United States Patent 3743783

A spectrum analyzer having an input, a mixer for feeding the input with a filtered speech signal and a frequency converted signal which is a function of physiological activity associated with the speech.

Application Number:
Publication Date:
Filing Date:
Primary Class:
Other Classes:
International Classes:
A61B5/12; G10L15/00; (IPC1-7): G10L1/12
Field of Search:
179/1VS,1SB,1SA 324
View Patent Images:
US Patent References:
3383466Nonacoustic measures in automatic speech recognition1968-05-14Hillix
2492062Device for portrayal of complex waves1949-12-20Potter
2378383Transient signal recordingreproducing device1945-06-19Arndt
2181265Signaling system1939-11-28Dudley

Other References:

Dolansky, On Certain Irregularities of Voiced Speech Waveforms, IEEE Vol. AU-16 March 1968, p. 51-56. .
Farr, Phonovid, A System for Recording Television Pictures on Phonograph Records, JAES, April 1968. .
Kersta, Amplitude Cross-Section Representation with the Sound Spectrograph, JASA Nov. 1948..
Primary Examiner:
Claffy, Kathleen H.
Assistant Examiner:
Leaheey, Jon Bradford
What is claimed is

1. In apparatus for simultaneously recording sound spectra and physiological data, the combination comprising;

2. The apparatus according to claim 1, wherein;

3. The apparatus according to claim 2, wherein;

4. The apparatus according to claim 1, wherein;


The present invention relates to spectrum analyzers and, more particularly, to audio frequency spectrum analyzers which produce permanent graphic recordings of complex waves.

Various types of spectrum analyzers are known and widely used in the analysis and description of voices and other complex acoustic signals.

While the recorded outputs of these spectrum analyzers, in the form of spectrograms, have been a substantial aid to those engaged in phonetic research, heretofore, these analyzers have been limited to the reproduction of information derived only from voice or other sound inputs. Such information is usually in the form of an energy distribution in frequency and time. From this information researchers can make inferences about the physiological formation of the sounds that are produced. It is, however, desirable that apparatus be available that can not only record the complex voice signals but also the physiological event that may have occurred prior to, at the time of, or subsequent to the formation of the actual sound output.


The present invention relates to an improvement to existing audio frequency spectrum analyzers which permits the simultaneous recording of physiological events as well as the complex acoustic signal onto the conventional spectrogram recording medium.

Basically, the present invention provides, in combination with conventional spectrum analyzers, input structure which combines the acoustic or speech signal with the physiological activity signal in such a manner that switching the spectrum analyzer from wide-band analysis to narrow-band analysis at a predetermined frequency will yield the physiological information alongside the speech analysis information, whereby the two can be readily compared in substantially real time.

This is accomplished by the provision of, in combination with a spectrum analyzer of conventional construction having an input, means for developing a signal in response to sounds, means for filtering said signal and feeding the same to mixing means, transducer means for converting a physiological activity, such as intraoral pressure, to an electrical signal and means for converting said signal to a varying frequency signal and feeding the same to a second input of said mixing means, the output of which is connected to the input of said spectrum analyzer.


For a fuller understanding of the present invention reference should now be had to the following detailed description of the same taken in conjunction with the accompanying drawings wherein;

FIG. 1 is a schematic representation of the apparatus of the present invention with conventional components illustrated in block form;

FIG. 2 is a first spectrogram illustrating a typical result utilizing the apparatus of the present invention; and

FIG. 3 is a second spectrogram.


Referring now to the drawings and, more particularly, to FIG. 1, the improved spectrum analyzer of the present invention is illustrated in schematic block form and comprises, a spectrum analyzer 10, mixing means 12 and two input circuits connected to the mixing means; a speech input circuit depicted generally by the numeral 14 and a physiological activity input circuit depicted generally by the numeral 16.

For the spectrum analyzer 10 any suitable conventional analyzer may be utilized of the type wherein a selective receiver is cyclically tuned through the desired frequency band, and functions to resolve a complex audio signal into its frequency and amplitude components as a function of time. A typical spectrum analyzer for this purpose is the "Sona-Graph" Model 6061-A manufactured by the Kay Elemetrics Corp.

The mixing means 12 has an output line 18 which is fed to the input, such as the microphone input, of the spectrum analyzer 10. Any conventional signal mixing structure may be utilized, such as a Claricon Transistorized Microphone mixer.

The speech input circuit 14 may comprise a microphone 140, the output from which is fed to filter means 142 which feeds the filtered signal from the microphone to one input of the mixing means 12 via line 20. Any conventional microphone or sound responsive device may be utilized such as, for example, an Electro-Voice Dynamic Microphone Model No. 631. The filter may comprise any suitable low-pass filter for passing frequencies below a predetermined value, such as 4,500 Hertz, which allows frequencies above this value to be employed for the response of physiological signals. A typical filter is the Krohn Hite Model 310-C.

The physiological activity input circuit 16 comprises transducer means 160 for converting a physiological activity, such as intraoral pressure, into an electrical signal, means 162 for amplifying the output of the transducer means and frequency conversion means 164 for developing a varying frequency signal which is a function of the amplified transducer means output signal. The output from the frequency conversion means is fed via line 22 to the second input of the mixing means 12. A suitable power supply 166, such as an Offner Type 392 power supply, is provided for circuit 16.

Transducer means 160 may comprise a Statham PM6OTC Differential pressure transducer; means 162 may comprise an Offner Type 492 data amplifier and the frequency conversion means 164 may comprise an Anadex Model DF 100 DC to Frequency converter. In addition to utilizing the sine wave normally delivered by the frequency converter, the output thereof may be modified to deliver a saw-tooth wave.

In operation, assuming it is desired to correlate the intraoral pressure variations in the formation of certain spoken words with the acoustic sounds themselves, a suitable pressure sensor such as a polyethelene tube may be placed the speaker's mouth to transmit the pressure variations to transducer 160, whereas the spoken sounds may be transmitted through microphone 140 to filter 142.

The varying intraoral pressure signal is converted by transducer 160 into a varying d.c. signal which is amplified by 162 and converted into a frequency varying signal by converter 164, the baseline frequency of which signal may be typically set at 4,700 Hertz.

The speech signal is filtered by low-pass filter 142 to frequencies below a predetermined value, such as 4,500 Hertz, and is mixed with the physiological activity signal from converter 164 in mixer 12 and then fed to the input of the spectrum analyzer 10.

With the spectrum analyzer set for wide-band analysis, the lower trace pattern of the speech signal is generated as a spectrogram on conventional spectrograph paper and is depicted in FIGS. 2 and 3 at S, which is expressed as the frequency distribution of the speech pattern as a function of time. Since conventional spectrum analyzers of the type mentioned herein typically utilize a wideband analysis of 300 Hz. for speech analysis, it is advisable to manually switch to a narrow band analysis to obtain a fine line trace of the physiological activity. Such switching can obviously be achieved by the built-in conventional band width control of typical spectrum analyzers. The expansion was set at 80 to 6,000 hertz. When the stylus of the spectrum analyzer reached the 4,500 hertz vertical position on the spectrographic paper, the expansion scale was turned off and switched from wide-band to narrow-band analysis, resulting in the upper trace shown in FIGS. 2 and 3 and depicted at P. This trace records the physiological activity expressed as intraoral pressure as a function of time. Alternatively, no switching is essential in the spectrum analyzer if the analysis begins at the narrow band width setting. The two signals will still be depicted as shown in FIGS. 2 and 3 with the exception that the harmonics of the lower trace will be resolved.

The FIG. 2 display is for the formation of the sound syllable /pa/ whereas the FIG. 3 display is for the formation of the sound syllable /ap/.

With the apparatus of the present invention many different kinds of displays of the speech spectrum and physiological activities can be obtained, depending upon the versatility of the user of the spectrum analyzer. The baseline for the pressure, or other physiological activity, recordings can be adjusted within the total range of the spectrum analyzer. The sensitivity of frequency change as a function of intraoral pressure variations is adjustable.

The present invention enables the graphic displays of sound spectrum to be obtained with those of any physiological activity in real time in a simple and inexpensive manner.

Although a preferred embodiment has been disclosed, many changes will occur to those skilled in this field. For example, additional outputs from the frequency converter 164 will permit multi-channel operation for a comparison of a variety of physiological events together with sound spectra. Such events, in addition to the intraoral pressure can comprise an electrocardiographic pattern, which can be recorded simultaneously with the acoustic pattern of the heartbeat. Depending upon the type of sensor or transducer most any response can be recorded on to the spectrogram.

It is therefore intended that the scope of the present invention is to be limited only by the scope of the appended claims.