Multi-feature speech/music discrimination system
2010-03-29 00:00:00
with representative pairs of histograms depicted in FIGS. 3-14. These figures pertain to a variety of different types of audio signals that were sampled at a rate of 22,050 samples per second and manually labelled as being speech or music. In the figures, the upper histogram of a pair depicts measured results for a number of samples of speech data, and the lower histogram depicts values for samples of music data. In all of the histograms, a log transformation is employed to provide a monotonic normalization of the values for the features. This normalization is preferred, since it has been found to improve the spread and conformity of the data over the applicable range of values. Thus, the x-axis values can be negative, for features in which the measured result is a fraction less than one, as well as positive. The y-axis represent the number of frames in which a given value was measured for that feature.

The histograms depicted in the figures are representative of the different results between speech and music that might be obtained for the respective features. In practice, actual results may vary, in dependence upon factors such as the size and makeup of the set of known samples that are used to derive training data, preprocessing of the signals that is used to generate spectrograms, and the like.

One of the features, depicted in FIGS. 3a and 3b, is the spectral centroid, which represents the balancing point of the spectral power distribution within a frame. Many types of music involve percussive sounds which, by including high-frequency noise, result in a higher spectral mean. In addition, excitation energies can be higher for music than for speech, in which pitch stays within a range of fairly low values. As a result, the spectral centroid for music is, on average, higher than that for speech, as depicted in FIG. 3b. In addition, the spectral centroid has higher values for unvoiced speech than it does for voiced speech. The spectral centroid for a frame occurring at time t is computed as follows ##EQU1##

where k is an index corresponding to a frequency, or small band of frequencies, within the overall measured spectrum, and Xt [k] is the power of the signal at the corresponding frequency band.

Another analysis feature, depicted in FIGS. 4a and 4b, is known as the spectral flux. This feature measures frame-to-frame spectral difference. Speech has a higher rate of change, and goes through more drastic frame-to-frame changes than music. As a result, the spectral flux value is higher for speech, particularly unvoiced speech, than it is for music. Also, speech alternates periods of transition, such as the boundaries between consonance and vowels, with periods of relative stasis, i.e. vowel sounds, whereas music typically has a more constant rate of change. Consequently, the spectral flux is highest at the transition between voiced and unvoiced sounds.

Another feature which is employed for speech/music discrimination is the zero-crossing rate, depicted in FIGS. 5a and 5b. This value is a measure of the number of time-domain zero- voltage crossings within a speech frame. In essence, the zero-crossing rate indicates the dominant frequency during the time period of the frame.

The next feature, depicted in FIGS. 6a and 6b, is the spectral roll-off point. This value measures the frequency below which 95% of the power in the spectrum resides. Music, due to percussive sounds, attack transients, and the like, has more energy in the high frequency ranges than speech. As a result, the spectral roll-off point exhibits higher values for music and unvoiced speech, and lower values for voiced speech. The spectral roll-off value for a frame is computed as follows:

SRt =K, where ##EQU2##

The next feature, depicted in FIGS. 7a and 7b, comprises the cepstrum resynthesis residual magnitude. The value for this feature is determined by first computing the cepstrum of the spectrogram by means of a Discrete Fourier Transform, as described for example in Bogert et al, The Frequency Analysis of Time Series for Echoes: Cepstrum, Pseudo-autocovariance, Cross-Cepstrum and Saphe Cracking, John Wiley and Sons, New York 1963, pp 209-243. The result is then smoothed over a time window, and the sound is resynthesized. The smooth spectrum is then compared to the original (unsmoothed) spectrum, to obtain an error value. A better fit between the two spectra is obtained for unvoiced speech than for voiced speech or music, due to the fact that unvoiced speech better fits a homomorphic single-source filter model than does music. In other words, the error value is higher for voiced speech and music. FIG. 7c illustrates an example of the difference between the smoothed and unsmoothed spectra for voiced speech. The cepstrum resynthesis residual magnitude is computed as follows: ##EQU3##

where Yt [k] is the resynthesized smoothed spectrum.

In addition to each...

Low profile keyboard device and system for recording and scoring music
2010-03-23 00:00:00
Vincent, teaches an electronic data storage system including a magnetic type recorder/replayer for recording spontaneous musical presentations for replay through a similar instrument. Tocapture the musical data, the invention also requires extensive and expensive modifications to the underside of each key in the instrument. See also U.S. Pat. No. 4,023,456, entitled "MUSIC ENCODING AND DECODING APPARATUS," to Groeschel, for yetanother example of how electronic switching to monitor keyboard action requires bulky circuitry and modification of the keyboard from within the instrument.

The sequencer is a viable alternative method of recording music which has been developed in the prior art, although early in its development, the sequencer was a massive network of electronics, often covering walls in a recording studio. Musicians are able to record and immediately play back music with the use of sequencers. A sequencer, in its simplest form, consists of a series of adjustable voltage memories stepped by a clock pulse. The typical analog sequencer uses potentiometersand variable resistors, each including a manually operable dial for establishing a certain DC voltage In order to load the sequencer, the musician manually sets each potentiometer. Thereafter, the bank of potentiometers is scanned sequentially and theDC voltages are read to a voltage controlled oscillator (VCO) which then produces the melody or the rhythm. The sequencer thus enables the musician to repeatedly listen to the melody and make changes by varying the potentiometer dials. Sequencers areused to create the familiar insistent machine-beat that has been used in electronic organs. See Keyboard Synthesizer Library, Vol. 3, Synthesizers and Computers, p. 37 (1985). While the sequencer produces the accompaniment, a musician can play the leadline of the same or another keyboard, or even another instrument.

With the advent of solid state electronics, smaller and more efficient electronics have been combined in the prior art to produce a digital sequencer. Typical digital sequencers utilize a Read/Write memory storing a plurality of words, each wordbeing coded to represent a note played on the keyboard. Once the memory has been coded, the sequencer can be used to play the keyboard instrument by reading back the data words in the memory in time sequence. See U.S. Pat. No. 3,890,871, entitled,"APPARATUS FOR STORING SEQUENCES OF MUSICAL TONES," to Oberheim; U.S. Pat. No. 4,160,399, entitled, "AUTOMATIC SEQUENCE GENERATOR FOR A POLYPHONIC TONE SYNTHESIZER," to Deutsch; and U.S. Pat. No. 4,487,101, entitled "DIGITAL SOLID STATE RECORDING OFTHE SIGNALS CHARACTERIZING THE PLAYING OF A MUSICAL INSTRUMENT," to Ellen. While providing an improved and efficient means of recording music, sequencers do not provide a written means of preserving music on musical score sheets. More importantly,however, sequencers require an electronic musical instrument and have not been adapted to conventional acoustic keyboard instruments, such as the piano.

The electronic music revolution has led to the invention of the synthesizer, an electronic musical instrument. Sequencers, as described above, have been incorporated into the synthesizer, so that while the musician plays music on a synthesizerkeyboard, sequencers within the synthesizer plays back various accompaniments that the musician loaded previously into the sequencer. The use of sequencers allows the musician to compose and record various tracks of music. The electronic instrumentsgenerate musical data consisting of a series of binary digits, called bits. A number of digits representing a complete musical expression, such as which note has been played and the particular style, is called a data word. The words are then stored ina memory unit which can store only a finite number of these binary data words. The length of the recorded music, therefore, is limited by the amount of memory in the solid state chips used in digital sequencers. Microprocessor technology provides themeans for storing lengthy sequences by transferring the digitized musical data stored in memory to peripheral devices such as computer diskettes. Examples of electronic musical instruments which incorporate microprocessor technology include the EnsoniqMirage鈩? various Korg polyphonic synthesizers, and the Casio CZ 101鈩?

The computer, especially the personal home computer, further revolutionized the electronic music industry with the creation of software capable of interpreting the notes played on the keyboard and printing the music in musical scored form. Themusic industry desired a communication standard to be used among the multitude of electronic music manufacturers and the multitude of available home computers. The standard decided upon was MIDI, an acronym for Musical Instrument Digital Interface. Inits simplest application, MIDI permits a musician to play two or more instruments from a single keyboard, in order to layer musical tone colors. In its most comprehensive application, MIDI provides the means for realizing a multi-track recorder or acomputer-based composing system by connecting several instruments to a master controller or computer. Computer software is available, furthermore, which can transform the music from digital format to a conventional musical score, both on the computerscreen and as printed out on paper in hard copy. Commercially available software which can convert MIDI data to scored music or to a format to be viewed on a computer terminal for editing purposes include the MIDI Performance Series鈩?by Passport,and the MPS鈩?wri...

Electronic music system and stringed instrument input device therefor
2010-03-02 00:00:00
a voltage controlled tone generator, or synthesizer, and an input device, in the form of a guitar or other fretted stringed instrument and associated electronic circuitry, for sequentially providing voltage signals, selected from a set of discretely different voltage levels each analogously related to a musical tone, for driving the tone generator. Each string-fret pair of the stringed instrument is assigned a given musical tone, preferably in accordance with normal tuning of the instrument, and means are provided for producing a corresponding voltage when a string-fret pair is closed by pressing the string against the fret. When two or more string-fret pairs are simultaneously closed, the output voltage corresponding to the highest frequency musical tone associated with the closed string-fret pairs is produced. In particular, different electrical voltages are applied to the instrument frets so as to apply such voltages to the strings when the strings are pressed into contact with the frets. A multiplexing system repetitively samples the string voltages, adds to each string voltage an offset voltage compensating for the musical intervals between the open strings, and processes the highest summed voltage for output to the tone generator.ClaimsWe claim:

1. An electronic music system comprising a voltage controlled tone generator, a stringed instrument having at least one string and a plurality of frets spaced from one another along thelength of said string with each string-fret pair representing an assigned musical tone, and means responsive to said string being pressed into contact with any one of said frets for producing and supplying to said voltage controlled tone generator, asthe driving input signal for said tone generator, a voltage signal having a voltage value analogously related to the frequency of the musical tone assigned to the contacting string-fret pair, said voltage controlled tone generator including means forproducing an intermediate signal having a frequency related to said input voltage signal, an amplifier having a voltage controlled gain for varying the amplitude of said intermediate signal, an envelope generator for providing a voltage waveformcontrolling the gain of said amplifier, and means for turning said envelope generator on to initiate the production of a new voltage waveform therefrom in response to said at least one string being brought into contact with any one of said frets.

2. An electronic music system comprising a voltage controlled tone generator, a stringed instrument having a plurality of spaced parallel strings located over a fret board having a plurality of frets extending transversely of said strings andspaced one from another along the length of said fret board with each string-fret pair representing an assigned musical tone, and means responsive of any one of said strings being pressed into contact with any one of said frets for producing andsupplying to said voltage controlled tone generator, as the driving input for said tone generator, a voltage signal having a voltage value analogously related to the frequency of the musical tone represented by the contacting string-fret pair.

3. A music system as defined in claim 2 further characterized by said voltage controlled tone generator including means for producing an intermediate signal having a frequency related to said input voltage signal, an amplifier having a voltagecontrolled gain for varying the amplitude of said intermediate signal, an envelope generator for producing a voltage waveform controlling the gain of said amplifier, and means for turning said envelope generator on to initiate the production of a newvoltage waveform therefrom in response to any one of said strings being brought into contact with any one of said frets.

4. A music system as defined in claim 3 further characterized by means for inhibiting the production of another voltage waveform from said envelope generator until after all of said strings are first out of contact with any of said frets.

5. An electronic music sy...

Generation of noise-like tones in an electronic musical instrument
2010-02-27 00:00:00
rateproportional to the pitch of the tone being generated, the clock means shifting said register, random address generating means for selectively transferring words from any one of a plurality of locations in the master data list memory to the shiftregister with each clock pulse, and means transferring the words in the shift register to said converter to convert said words to an analog voltage whose amplitude is controlled by the digital values of said words stored in the shift register.

8. A tone synthesizer comprising source means providing a group of words representing respectively the amplitudes of equally spaced points defining the waveform of a musical tone, digital-to-analog converter, means transferring said group ofwords in timed sequence from the source means and applying the words to the converter, and a random signal generator for generating timing pulses at random time intervals, said transferring means including means responsive to the timing pulses from saidrandom signal generator for modifying the values of those digital words transferred in time coincidence with the pulses from the random signal generator.DescriptionFIELD OF THE INVENTION

This invention relates to musical tone synthesizers, and more particularly, to a noise generator for a digital tone generator.

BACKGROUND OF THE INVENTION

The generation of musical tones electronically, either by analog or digital circuits, is well known. In attempting to duplicate the sounds of conventional musical instruments it may be desirable to superimpose sounds which can only becharacterized as "noise" onto the musical tones. Such added noise may be introduced to simulate the air noise, hiss, or breathiness characteristic of wind-operated instruments, such as the organ pipes of a conventional organ, or other types of windinstruments. In prior art digital type organs tones have been created imitative to noisy wind-blown organ pipes, by using a frequency modulation technique. This has been accomplished by adding or subtracting a fixed constant to the frequency numberused to address the tone data. Alternatively, the noise has been added to the reference voltage of the analog output signal from the digital-to-analog convert...

Electronic musical instrument with exponential keyboard and voltage controlled oscillator
2010-02-26 00:00:00
selects a control voltage, from an exponential voltage divider, for controlling the frequency of a voltage controlled oscillator, which produces a frequency which is directly proportional to the control voltage and inversely proportional to a reference voltage. The reference voltage compensates for variations in the level of the supply voltage, so that the oscillator frequency is independent of the supply voltage.ClaimsWhat is claimed is:

1. An electronic musical instrument having a voltage controlled oscillator for producing a sound signal having a frequency proportional to a control voltage applied to it, akeyboard having a plurality of keys, a plurality of switches, one for each of said keys, each adapted to be operated by depression of its associated key, and a voltage divider connected with said switches for connecting a control voltage to saidoscillator which corresponds to the position of the key associated with an operated one of said switches, said voltage divider comprising a plurality of resistance elements connected in series, each of said elements having different resistance valueswhich bear an exponential relation to the resistance values of the adjacent connected resistors such that the voltage at successive junctions of said resistance elements correspond to a geometric series, said resistance elements being formed of the samematerial and being physically located in close physical juxtaposition with each other, so that all said resistors are maintained at approximately the same temperature, with approximately constant relative resistances.

2. Apparatus according to claim 1 wherein said resistance elements are formed simultaneously as portions of a single integrated thick-film circuit.

3. In an electronic musical instrument having an electrical power supply, a voltage controlled oscillator for producing a sound signal having a frequency proportional to a control voltage applied to it, a keyboard having a plurality of keys, aplurality of switches, one for each of said keys, each adapted to be operated by depression of its associated key, and connecting means connected with said switches for connecting a control voltage to said oscillator which corresponds to the position ofthe key associated with an operated one of said switches, the combination comprising a reference voltage generator connected to said electrical power supply for producing a reference voltage, and means connecting said oscillator to said reference voltagegenerator, said reference voltage generator being adapted to produce a shift in the level of said reference voltage in response to a change in the level of voltage of said electrical power supply, said shift having a magnitude and direction tending tocompensate for said change in power supply voltage level, whereby said oscillator frequency is substantially independent of said change.

4. Apparatus according to claim 3, wherein said reference voltage generator comprises an inverter having an input connected with said power supply.

5. Apparatus according to claim 4, wherein said oscillator comprises an integrator for integrating a voltage derived from said voltage divider, a comparator connected to said integrator and operative to compare an output produced by saidintegrator with said reference voltage, and means connected with said comparator and operative upon a comparison of said integrator output and said reference voltage for resetting said integrator for a subsequent cycle of integration.

6. An electronic musical instrument having a voltage controlled oscillator for producing a sound signal having a frequency proportional to a control voltage applied to it, a keyboard having a plurality of keys, a plurality of switches, one foreach of said keys, each adapted to be operated by depression of its associated key, a voltage divider connected with said switches for connecting a control voltage to said oscillator which corresponds to the position of the key associated with anoperated one of said switches, said voltage divider comprising a plurality of resistance elements connected in series, each of said elements having resistance values which bear an exponential relation to the resistance values of adjacent connectedresistors su...

Electronic musical instrument with means for automatically generating chords and harmony
2010-02-05 00:00:00
120 are connected to the keyboard switches and the function switches, through diodes 121, as illustrated in FIG. 3. The keyboard 10 is represented schematically, and under each key of the keyboard 10 there is a keyboard switch 124which is actuated when its corresponding key is depressed. Each of the switches 124 is a normally open, single-pole, single-throw switch. One terminal of the switch is connected to one of the X lines 120, and the switches 124 which are associated withcorresponding keys of the keyboard 10 in different octaves are connected to the same X line 120. Thus, the common poles of all of the A keys are connected together, through diodes 121, to one of the X lines 120, the common poles of all of the B-flatkeys 124 are connected in common to a second of the X lines, etc.

The other pole of each of the switches 120 is connected to one of several Y lines 126. The Y lines 126 are connected to the switches 124 in accordance with their octave location relative to the keyboard. Thus, the first of the Y lines 120 isconnected in common to the second terminal of the first twelve switches 124 (for the highest octave), the next Y line is connected in common to the next twelve switches 124, etc. A total of sixty-four keys are provided in the keyboard 10, and there aresixty-four switches 124 connected in the manner partially illustrated in FIG. 3. The last of the Y lines 126 is connected to four switches 124 of the lower left-hand end of the keyboard 10 (not shown), and to eight function control switches 129 through136. The Y line is connected to the second pole of the switches 129 through 136, with the first terminal of the switches 129 through 136 connected to the last eight of the X lines 120.

In operation, one of the X lines is energized with a signal in the manner described above, in accordance with the state of the counter 108. This signal is then applied to all of the switches associated with keys for the same note name and, insome cases, to one of the function control switches 128-136. Each of these switches 124 which is closed completes a path from the energized X line to one of the Y lines 126. The Y line which is energized by any closed switch depends upon the positionof that switch in the keyboard. A switch in the highest octave will energize the first Y line, a switch in the second highest octave will energize the second Y line, etc. If the energized X line is connected to a closed function control switch, the lastof the Y lines is energized. For any given combination of X and Y lines there is one and only one switch which can complete a path, and this one switch may be either a keyboard switch 124 or a function control switch 129-136.

The Y lines 126 are connected to six inputs of a six-channel multiplexer 128 (FIG. 4a). The multiplexer 128 receives its control inputs from the output lines 116 of the counter 114, and functions to connect one and only one of the Y lines 126 toan output 130 of the multiplexer 128, in accordance with the state of the counter 114. The output line 130 thus contains, for each cycle of operation of the multiplexer 32, a train of pulses, including one pulse for each of the operated switches 124 and129-136. The pulses are encoded in time position representative of the switches which are operated. Hereinafter the seventy-two pulse times of each cycle will be referred to as pulse time 1 through pulse time 72. Pulses occurring during the firstsixty-four pulse times represent operated keys of the keyboard, while the last eight pulse times represent operated ones of the function switches.

Note Latches

The identification of the note corresponding to a pulse within the first sixty-four pulse times corresponds to the binary coded output on the lines 110 at the timme of occurrence of such pulse on the line 130. Similarly, the octave of such note(the octave of the keyboard in which its associated key is located) is identified by the binary code on the output lines 116. Thus, the simultaneous condition of the outputs 110 and 116 uniquely represents the specific switch which is being scanned atany instant. These seven lines, viz., the four lines 110 and the three lines 116, are connected to seven inputs of a group of A latches 132 and to seven inputs of a group of B latches 134. The A latches 132 are provided with a set input 136, and, whenthe set input is energized with a set pulse, the latches 132 are set in accordance with the signals then present on the lines 110 and 116. That is, a high level on one of the lines 110 and 116 at the set pulse time causes its respective latch to be setin one state, and a low level then causes it to be reset to the opposite state. A similar set terminal 138 is provided for the B latches. The set inputs 136 and 138 are energized with set pulses at specific times, as described more fully hereinafter.

Seven output lines 140 connected to the A latches continuously manifest voltage levels representative of the states of the individual latches 132. Similarly, seven output ines 142 manifest the states of the B latches 134. The lines 140 areconnected to seven inputs of a group of C latches 144, which is provided with a set input 136. When the set input 136 is energized, the C latches 144 are set in accordance with the signals on the lines 140. Similarly, the lines 142 are connected toseven inputs of a group of D latches 148, which has a set input terminal 150. When the set input terminal 150 is energized, the D latches 148 are set in accordance with the signals on the lines 142. The set input terminals 146 and 150 are energizedwith set pulses together by signals on a line 152, derived in a manner described more fully hereinafter.

When the C and D latches are set, outputs on lines 154 and 156 manifest the states of these latches. The four lines 154, which carry signals derived initially from the outputs of the counter 108, are connected to four inputs of a one-of-twelvedecoder 158, which functions to energize one of twelve output lines 160, in accordance with the binary representation of the signals on the lines 154. Similarly, the four of the lines 156, which also manifest information originally derived from thecounter 108, are connected to four inputs of a one-of-twelve decoder 162, which manifests an output on one of its twelve output lines 164, in accordance with the binary representation of its four inputs lines.