CA1184506A - Method and system for discriminating human voice signal - Google Patents
Method and system for discriminating human voice signalInfo
- Publication number
- CA1184506A CA1184506A CA000375651A CA375651A CA1184506A CA 1184506 A CA1184506 A CA 1184506A CA 000375651 A CA000375651 A CA 000375651A CA 375651 A CA375651 A CA 375651A CA 1184506 A CA1184506 A CA 1184506A
- Authority
- CA
- Canada
- Prior art keywords
- signal
- control
- input signal
- input
- human voice
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
- 238000000034 method Methods 0.000 title claims description 8
- 230000005236 sound signal Effects 0.000 claims abstract description 10
- 238000001228 spectrum Methods 0.000 claims abstract description 7
- 210000001260 vocal cord Anatomy 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 8
- 238000009499 grossing Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 14
- 206010044565 Tremor Diseases 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 230000035939 shock Effects 0.000 description 4
- 230000001755 vocal effect Effects 0.000 description 4
- 150000002500 ions Chemical class 0.000 description 3
- 230000002238 attenuated effect Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 241000331231 Amorphocerini gen. n. 1 DAD-2008 Species 0.000 description 1
- NLZUEZXRPGMBCV-UHFFFAOYSA-N Butylhydroxytoluene Chemical compound CC1=CC(C(C)(C)C)=C(O)C(C(C)(C)C)=C1 NLZUEZXRPGMBCV-UHFFFAOYSA-N 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- JXSJBGJIGXNWCI-UHFFFAOYSA-N diethyl 2-[(dimethoxyphosphorothioyl)thio]succinate Chemical compound CCOC(=O)CC(SP(=S)(OC)OC)C(=O)OCC JXSJBGJIGXNWCI-UHFFFAOYSA-N 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000003137 locomotive effect Effects 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 210000000214 mouth Anatomy 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 229940061319 ovide Drugs 0.000 description 1
- 210000003800 pharynx Anatomy 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 210000003437 trachea Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/78—Detection of presence or absence of voice signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R2225/00—Details of deaf aids covered by H04R25/00, not provided for in any of its subgroups
- H04R2225/43—Signal processing in hearing aids to enhance the speech intelligibility
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Tone Control, Compression And Expansion, Limiting Amplitude (AREA)
- Circuit For Audible Band Transducer (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
ABSTRACT
The human voice is discriminated from other sounds in an audio signal by first recognizing that the human voice includes a fundamental frequency saw-tooth wave in the low bass region. Sounds in this region are also important for effects where body felt vibrations are to be transmitted direct to a listener or appreciator of the sound. The gain of an amplifier for using such sound effects is controlled in accordance with frequencies in the audio spectrum below that containing appreciable human voice frequency components. Apparatus detects the level of the total audio signal and subtracts the output of a human voice detector circuit, to provide the control for the amplifier. Human voice detector circuits exploit the fact that the human voice signal can be regarded as an amplitude modulated carrier wave. Sound effect frequencies in the range below that containing substantial frequency components of the human voice may be added to the detected total signal level from which the output of the human voice detector is subtracted before application to the amplifier.
The human voice is discriminated from other sounds in an audio signal by first recognizing that the human voice includes a fundamental frequency saw-tooth wave in the low bass region. Sounds in this region are also important for effects where body felt vibrations are to be transmitted direct to a listener or appreciator of the sound. The gain of an amplifier for using such sound effects is controlled in accordance with frequencies in the audio spectrum below that containing appreciable human voice frequency components. Apparatus detects the level of the total audio signal and subtracts the output of a human voice detector circuit, to provide the control for the amplifier. Human voice detector circuits exploit the fact that the human voice signal can be regarded as an amplitude modulated carrier wave. Sound effect frequencies in the range below that containing substantial frequency components of the human voice may be added to the detected total signal level from which the output of the human voice detector is subtracted before application to the amplifier.
Description
METHOD ~D SYSTEM FOR DISCRIMINATLNG A HUMAN VOICE SIGNAL
This invention relates to a method for discriminating between a human voice signal and any other sound signal, and also to a system for carrying out the method.
In the human living environment, various sounds are present. When re-cording and reproducing the sounds, high sensitivity and high fidelity are re-qulred.
In the genera] case, all the sounds which enter an appropriate micro-p~lone llave only to be equally amplified and reproduced. In special instances, I() however, it becomes necessary to classify input signals depending on their pro-pertles or types and to subject some types of signals to processing different from those of other types~
Thus, one may wish, for example, to derive human voices from among noise or to p~ck up only the voice of an announcer separately from the other sounds, such as music when recording a radio or television broadcast. This technique has also been proposed and applied to generation of body-felt vibration.
~ body-felt vibration generator is one which, gives ambience to sound repro~ cin~ equ;ipment/ by converting signals of lower bass sound, e.g., below 15~ tlz ~o mechanlcal vibrations so as to present these directly to the human body as hotly-Eelt vibrations. The sounds and the body felt vibrations of the lower bass sound region are simultaneously presented to a listener (or better, the "appreciator') thereby to intensify a heavy lower bass sound feeling. The device has achieved effects principally for the appreciation of music.
It has been experimentally confirmed that, when the device is combined with sounds other than music, for example, documentary sounds and sound effects attendant on earth tremors, or shocks such as gun blasts, the operating sound of a street car or steam locomotive, the noise emitted by felling of a large tree, e~p~osion, or the running and engine sounds of an automobile, tractor etc., ambience and dramatic effects which cannot be obtained with conventional sound reproducing equipment can be brought forth.
When such a device is incorporated into the cinema, television or the like~ presence and numerous dramatic effects are obtained. A film or televi-sion drama, for instance, is in general, composed of the combination of human voices (conversation), music, and the sound effects mentioned.
When such sound signals are presented to the appreciator in ~he form of body-felt vibrations without performing any additional processing, the same strong effects can be expected from sounds associated with earth tremor and other shocks, as from normal conversation. This has been experimentally shown to cause a very unnatural feeling.
In order to produce good dramatic effects, the advent of a device which Ls responsive to the muslcal and effect sounds but is not responsive to the hu-m~l~ VV'lC~ .18 de~ired.
The production and features of the human voice will now be considered.
When the vocal cords vibrate to open and close the trachea at its natural fre-ency, pe~odlc pulsating pressure waves are generated at the vocal cords. It Is Icnown these waves are approximately saw-tooth in shape with a fundamental ~rc~u~ncy at approximately 110 Hz, but which varies over quite a wide range,
This invention relates to a method for discriminating between a human voice signal and any other sound signal, and also to a system for carrying out the method.
In the human living environment, various sounds are present. When re-cording and reproducing the sounds, high sensitivity and high fidelity are re-qulred.
In the genera] case, all the sounds which enter an appropriate micro-p~lone llave only to be equally amplified and reproduced. In special instances, I() however, it becomes necessary to classify input signals depending on their pro-pertles or types and to subject some types of signals to processing different from those of other types~
Thus, one may wish, for example, to derive human voices from among noise or to p~ck up only the voice of an announcer separately from the other sounds, such as music when recording a radio or television broadcast. This technique has also been proposed and applied to generation of body-felt vibration.
~ body-felt vibration generator is one which, gives ambience to sound repro~ cin~ equ;ipment/ by converting signals of lower bass sound, e.g., below 15~ tlz ~o mechanlcal vibrations so as to present these directly to the human body as hotly-Eelt vibrations. The sounds and the body felt vibrations of the lower bass sound region are simultaneously presented to a listener (or better, the "appreciator') thereby to intensify a heavy lower bass sound feeling. The device has achieved effects principally for the appreciation of music.
It has been experimentally confirmed that, when the device is combined with sounds other than music, for example, documentary sounds and sound effects attendant on earth tremors, or shocks such as gun blasts, the operating sound of a street car or steam locomotive, the noise emitted by felling of a large tree, e~p~osion, or the running and engine sounds of an automobile, tractor etc., ambience and dramatic effects which cannot be obtained with conventional sound reproducing equipment can be brought forth.
When such a device is incorporated into the cinema, television or the like~ presence and numerous dramatic effects are obtained. A film or televi-sion drama, for instance, is in general, composed of the combination of human voices (conversation), music, and the sound effects mentioned.
When such sound signals are presented to the appreciator in ~he form of body-felt vibrations without performing any additional processing, the same strong effects can be expected from sounds associated with earth tremor and other shocks, as from normal conversation. This has been experimentally shown to cause a very unnatural feeling.
In order to produce good dramatic effects, the advent of a device which Ls responsive to the muslcal and effect sounds but is not responsive to the hu-m~l~ VV'lC~ .18 de~ired.
The production and features of the human voice will now be considered.
When the vocal cords vibrate to open and close the trachea at its natural fre-ency, pe~odlc pulsating pressure waves are generated at the vocal cords. It Is Icnown these waves are approximately saw-tooth in shape with a fundamental ~rc~u~ncy at approximately 110 Hz, but which varies over quite a wide range,
2~ particularly in ~he voice of an adult male during ordinary conversation. The saw~tooth waves generated at the vocal-cord sound source are modified by com-plicated formants such as by resonance and antiresonance while propagating through the vocal tube consisting of the pharynx, oral cavity, nasal cavity etc./ and they are emitted from the lips as the voice. (In general, the reso-nance frequencies are distributed at frequencies considerably higher than the fundamental saw-tooth wave frequency.
More particularly in accordance with one aspect of the inventi.on there is provided, a method for de~ermining the presence of a human voice in a sound sienal co~priain~:
providi~ An input sound si~n~l havin~ an input si~nal level;
detecting said input sound signsl ~o determi~e the presence of fund~mental wave wherein a human voice is represented as a s~rrier wave p~oduced by voice formants and a f~ndamental wave produced by the vocal co~d~ modulatin~ sGid carrier w~ve;
providing an emphasized output in response to said detectin~ to 1~ p~ovide a representation of snid fundament~l wave at an output level; and comparin~ the output level of said fundamental wave with the si~nal levcl of said inpu~ signal to determine the presence of a human voice i~
said input sound ~i~nal.
In accordance wlth a second aspect of the invention there is provided, a human voice signal discriminating system comprising:
means for providing an input signal which may contain human voice slgnals;
control means, having an input terminal, output terminal and control tcrmlnal, for receiving said illpUt slgnal at said input terminal and varlably providing said input signal at said output terminal in response ~o 2() a control signal applied to said control terminal; and means coupled to receive said input signal and provide a control signal to said control terminal in response to the detection of frequencles ln said input slgnal less than the fundamental wave component of h~man speech. The control means may be a voltage controlled variable gain amplifier or a voltage control variable frequency filter or an analog - 2a -switch. There may be first means for low pass filtering the Lnput signal received by the input signal providlng means and s2cond means for low pass filtering the input signal received by the input terminal, the first means having a frequency cut off lower than the second means.
In accordance with a third aspect of the invention there is provided, a ~ystem for discriminating a human voice slgnal comprislng:
means for providlng an input signal which may contain a human voice signa];
lnput level deriving circuit means coupled to receive said input signal for rectifying said input signal and smoothlng said rectified signal to provide a direct current output propor~ional to a level of said input ~ignal;
a human voice fundamental wave component level deriving circuit me~n~ coupled to receive said input signal and provide an outpu~ in response to the det:ection of a fundemental wave component of the human voice in said Lnput signal;
arithmetic circuit means coupled to receive said outputs from said L~put level derlving circuit means and said fundamental wave component level (lerLvlng circult means for subtracting one from the other to produce a 2.0 di~ference output;
control means having a control terminal and coupled t~ receive said input signal and provlde a variable level output of said input signal in response to a control slgnal applied to said control terminal, said difference output being coupled as the control signal applied to said control terminal.
In accordance with ~ fourth aspect of the lnvention there is provided, a humnn voice signnl discriminatin~ system, comprislng:
menns for providing ~n input signal which may cont~in human voice signals;
control menns having an input terminal, ou~put terminal and control terminal, for receiving snid input signal at snid input terminal and variQbly providing said input signal at said output terminal in response to control si~nal applied to said control terminal; And a low base sound deriving circuit means coupled to receive snid input si~nal and provide a control signal to shid control terminal in response to frequeneies ~n said input signnl which are less than th~t of the eundamental wave co~ponent of human speech in said input signal.
In accorda~ce with a fifth aspect of the invention there is provided a human voice signsl discri~inating system co~p~ising:
~enns for providin~ an input signal which ~ay contain hu~an voice ~l~n41u;
~ ean3 for low-p~ss filtering sald input signal to provide a fllte~ed input signal hnving frequencies in Q r~nge of 0-150 Hz;
control me~ns having an input terminal, output terminal, nnd eontrol te~minnll for receiving said filtered input signal at sQid input terminnl ~nd variably providing said filtered input signal at said output t~rmLnnl in renponse to a control signal applied to said control ter~inal;
~nd means for low-pnss filtering snid input signal to provide ~ control slgnnl to said control terminal in response to ~requency region below the lowest no~mal point of the humnn voice spectrum in said input si~nal to cause said control means to pass said erequencies in snid range of 0-150 Hz to snid output terminal in response to said control signal resulting from frequencies in said frequ0ncy re~ion.
In the description which follows5 reference Will be made to the accompanying drawings, ln which, ~ 2c -Figure 1 is a block diagram of a system embodying the invention, Figure 2(a) to 2(d) are waveform diagrams schematically showing the waveforms of a human voice and sounds of music etc., Figure 3 is a block diagram of an improved system embodying the inven-tion, Figure 4 is a block diagram detailing the detector circuit shown in Figure 3 for the fundamental of a human-voice signal, Figure 5 and 6 are block diagrams each showing practical detector cir-cultr" and 1() Figure 7 is a block diagram of a further arrangement embodying the in-vention.
Human voice waveforms are schematically illustrated in Figures 2(a) to 2(c). Figure 2(a) shows the saw-tooth wave of the vocal-cord sound source, whlle Figure 2(b) shows the voice waveform given the formants in the vocal tube. ~he saw-tooth waves produced at the vocal cords include abundant higher harmonics and hence, when they are given the formants in the resonance circuit part o the vocal tube, they become as shown in Figure 2(b).
a~ seen ~rom this figure, the ~undamental wave COmpOnellt decreases, and ttle Eormant frequency components increase considerably.
On the other hand the amplitude modulation of a carrier wave by a saw-tooth waveform gives the resultant shown in Figure 2(c). When Figures 2(b) and 2(c) are compared, it will be seen that although the processes of production of the waveforms a-re quite different, both are very similar macroscopically.
Therefore, if the voice waveform in Figure 2(b) is regarded as a kind of amplitude-modulated wave, its detection permits one readily to derive the fundamental frequency component of the vocal-cord sound source.
A system here described has a lower bass sound region deriving circuit ._ _. .
which obtains the bass frequency components which are lower than the fundamental wave components of the human voice and a control means such as, for example, a voltage-controlled variable gain amplifier whose gain varies in correspondence with the voltage applied to its control terminal, with the control means is controlled by the output of the lower bass region deriving circuit, thus a sound accompanied by shock such as earth tremor which carries ab(llldant ~requency components below approximately 60 Hz and a human voice which carrie~ a]most none of these components can be discriminated from one another.
Further~ described is a system with an input level deriving circuit whictl rectifies an input signal and integrates the rectified signal to provide a d,c. signal proportional to the input signal level, and a human-voice fundamental wave deriving circuit which obtains the fundamental wave component of a vocal cord sound source. Outputs from these two circuits are connected to a dlfferential circuit which subtracts an output of the human voice fundamental wave derlving circult from an output of the input level deriving circuit, and the ou~put of the differential circuit is connected to the control ~erminal of said control mean~.
Speclfic embodiments of the invent-Lon wlll now be described in conJllnctlon with a body-felt vlbration generator which is one example of the appllcation, When the frequency components of humall voices, music, sound ?~ ffects etc. are studied, the low region of the lluman voice (chiefly, the male voice) extends sometimes to just below 80 ~Iz with the high region extending above 10 kHz in both the male and female voices (dependent to some extent upon recording circumstances and individual differences).
The sound effects are of very wide variety and have varied frequency spectra. An earth tremor or such a sound giving a shock feeling is characteri7ed by including wide frequency distribution and abundant lower bass sound components~ Musical sounds often have a frequency distribution which is uniform and~wide compared with those of the other sounds.
One measure which suggests itself for distinguishing the human voice from the other sounds is to pass them through an appropriate filter. However, a low-pass filter which can attenuate the fundamental wave component of the human voice extending below 80 Hz has in practice an upper cutoff frequency of 60 1-1z or even less. This poses a problem. In frequency components below approximately 150 ~1z which are effective for body-felt vibrations, the frequency cornponent density is greatest in the spectrum around 100 Hz, with the dlstribution density at and below 50 - ~0 Hz falling off. Thus signals passed through such a low-pass filter have most of the components for the body-felt vibrations attenuated, and the effect of the body-felt vibration generator is detrimentally affected.
~ arth tremor includes sufficient lower bass sound components of and below 60 ~1z and it can give rise to the body-felt vibrations. Without the components near 100 Hz, however, the feeling of "presence" is lessened, and ~mvivid Elat vibrations are produced. This situation can be likened to the sound from a loudspeaker assembly from which the tweeter has been detached so tha~ It does not give hlgh register sounds.
FLgure 1 shows an example of a circuit which has eliminated such problems, ~n input terrnlnal, receives an audio signal from an amplifier (not 2~ S~lOWn), A ~irst low-pass filter 3 having an upper cutoff frequency of approximately 150 1-1z and a voltage-controlled variable gain amplifier (VCA) 4 are connected between the input terminal 1 and an output terminal 2.
'1'he low-pass filter 3 derives lower-bass sound components below approximately 150 Hz useful for supply to a body-felt vibration generator. The voltage-controlled variable gain amplifier 4 has its gain controlled by application of a voltage to the control port 4a. In the present example, an amplifier whose gain increases w-ith applied positive voltage is employed.
Furt~er connected to t~1e input terminal 1 is the input side of a second ]ow~pass filter 5 having an upper cutoff frequency of approximately 60 Hz. The output side of the second low--pass filter 5 is connected to the control port 4a through a rectifier circuit 6 and an integrator circuit 7.
When a sound efEect such as an earth tremor with abundant lower bass sound components of and below 60 Hz enters the input at terminal 1 the two low-pass filters allow passage of these lower bass components. The output sign<ll from the second low-pass filter 5 is rectified by the circuit 6, and is then integrated in circuit 7. The output of the integrator ~ircuit 7 is appliecl to the control terminal 4a of amplifier 4, to increase its gain. Thus, the lower bass signal passed Lhrough the first low~pass filter 3 is strongly amplified and delivered to the output terminal 2. Accordingly, a body-felt vlbration generator can be drLven by the output from terminal 2, and an appreclator reeeive the body-felt vibrat:ions produced. Since the output signal includes frequency components near 100 Hz, realistic vibration can be achieved.
Now, consider the case where a human voice signal is the input signal.
~llCIl a sLgnal includes very lLttle lower bass sound of and below 60 Hz. No na'l thereore pasoes through the second low~pass filter 5, and the output of the ~ egrator circult 7 is low. The gain of the voltage-controlled variable gaLn aml)Llfler ~ is also low, and the amplification of the signal arriving from 2(~ ~:he Low-pass fllter 3 is low, and the output signal at terminal 2 is small.
The human voice thus gives rise to essentially no body felt vibration.
In musical and other sound effects from which body-felt vlbrations are desirably,generated, there are many which do not include or anly include at low level the components of and below 60 Hz~ The apparatus of Figure 1 cannot generate body-felt vibration when components of and below 60 Hz are not present.
Figure 3 is a block diagram of an apparatus which meets this problem.
This circui~ serves to emphasize and derive the fundamental wave component generated by a vocal cord sound source, and then compares this with the total input signal level, thereby to discri~inate the human voice from other sound.
The apparatus comprises an input level deriving circuit 8, a human-voice fundamental wave deriving circuit 9, an arithmetic circuit lO, and also the low-pass filter 3 and the voltage-controlled variable gain ampl~fier 4 described earller.
FirstLy, the human-voice fundamental wave deriving circuit 9 will be desc:ribed in detail. Basicall~, this is an amplitude modulation detector circuit and will be explained in conjunction with ~he simplified circuit shown in ligure 4. Figure 4 consists of a detector ll and a band-pass filter 12 for selecting the frequency rallge of 80 to 300 Hæ. The fundamental wave component o~ tlle vocal-cord sourld source can be derived from this.
Ilowever, a waveform o a sound other than the human voice, for example, a muslcal sound ls as shown in Figure 2(d). This signal covers a wide frequency spectrum including both low and high frequencies and when detected all~l pasc3ed ~hrough the band-pass filter in the circuit of Figure 4, produces a t~latlveLy small ou~put. Since the music waveform covers a wide frequency rall~e, the percentage of the signal component passing through the band-pass eLItcr l?. rclatLvely to the total input signal ls small. ~Lso, in contrast to .0 that sho~m in Figure 2(b), the music waveform is not of the definite shape whlch resembles an amplitude-modulated carrier wave.
~ ccordLngly, when the input signal and output signal of the circuit of Figure 4 are respectively rectified, appropriately weighted and are compared and calcula~ed, it becomes possible to distinguish the human voice from other sounds. Where an input signal at a chosen fixed level is app]ied to the circuit of Flgure 4, signals at relatively high output can be regarded as those of the human voice, whereas those at relatively low output can be taken to be otller sounds..
A waveform given the formants is not a simple sinu.soidal wave as in a carrier wave shown in Figure 2(c), and the human voice waveform shown in Figure 2(b) is not generally vertically symmetric. For this reason, when detecting the input signal with the circuit of Figure 4~ the solution of the cal.culation difers greatly depending upon whether the envelope on the positive or the negat:lve side is used.
In the waveform illustrated in Figure 2(b), it is apparent that the d:Lstinguishability is enhanced on the positive side of the envelope because the output at terminal in Figure 4 is thus increased. The waveform shown in F:Lgure 2(b) has hi~h peaks on the positive going side, but there is the equally frequent inverse case where higher peaks are negative go~ng. The detector ll o~ F:lgure 4 must therefore derive the higher level envelope regardless oi whether l:his :Ls positive and negative going.
In order to fulfill this condition, the detection circuit may be of the tull-w~vQ rect.i~icatlon type. Since, however~ the full-wave rectifier circuit nlso acts ns ~ ~requency multLplier, the input frequency is doubled. Let it nOw he a~,sumed that thQ det.ection circuit 11 i5 a full-wave rectification type In the c;Lre~ of Figure ~ and that a slnusoidal wave of 50 Hz has entered the lnput. ~ pul~scltlng waveform with a fundamental wave component of 100 H~
(double frequency) appears at the detector output, with the result that an output from the band-pass filter 12 having the pass band of 80 - 300 Hz is obtained~
However, should the lower bass sound attendant on music or earth tremor enter the input, its frequency will be multiplied to provide a derived outputO
In some instances however this gives the drawback that the distinguishability trom the human voice is lowered.
To eliminate this disadvantage, the circuit showll in Figure 5 includes a high-pas6~f~1ter 13 which attenuates the lower bass region which is not
More particularly in accordance with one aspect of the inventi.on there is provided, a method for de~ermining the presence of a human voice in a sound sienal co~priain~:
providi~ An input sound si~n~l havin~ an input si~nal level;
detecting said input sound signsl ~o determi~e the presence of fund~mental wave wherein a human voice is represented as a s~rrier wave p~oduced by voice formants and a f~ndamental wave produced by the vocal co~d~ modulatin~ sGid carrier w~ve;
providing an emphasized output in response to said detectin~ to 1~ p~ovide a representation of snid fundament~l wave at an output level; and comparin~ the output level of said fundamental wave with the si~nal levcl of said inpu~ signal to determine the presence of a human voice i~
said input sound ~i~nal.
In accordance wlth a second aspect of the invention there is provided, a human voice signal discriminating system comprising:
means for providing an input signal which may contain human voice slgnals;
control means, having an input terminal, output terminal and control tcrmlnal, for receiving said illpUt slgnal at said input terminal and varlably providing said input signal at said output terminal in response ~o 2() a control signal applied to said control terminal; and means coupled to receive said input signal and provide a control signal to said control terminal in response to the detection of frequencles ln said input slgnal less than the fundamental wave component of h~man speech. The control means may be a voltage controlled variable gain amplifier or a voltage control variable frequency filter or an analog - 2a -switch. There may be first means for low pass filtering the Lnput signal received by the input signal providlng means and s2cond means for low pass filtering the input signal received by the input terminal, the first means having a frequency cut off lower than the second means.
In accordance with a third aspect of the invention there is provided, a ~ystem for discriminating a human voice slgnal comprislng:
means for providlng an input signal which may contain a human voice signa];
lnput level deriving circuit means coupled to receive said input signal for rectifying said input signal and smoothlng said rectified signal to provide a direct current output propor~ional to a level of said input ~ignal;
a human voice fundamental wave component level deriving circuit me~n~ coupled to receive said input signal and provide an outpu~ in response to the det:ection of a fundemental wave component of the human voice in said Lnput signal;
arithmetic circuit means coupled to receive said outputs from said L~put level derlving circuit means and said fundamental wave component level (lerLvlng circult means for subtracting one from the other to produce a 2.0 di~ference output;
control means having a control terminal and coupled t~ receive said input signal and provlde a variable level output of said input signal in response to a control slgnal applied to said control terminal, said difference output being coupled as the control signal applied to said control terminal.
In accordance with ~ fourth aspect of the lnvention there is provided, a humnn voice signnl discriminatin~ system, comprislng:
menns for providing ~n input signal which may cont~in human voice signals;
control menns having an input terminal, ou~put terminal and control terminal, for receiving snid input signal at snid input terminal and variQbly providing said input signal at said output terminal in response to control si~nal applied to said control terminal; And a low base sound deriving circuit means coupled to receive snid input si~nal and provide a control signal to shid control terminal in response to frequeneies ~n said input signnl which are less than th~t of the eundamental wave co~ponent of human speech in said input signal.
In accorda~ce with a fifth aspect of the invention there is provided a human voice signsl discri~inating system co~p~ising:
~enns for providin~ an input signal which ~ay contain hu~an voice ~l~n41u;
~ ean3 for low-p~ss filtering sald input signal to provide a fllte~ed input signal hnving frequencies in Q r~nge of 0-150 Hz;
control me~ns having an input terminal, output terminal, nnd eontrol te~minnll for receiving said filtered input signal at sQid input terminnl ~nd variably providing said filtered input signal at said output t~rmLnnl in renponse to a control signal applied to said control ter~inal;
~nd means for low-pnss filtering snid input signal to provide ~ control slgnnl to said control terminal in response to ~requency region below the lowest no~mal point of the humnn voice spectrum in said input si~nal to cause said control means to pass said erequencies in snid range of 0-150 Hz to snid output terminal in response to said control signal resulting from frequencies in said frequ0ncy re~ion.
In the description which follows5 reference Will be made to the accompanying drawings, ln which, ~ 2c -Figure 1 is a block diagram of a system embodying the invention, Figure 2(a) to 2(d) are waveform diagrams schematically showing the waveforms of a human voice and sounds of music etc., Figure 3 is a block diagram of an improved system embodying the inven-tion, Figure 4 is a block diagram detailing the detector circuit shown in Figure 3 for the fundamental of a human-voice signal, Figure 5 and 6 are block diagrams each showing practical detector cir-cultr" and 1() Figure 7 is a block diagram of a further arrangement embodying the in-vention.
Human voice waveforms are schematically illustrated in Figures 2(a) to 2(c). Figure 2(a) shows the saw-tooth wave of the vocal-cord sound source, whlle Figure 2(b) shows the voice waveform given the formants in the vocal tube. ~he saw-tooth waves produced at the vocal cords include abundant higher harmonics and hence, when they are given the formants in the resonance circuit part o the vocal tube, they become as shown in Figure 2(b).
a~ seen ~rom this figure, the ~undamental wave COmpOnellt decreases, and ttle Eormant frequency components increase considerably.
On the other hand the amplitude modulation of a carrier wave by a saw-tooth waveform gives the resultant shown in Figure 2(c). When Figures 2(b) and 2(c) are compared, it will be seen that although the processes of production of the waveforms a-re quite different, both are very similar macroscopically.
Therefore, if the voice waveform in Figure 2(b) is regarded as a kind of amplitude-modulated wave, its detection permits one readily to derive the fundamental frequency component of the vocal-cord sound source.
A system here described has a lower bass sound region deriving circuit ._ _. .
which obtains the bass frequency components which are lower than the fundamental wave components of the human voice and a control means such as, for example, a voltage-controlled variable gain amplifier whose gain varies in correspondence with the voltage applied to its control terminal, with the control means is controlled by the output of the lower bass region deriving circuit, thus a sound accompanied by shock such as earth tremor which carries ab(llldant ~requency components below approximately 60 Hz and a human voice which carrie~ a]most none of these components can be discriminated from one another.
Further~ described is a system with an input level deriving circuit whictl rectifies an input signal and integrates the rectified signal to provide a d,c. signal proportional to the input signal level, and a human-voice fundamental wave deriving circuit which obtains the fundamental wave component of a vocal cord sound source. Outputs from these two circuits are connected to a dlfferential circuit which subtracts an output of the human voice fundamental wave derlving circult from an output of the input level deriving circuit, and the ou~put of the differential circuit is connected to the control ~erminal of said control mean~.
Speclfic embodiments of the invent-Lon wlll now be described in conJllnctlon with a body-felt vlbration generator which is one example of the appllcation, When the frequency components of humall voices, music, sound ?~ ffects etc. are studied, the low region of the lluman voice (chiefly, the male voice) extends sometimes to just below 80 ~Iz with the high region extending above 10 kHz in both the male and female voices (dependent to some extent upon recording circumstances and individual differences).
The sound effects are of very wide variety and have varied frequency spectra. An earth tremor or such a sound giving a shock feeling is characteri7ed by including wide frequency distribution and abundant lower bass sound components~ Musical sounds often have a frequency distribution which is uniform and~wide compared with those of the other sounds.
One measure which suggests itself for distinguishing the human voice from the other sounds is to pass them through an appropriate filter. However, a low-pass filter which can attenuate the fundamental wave component of the human voice extending below 80 Hz has in practice an upper cutoff frequency of 60 1-1z or even less. This poses a problem. In frequency components below approximately 150 ~1z which are effective for body-felt vibrations, the frequency cornponent density is greatest in the spectrum around 100 Hz, with the dlstribution density at and below 50 - ~0 Hz falling off. Thus signals passed through such a low-pass filter have most of the components for the body-felt vibrations attenuated, and the effect of the body-felt vibration generator is detrimentally affected.
~ arth tremor includes sufficient lower bass sound components of and below 60 ~1z and it can give rise to the body-felt vibrations. Without the components near 100 Hz, however, the feeling of "presence" is lessened, and ~mvivid Elat vibrations are produced. This situation can be likened to the sound from a loudspeaker assembly from which the tweeter has been detached so tha~ It does not give hlgh register sounds.
FLgure 1 shows an example of a circuit which has eliminated such problems, ~n input terrnlnal, receives an audio signal from an amplifier (not 2~ S~lOWn), A ~irst low-pass filter 3 having an upper cutoff frequency of approximately 150 1-1z and a voltage-controlled variable gain amplifier (VCA) 4 are connected between the input terminal 1 and an output terminal 2.
'1'he low-pass filter 3 derives lower-bass sound components below approximately 150 Hz useful for supply to a body-felt vibration generator. The voltage-controlled variable gain amplifier 4 has its gain controlled by application of a voltage to the control port 4a. In the present example, an amplifier whose gain increases w-ith applied positive voltage is employed.
Furt~er connected to t~1e input terminal 1 is the input side of a second ]ow~pass filter 5 having an upper cutoff frequency of approximately 60 Hz. The output side of the second low--pass filter 5 is connected to the control port 4a through a rectifier circuit 6 and an integrator circuit 7.
When a sound efEect such as an earth tremor with abundant lower bass sound components of and below 60 Hz enters the input at terminal 1 the two low-pass filters allow passage of these lower bass components. The output sign<ll from the second low-pass filter 5 is rectified by the circuit 6, and is then integrated in circuit 7. The output of the integrator ~ircuit 7 is appliecl to the control terminal 4a of amplifier 4, to increase its gain. Thus, the lower bass signal passed Lhrough the first low~pass filter 3 is strongly amplified and delivered to the output terminal 2. Accordingly, a body-felt vlbration generator can be drLven by the output from terminal 2, and an appreclator reeeive the body-felt vibrat:ions produced. Since the output signal includes frequency components near 100 Hz, realistic vibration can be achieved.
Now, consider the case where a human voice signal is the input signal.
~llCIl a sLgnal includes very lLttle lower bass sound of and below 60 Hz. No na'l thereore pasoes through the second low~pass filter 5, and the output of the ~ egrator circult 7 is low. The gain of the voltage-controlled variable gaLn aml)Llfler ~ is also low, and the amplification of the signal arriving from 2(~ ~:he Low-pass fllter 3 is low, and the output signal at terminal 2 is small.
The human voice thus gives rise to essentially no body felt vibration.
In musical and other sound effects from which body-felt vlbrations are desirably,generated, there are many which do not include or anly include at low level the components of and below 60 Hz~ The apparatus of Figure 1 cannot generate body-felt vibration when components of and below 60 Hz are not present.
Figure 3 is a block diagram of an apparatus which meets this problem.
This circui~ serves to emphasize and derive the fundamental wave component generated by a vocal cord sound source, and then compares this with the total input signal level, thereby to discri~inate the human voice from other sound.
The apparatus comprises an input level deriving circuit 8, a human-voice fundamental wave deriving circuit 9, an arithmetic circuit lO, and also the low-pass filter 3 and the voltage-controlled variable gain ampl~fier 4 described earller.
FirstLy, the human-voice fundamental wave deriving circuit 9 will be desc:ribed in detail. Basicall~, this is an amplitude modulation detector circuit and will be explained in conjunction with ~he simplified circuit shown in ligure 4. Figure 4 consists of a detector ll and a band-pass filter 12 for selecting the frequency rallge of 80 to 300 Hæ. The fundamental wave component o~ tlle vocal-cord sourld source can be derived from this.
Ilowever, a waveform o a sound other than the human voice, for example, a muslcal sound ls as shown in Figure 2(d). This signal covers a wide frequency spectrum including both low and high frequencies and when detected all~l pasc3ed ~hrough the band-pass filter in the circuit of Figure 4, produces a t~latlveLy small ou~put. Since the music waveform covers a wide frequency rall~e, the percentage of the signal component passing through the band-pass eLItcr l?. rclatLvely to the total input signal ls small. ~Lso, in contrast to .0 that sho~m in Figure 2(b), the music waveform is not of the definite shape whlch resembles an amplitude-modulated carrier wave.
~ ccordLngly, when the input signal and output signal of the circuit of Figure 4 are respectively rectified, appropriately weighted and are compared and calcula~ed, it becomes possible to distinguish the human voice from other sounds. Where an input signal at a chosen fixed level is app]ied to the circuit of Flgure 4, signals at relatively high output can be regarded as those of the human voice, whereas those at relatively low output can be taken to be otller sounds..
A waveform given the formants is not a simple sinu.soidal wave as in a carrier wave shown in Figure 2(c), and the human voice waveform shown in Figure 2(b) is not generally vertically symmetric. For this reason, when detecting the input signal with the circuit of Figure 4~ the solution of the cal.culation difers greatly depending upon whether the envelope on the positive or the negat:lve side is used.
In the waveform illustrated in Figure 2(b), it is apparent that the d:Lstinguishability is enhanced on the positive side of the envelope because the output at terminal in Figure 4 is thus increased. The waveform shown in F:Lgure 2(b) has hi~h peaks on the positive going side, but there is the equally frequent inverse case where higher peaks are negative go~ng. The detector ll o~ F:lgure 4 must therefore derive the higher level envelope regardless oi whether l:his :Ls positive and negative going.
In order to fulfill this condition, the detection circuit may be of the tull-w~vQ rect.i~icatlon type. Since, however~ the full-wave rectifier circuit nlso acts ns ~ ~requency multLplier, the input frequency is doubled. Let it nOw he a~,sumed that thQ det.ection circuit 11 i5 a full-wave rectification type In the c;Lre~ of Figure ~ and that a slnusoidal wave of 50 Hz has entered the lnput. ~ pul~scltlng waveform with a fundamental wave component of 100 H~
(double frequency) appears at the detector output, with the result that an output from the band-pass filter 12 having the pass band of 80 - 300 Hz is obtained~
However, should the lower bass sound attendant on music or earth tremor enter the input, its frequency will be multiplied to provide a derived outputO
In some instances however this gives the drawback that the distinguishability trom the human voice is lowered.
To eliminate this disadvantage, the circuit showll in Figure 5 includes a high-pas6~f~1ter 13 which attenuates the lower bass region which is not
3$
significant for distinguishing the human voiceO Even when the succeeding full-wave detector circuit 14 performs the frequency multiplication, the component passing the band~pass filter l~ is reduced~ sy experiment, it was found suit-able to set the cutoff frequency of the highpass filter 13 at approximately 130 The output of the band-pass filter 12 is rectified by the succeeding rec~:lfler circuit (AC-to-DC converter) 15, and the output of circuit 15 has its higller frequency components attenuated by the low-pass filter 16. ~ DC output proportional to the output of the band-pass filter 12 is produced.
I:igure 6 shows an example of another circuit arrangement which carries out the same opera~ion. This circuit modifies the intermediate portion of the circuit of FLgure 5, Detector circuits 17 and 18 are not of the full-wave rectlflc~ltLon type l-aving the frequency Multiplying action but are of the half-wave rectification type. The positive going envelope is thus derived by cir-cuit 17 and the negative going envelope by circuit 18. The negative side en--velope has lts polarlty reversed by an invertion circuit 19~
Th~ ~espective detected OUtp~ltS are applied to band-pass filters 20 and 21 and are re-tified by recti~ler clrcuits 22 and 23, and thereafter, applled to a low-pass filter 16. ~ligh-pass filter 13 at the input is not in principle re~lulred, but its presence enhances the dis~inguisability. The detectors and rectifier circuits in Figures 4 to 6 should desirably be linear in order to reduce operation errors from nonllnearity.
The input level derlving circuit 8 of Figure 3 will now be described.
This circuit integrates or smooths the input signal as rectified and provides a DC level proportional to the input level.
The input sides of the input level deriving circuit 8 and the human-voice fundamental wave deriving circuit 9 are connected to the input terminal 1 together ~ith the input side of the lowpass filter 3. The outputs of the circuit 8 and the circuit 9 are respectively connected to input terminals lOa and lOb of the arithmetic circuit 10.
'rhe output signal of circuit 8 and the output signal circuit 9 are weigilted to appropriate levels. Appropriate weighting such as by selection of the time constant of the integrator circuit 7 and/or the cutoff frequency of the smoothing low-pass filter 3 is applied.
The arithmetic circuit lO operates on the appropriately weighted outputs to subtract that of circuit 9 from that of circuit 8. Where the input signal is the human voice, the output of circuit 9 becomes greater tharl that of the lnput level deriving circuit 8 owing to the appropriate weighting, and hence, the output o the aritt-metic circuit lO is negative. ~n contrast, where the input ~ignal is muslc or other sound, the output of the human-voice fundamental wave derlving circuit 9 becomes smaller than that of the input level deriving cLrcuit 8, and hence, the arithmetic circuit 10 has a positive output~
'I'he input ~ignal also passes through the low-pass filter 3 (having an p~ cutof~ Prequency of approximately 150 Hz) for deriving components ~L~ t-Lve ~or ~ody-~elt vLbrations and is applied to the variable gain amplLeler l~ When the lnput signal is the humcln voice, the output of amplifier (-~ is reduced, whereas when the input signal is any other sound signal, the gain Oe amplifier 4 is increased resulting in a greater output for the body-felt vLbration generator.
The voltage-controlled variable gain amplifier 4 can be replaced by a voltage-controlled variable frequency filter (VCF)~ It can also be replaced b~
a gate based on an analog switch, or the like. Since, however, the binary "on"-"of~" switching in this case gives rise to an unnatural feeling, the gentle and continuous "on"~"off" control from the analog increase and decrease of tlle voltage-controlled variable gain amplifier 4, etc. is more favorable.
_ ~.
In the above description of the human voice only cases of voiced phones have been considered. Since unvoiced phones (consonants produced without vibration of the vocal cords) have their spectra in the higher frequency audio band, they are cut by the low-pass filter 3 (Figure 1 or Figure 3) having an upper cutoff frequency of approximately 150 Hz. and cause no problem.
Even Ln the case of voiced phones, voices of high fundamental wave Lrequenc:Les such as children's voices, the female voi.ce or the highpitched singl.ng voice present no problem for the same reason. It has been experimentally verified that the difference in distinguishabil.ity of voices in different languages is scarcely detectable, which is thought also to be due to t~l.ls.
Figure 7 is a block dLagram o.f a practical circuit in which the elements s)f botll the clrcuits 1n Il':Lgures 1 and 3 are coml~ined. Two input terminals 1 are disposed so as to receive signals through two channels whicil are mixed in clrcu:it 24. On the output s:Lde of the mixer circuit 24 is a subsonic filter 25 ~r r~movLng unnecessary subsonic waves of and below 20 H~.
Tllc output side of the subsonlc filter 25 is branched to feed firstly a vol.tage-controlled variable ga:l.n amplifler 4 through a l.o~-pass filter 3 for ~l~rLvl.llg a body-felt v:Lhration signal (havillg an upyer cutoff frequency of lS0 11~) and secondly to feed an equalizer circuit 26. The equalizer circuit 26 g>:Lves the lower bass region and the high-pitched region appropriate equali~ing curves in advance in order to enhance the human-voice disti.nguishability.
Succeeding the equalizer circuit 26 is a level compressor circuit 27, to the output of which the inputs of a lower bass region derivi.ng circuit 28, an input level de~iving circuit 8, and a human-voice f-mdamental wave deriving circuit 9 are connectèd. The output sides of these circuit.s 28, 8 and 9 are respectlvely connected to the inputs 10c, lOa and lOb of an arithmetic circuit 1~.-The level compressor circuit 27 compresses the dynamic range of the input signal to prevent errors in discrimination and operation. If the level is too low, discrimination is difficult, if it is too high, the operational amplifier circuitry may be saturated. ~he lower bass sound region deriving circuit 28 detects very low bass sound components of sound ef~ects and corresponds to the part of Figure 1 enclosed by the chain-dotted line. ~he input level deriving circuit 8 and the human-voice fundamental wave deriving circuit 9 correspond to the circuits carrying the same respective reference numerals in Figure 3.
In this arrangement, the arithmetic circuit 10 functions ~o sub~ract the output of tne human-voice fundamental wave deriving circuit 9 from the sum of the outputs of the very low-bass sound region deriving circuit 28 and that of the input level derlvLng circuit 80 The operation is the same as described for lïigure 3 in that the solution of the resultant difference is used to control of the variable gain amplifier ~. When the weighting of the very low-bass region derlving circuit 28 is comparatively high and the circuit is set so as to p~event response to levels below a threshold value, good results are obtained.
~0
significant for distinguishing the human voiceO Even when the succeeding full-wave detector circuit 14 performs the frequency multiplication, the component passing the band~pass filter l~ is reduced~ sy experiment, it was found suit-able to set the cutoff frequency of the highpass filter 13 at approximately 130 The output of the band-pass filter 12 is rectified by the succeeding rec~:lfler circuit (AC-to-DC converter) 15, and the output of circuit 15 has its higller frequency components attenuated by the low-pass filter 16. ~ DC output proportional to the output of the band-pass filter 12 is produced.
I:igure 6 shows an example of another circuit arrangement which carries out the same opera~ion. This circuit modifies the intermediate portion of the circuit of FLgure 5, Detector circuits 17 and 18 are not of the full-wave rectlflc~ltLon type l-aving the frequency Multiplying action but are of the half-wave rectification type. The positive going envelope is thus derived by cir-cuit 17 and the negative going envelope by circuit 18. The negative side en--velope has lts polarlty reversed by an invertion circuit 19~
Th~ ~espective detected OUtp~ltS are applied to band-pass filters 20 and 21 and are re-tified by recti~ler clrcuits 22 and 23, and thereafter, applled to a low-pass filter 16. ~ligh-pass filter 13 at the input is not in principle re~lulred, but its presence enhances the dis~inguisability. The detectors and rectifier circuits in Figures 4 to 6 should desirably be linear in order to reduce operation errors from nonllnearity.
The input level derlving circuit 8 of Figure 3 will now be described.
This circuit integrates or smooths the input signal as rectified and provides a DC level proportional to the input level.
The input sides of the input level deriving circuit 8 and the human-voice fundamental wave deriving circuit 9 are connected to the input terminal 1 together ~ith the input side of the lowpass filter 3. The outputs of the circuit 8 and the circuit 9 are respectively connected to input terminals lOa and lOb of the arithmetic circuit 10.
'rhe output signal of circuit 8 and the output signal circuit 9 are weigilted to appropriate levels. Appropriate weighting such as by selection of the time constant of the integrator circuit 7 and/or the cutoff frequency of the smoothing low-pass filter 3 is applied.
The arithmetic circuit lO operates on the appropriately weighted outputs to subtract that of circuit 9 from that of circuit 8. Where the input signal is the human voice, the output of circuit 9 becomes greater tharl that of the lnput level deriving circuit 8 owing to the appropriate weighting, and hence, the output o the aritt-metic circuit lO is negative. ~n contrast, where the input ~ignal is muslc or other sound, the output of the human-voice fundamental wave derlving circuit 9 becomes smaller than that of the input level deriving cLrcuit 8, and hence, the arithmetic circuit 10 has a positive output~
'I'he input ~ignal also passes through the low-pass filter 3 (having an p~ cutof~ Prequency of approximately 150 Hz) for deriving components ~L~ t-Lve ~or ~ody-~elt vLbrations and is applied to the variable gain amplLeler l~ When the lnput signal is the humcln voice, the output of amplifier (-~ is reduced, whereas when the input signal is any other sound signal, the gain Oe amplifier 4 is increased resulting in a greater output for the body-felt vLbration generator.
The voltage-controlled variable gain amplifier 4 can be replaced by a voltage-controlled variable frequency filter (VCF)~ It can also be replaced b~
a gate based on an analog switch, or the like. Since, however, the binary "on"-"of~" switching in this case gives rise to an unnatural feeling, the gentle and continuous "on"~"off" control from the analog increase and decrease of tlle voltage-controlled variable gain amplifier 4, etc. is more favorable.
_ ~.
In the above description of the human voice only cases of voiced phones have been considered. Since unvoiced phones (consonants produced without vibration of the vocal cords) have their spectra in the higher frequency audio band, they are cut by the low-pass filter 3 (Figure 1 or Figure 3) having an upper cutoff frequency of approximately 150 Hz. and cause no problem.
Even Ln the case of voiced phones, voices of high fundamental wave Lrequenc:Les such as children's voices, the female voi.ce or the highpitched singl.ng voice present no problem for the same reason. It has been experimentally verified that the difference in distinguishabil.ity of voices in different languages is scarcely detectable, which is thought also to be due to t~l.ls.
Figure 7 is a block dLagram o.f a practical circuit in which the elements s)f botll the clrcuits 1n Il':Lgures 1 and 3 are coml~ined. Two input terminals 1 are disposed so as to receive signals through two channels whicil are mixed in clrcu:it 24. On the output s:Lde of the mixer circuit 24 is a subsonic filter 25 ~r r~movLng unnecessary subsonic waves of and below 20 H~.
Tllc output side of the subsonlc filter 25 is branched to feed firstly a vol.tage-controlled variable ga:l.n amplifler 4 through a l.o~-pass filter 3 for ~l~rLvl.llg a body-felt v:Lhration signal (havillg an upyer cutoff frequency of lS0 11~) and secondly to feed an equalizer circuit 26. The equalizer circuit 26 g>:Lves the lower bass region and the high-pitched region appropriate equali~ing curves in advance in order to enhance the human-voice disti.nguishability.
Succeeding the equalizer circuit 26 is a level compressor circuit 27, to the output of which the inputs of a lower bass region derivi.ng circuit 28, an input level de~iving circuit 8, and a human-voice f-mdamental wave deriving circuit 9 are connectèd. The output sides of these circuit.s 28, 8 and 9 are respectlvely connected to the inputs 10c, lOa and lOb of an arithmetic circuit 1~.-The level compressor circuit 27 compresses the dynamic range of the input signal to prevent errors in discrimination and operation. If the level is too low, discrimination is difficult, if it is too high, the operational amplifier circuitry may be saturated. ~he lower bass sound region deriving circuit 28 detects very low bass sound components of sound ef~ects and corresponds to the part of Figure 1 enclosed by the chain-dotted line. ~he input level deriving circuit 8 and the human-voice fundamental wave deriving circuit 9 correspond to the circuits carrying the same respective reference numerals in Figure 3.
In this arrangement, the arithmetic circuit 10 functions ~o sub~ract the output of tne human-voice fundamental wave deriving circuit 9 from the sum of the outputs of the very low-bass sound region deriving circuit 28 and that of the input level derlvLng circuit 80 The operation is the same as described for lïigure 3 in that the solution of the resultant difference is used to control of the variable gain amplifier ~. When the weighting of the very low-bass region derlving circuit 28 is comparatively high and the circuit is set so as to p~event response to levels below a threshold value, good results are obtained.
~0
Claims (13)
PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:
1. A method for determining the presence of a human voice in a sound signal comprising:
providing an input sound signal having an input signal level;
detecting said input sound signal to determine the presence of fundamental wave wherein a human voice is represented as a carrier wave produced by voice formants and a fundamental wave produced by the vocal cords modulating said carrier wave;
providing an emphasized output in response to said detecting to provide a representation of said fundamental wave at an output level; and comparing the output level of said fundamental wave with the signal level of said input signal to determine the presence of a human voice in said input sound signal.
providing an input sound signal having an input signal level;
detecting said input sound signal to determine the presence of fundamental wave wherein a human voice is represented as a carrier wave produced by voice formants and a fundamental wave produced by the vocal cords modulating said carrier wave;
providing an emphasized output in response to said detecting to provide a representation of said fundamental wave at an output level; and comparing the output level of said fundamental wave with the signal level of said input signal to determine the presence of a human voice in said input sound signal.
2. A human voice signal discriminating system, comprising:
means for providing an input signal which may contain human voice signals;
control means, having an input terminal, output terminal and control terminal, for receiving said input signal at said input terminal and variably providing said input signal at said output terminal in response to n control signal applied to said control terminal; and means coupled to receive said input signal and provide a control signal to said control terminal in response to the detection of frequencies in said input signal less than the fundamental wave component of human speech.
means for providing an input signal which may contain human voice signals;
control means, having an input terminal, output terminal and control terminal, for receiving said input signal at said input terminal and variably providing said input signal at said output terminal in response to n control signal applied to said control terminal; and means coupled to receive said input signal and provide a control signal to said control terminal in response to the detection of frequencies in said input signal less than the fundamental wave component of human speech.
3. The system as claimed in claim 2, wherein said control means is a voltage-controlled variable gain amplifier which provides a variable gain in response to a voltage applied to the control terminal, said means for providing a control. signal providing a variable voltage control signal in response to said detection of frequencies.
4. The system as claimed in claim 2, wherein said control means is a voltage-controlled variable frequency filter which varies its frequency in response to a voltage applied to the control terminal, said means for providing a control signal providing a variable voltage control signal in response to said detection of frequencies.
5. The system as claimed in claim 2, wherein said control means is an analog switch which provides or inhibits said input signal at its output terminal in response to a voltage applied to its control terminal, said means for providing a control signal providing a variable voltage control signal in response to said detection of frequencies.
6. The system as claimed in claims 3, 4 or 5 further including first means for low pass filtering said input signal received by said means for providing, and second means for low pass filtering said input signal received by said input terminal, said first means having a cutoff for frequency lower than said second means.
7. A system for discriminating a human voice signal comprising:
means for providing an input signal which may contain a human voice signal;
input level deriving circuit means coupled to receive said input signal for rectifying said input signal and smoothing said rectified signal to provide a direct current output proportional to a level of said input signal;
a human voice fundamental wave component level deriving circuit means coupled to receive said input signal and provide an output in response to the detection of a fundemental wave component of the human voice in said input signal;
arithmetic circuit means coupled to receive said outputs from said input level deriving circuit means and said fundamental wave component level deriving circuit means for subtracting one from the other to produce a difference output;
control means having a control terminal and coupled to receive said input signal and provide a variable level output of said input signal in response to a control signal applied to said control terminal, said difference output being coupled as the control signal applied to said control terminal.
means for providing an input signal which may contain a human voice signal;
input level deriving circuit means coupled to receive said input signal for rectifying said input signal and smoothing said rectified signal to provide a direct current output proportional to a level of said input signal;
a human voice fundamental wave component level deriving circuit means coupled to receive said input signal and provide an output in response to the detection of a fundemental wave component of the human voice in said input signal;
arithmetic circuit means coupled to receive said outputs from said input level deriving circuit means and said fundamental wave component level deriving circuit means for subtracting one from the other to produce a difference output;
control means having a control terminal and coupled to receive said input signal and provide a variable level output of said input signal in response to a control signal applied to said control terminal, said difference output being coupled as the control signal applied to said control terminal.
8. The system as claimed in claim 7, wherein said control means is a voltage-controlled variable gain amplifier having a gain which varies in response to a voltage applied to said control terminal.
9. The system as claimed in claim 7, wherein said control means is a voltage-controlled variable frequency filter having a frequency which varies in response to a voltage applied to said control terminal.
10. The system as claimed in claim 7, wherein said control means is an analog switch which selectively provides or inhibits said control means output in response to a voltage applied to said control terminal.
11. The system as claimed in claims 8, 9 or 10 further comprising low pass filter means for receiving said input signal and providing a low pass filtered output, said low pass filtered output being coupled to said arithmetic circuit means and summed with the outputs producing said difference output.
12. A human voice signal discriminating system, comprising:
means for providing an input signal which may contain human voice signals;
control means having an input terminal, output terminal and control terminal, for receiving said input signal at said input terminal and variably providing said input signal at said output terminal in response to a control signal applied to said control terminal; and a low base sound deriving circuit means coupled to receive said input signal and provide a control signal to said control terminal in response to frequencies in said input signal which are less than that of the fundamental wave component of human speech in said input signal.
means for providing an input signal which may contain human voice signals;
control means having an input terminal, output terminal and control terminal, for receiving said input signal at said input terminal and variably providing said input signal at said output terminal in response to a control signal applied to said control terminal; and a low base sound deriving circuit means coupled to receive said input signal and provide a control signal to said control terminal in response to frequencies in said input signal which are less than that of the fundamental wave component of human speech in said input signal.
13. A human voice signal discriminating system comprising:
means for providing an input signal which may contain human voice signals;
means for low-pass filtering said input signal to provide a filtered input signal having frequencies in a range of 0-150 Hz;
control means having an input terminal, output terminal, and control terminal, for receiving said filtered input signal at said input terminal and variably providing said filtered input signal at said output terminal in response to a control signal applied to said control terminal;
and means for low-pass filtering said input signal to provide a control signal to said control terminal in response to frequency region below the lowest normal point of the human voice spectrum in said input signal to cause said control means to pass said frequencies in said range of 0-150 Hz to said output terminal in response to said control signal resulting from frequencies in said frequency region.
means for providing an input signal which may contain human voice signals;
means for low-pass filtering said input signal to provide a filtered input signal having frequencies in a range of 0-150 Hz;
control means having an input terminal, output terminal, and control terminal, for receiving said filtered input signal at said input terminal and variably providing said filtered input signal at said output terminal in response to a control signal applied to said control terminal;
and means for low-pass filtering said input signal to provide a control signal to said control terminal in response to frequency region below the lowest normal point of the human voice spectrum in said input signal to cause said control means to pass said frequencies in said range of 0-150 Hz to said output terminal in response to said control signal resulting from frequencies in said frequency region.
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP54290/1980 | 1980-04-21 | ||
| JP1980054290U JPS6137040Y2 (en) | 1980-04-21 | 1980-04-21 | |
| JP5268280A JPS56149097A (en) | 1980-04-21 | 1980-04-21 | Human voice signal identifying method |
| JP52682/1980 | 1980-04-21 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1184506A true CA1184506A (en) | 1985-03-26 |
Family
ID=26393319
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA000375651A Expired CA1184506A (en) | 1980-04-21 | 1981-04-16 | Method and system for discriminating human voice signal |
Country Status (5)
| Country | Link |
|---|---|
| CA (1) | CA1184506A (en) |
| DE (1) | DE3115801A1 (en) |
| FR (1) | FR2480978B1 (en) |
| GB (2) | GB2077078B (en) |
| IT (1) | IT1142404B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2547690B1 (en) * | 1983-06-20 | 1985-12-06 | Capel Raymond | ELECTRONIC DEVICE FOR CREATING A LIGHT ATMOSPHERE COMPARABLE TO SUNSET |
| EP2254349A3 (en) * | 2003-03-03 | 2014-08-13 | Phonak AG | Method for manufacturing acoustical devices and for reducing wind disturbances |
Family Cites Families (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3102236A (en) * | 1960-05-05 | 1963-08-27 | Collins Radio Co | Squelch circuit controlled by demodulated voice signal |
| US3377428A (en) * | 1960-12-29 | 1968-04-09 | Ibm | Voiced sound detector circuits and systems |
| US3336533A (en) * | 1964-03-13 | 1967-08-15 | Motorola Inc | Electronic circuit |
| GB1094357A (en) * | 1964-12-09 | 1967-12-13 | Mullard Ltd | Improvements in or relating to speech filters |
| GB1120541A (en) * | 1965-08-11 | 1968-07-17 | Dolby Ray Milton | Improvements in noise reduction systems |
| US3513260A (en) * | 1967-10-13 | 1970-05-19 | Ibm | Speech presence detector |
| FR2050782A5 (en) * | 1969-06-25 | 1971-04-02 | Cit Alcatel | |
| US3678416A (en) * | 1971-07-26 | 1972-07-18 | Richard S Burwen | Dynamic noise filter having means for varying cutoff point |
| US3798559A (en) * | 1971-04-20 | 1974-03-19 | Matsushita Electric Ind Co Ltd | Noise reduction system |
| JPS5154703A (en) * | 1974-11-08 | 1976-05-14 | Oki Electric Ind Co Ltd | |
| US4051331A (en) * | 1976-03-29 | 1977-09-27 | Brigham Young University | Speech coding hearing aid system utilizing formant frequency transformation |
| US4025721A (en) * | 1976-05-04 | 1977-05-24 | Biocommunications Research Corporation | Method of and means for adaptively filtering near-stationary noise from speech |
| GB1591869A (en) * | 1978-05-31 | 1981-06-24 | Pye Elecronic Products Ltd | Alternating signal detector |
-
1981
- 1981-04-16 CA CA000375651A patent/CA1184506A/en not_active Expired
- 1981-04-17 IT IT8148303A patent/IT1142404B/en active
- 1981-04-18 DE DE19813115801 patent/DE3115801A1/en active Granted
- 1981-04-21 FR FR8107876A patent/FR2480978B1/en not_active Expired
- 1981-04-21 GB GB8112276A patent/GB2077078B/en not_active Expired
-
1984
- 1984-02-09 GB GB08403418A patent/GB2143105B/en not_active Expired
Also Published As
| Publication number | Publication date |
|---|---|
| GB2077078A (en) | 1981-12-09 |
| IT1142404B (en) | 1986-10-08 |
| GB2143105A (en) | 1985-01-30 |
| GB8403418D0 (en) | 1984-03-14 |
| DE3115801A1 (en) | 1982-01-14 |
| FR2480978A1 (en) | 1981-10-23 |
| FR2480978B1 (en) | 1985-12-13 |
| GB2143105B (en) | 1985-07-24 |
| GB2077078B (en) | 1984-09-12 |
| DE3115801C2 (en) | 1987-09-03 |
| IT8148303A0 (en) | 1981-04-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US4698842A (en) | Audio processing system for restoring bass frequencies | |
| US4162462A (en) | Noise reduction system | |
| US4322641A (en) | Noise reduction system | |
| JPH0317491Y2 (en) | ||
| US4622692A (en) | Noise reduction system | |
| US4506379A (en) | Method and system for discriminating human voice signal | |
| US4627094A (en) | Electronic audio signal processor | |
| JPS5746306A (en) | Aural signal muting circuit of magnetic recorder and reproducer | |
| US3784749A (en) | Noise eliminating device | |
| US4164626A (en) | Pitch detector and method thereof | |
| CA1184506A (en) | Method and system for discriminating human voice signal | |
| US3668322A (en) | Dynamic presence equalizer | |
| US3600516A (en) | Voicing detection and pitch extraction system | |
| US5841934A (en) | Method for noise suppression and noise-suppression circuit for video recorders | |
| JPH0339839Y2 (en) | ||
| US3735275A (en) | Dynamic presence filter | |
| JPS56137722A (en) | Noise reduction circuit | |
| JPH0427559B2 (en) | ||
| JPS627224A (en) | Reception signal detecting system | |
| JPS6246348Y2 (en) | ||
| JPH0739090Y2 (en) | Sound quality improvement circuit | |
| JPS63234680A (en) | Reproducing device for television sound | |
| JPS6251522B2 (en) | ||
| JPS5593513A (en) | Recording device of pulse train signal | |
| JPS59219793A (en) | Fundamental extractor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |