative respectively of the contribution of noise to
26 may he located either with the processing apparatus each of said spectrum control signals, means for selec-
of the analyzer or at a distant location. That is to say, tively averaging said noise representative signals, and
transmission channel 30 may be of any desired extent, means for selectively combining said averaged representa-
The above-described arrangements are, of course, tive signals algebraically with said spectrum control sig-
merely illustrative of the application of the principles of 7n, nals.
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
 Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3
3,403,224
7
8. Apparatus for suppressing noise in a speech signal
which comprises, a source of a speech signal, means for
deriving from said speech signal a plurality of spectrum
control signals representative of the speech energy falling
within a plurality of frequency sub-bands which collec-
tively embrace a selected frequency band of said speech
signal, means responsive to minima in each of said spec-
trum control signals for developing a group of average
level signals representative of the contribution of noise
to each of said control signals, and means for subtracting
said average level signals from said corresponding control
signals.
9. Apparatus for suppressing noise in a speech signal
which comprises, a source of a speech signal, means for
deriving from said speech signal a plurality of spectrum
control signals representative of the speech energy falling
within a plurality of frequency sub-bands which collec-
tively embrace the frequency hand of said speech signal,
means re.,ponsive to signal minima in each of said spec-
trum control signals for developing a group of average
level signals representative of the contribution of noise
to each of said control signals, means for selectively ad-
justing the magnitudes of said average level signals, means
for selectively combining said adjusted average level sig-
nals, and means for subtracting said combined average
level signal from selected ones of said control signals.
1.0. In a narrow band speech transmission system which
comprises. in combination, means for deriving from an
applied speech wave a plurality of spectrum control sig-
nals representative of the speech energy falling within a
plurality of frequency sub-hands collectively embracing
the frequency hand of said speech wave, means for de-
riving from said speech wave a pitch control signal repre-
sentative of the momentary pitch of said wave, and means
8
responsive to said spectrum control signals and to said
pitch control signal for reconstituting said speech wave;
means for eliminating noise from each of said spectrum
control signals which comprises, means for developing a
group of signals representative respectively of the contri-
bution of noise to each of said control signals, means for
selectively averaging said noise representative signals,
and means for selectively combining said noise representa-
tive signals algebraically with said control signals.
11. The method of eliminating noise from a message
signal which includes the steps of: analyzing a message
signal to develop a measure of the energy in each of a
selected number of frequency sub-bands of the signal,
developing a running estimate of the contribution of noise
to each of said measures, changing each of said running
estimates whenever the level of the corresponding measure
of energy falls below a pre-established level, smoothing
the changed running estimates, modifying the magnitudes
of the smoothed running estimates by a selected factor,
and selectively reducing the magnitudes of the individual
measures of energy by the magnitudes of the modified
running estimates.
References Cited
UNITED STATES PATENTS
1,968,460
7/1934
Llewellyn ------------ 179-78
3,238,457
3/1966
Boymel et al. -------- 325-67
3,204,119
8/1965
Gray --------------- 328-163
3,180,936
4/1965
Schroeder ----------- 179-1.8
KATHLEEN H. CLAFFY, Primary Examiner.
R. P. TAYLOR, Assistant Examiner.
Approved For Release 2007/09/21 : CIA-RDP81-0012OR000100030010-3