CN1170925A - Karaoke singer rating apparatus and method, and storage medium therefor - Google Patents

Karaoke singer rating apparatus and method, and storage medium therefor Download PDF

Info

Publication number
CN1170925A
CN1170925A CN97112227A CN97112227A CN1170925A CN 1170925 A CN1170925 A CN 1170925A CN 97112227 A CN97112227 A CN 97112227A CN 97112227 A CN97112227 A CN 97112227A CN 1170925 A CN1170925 A CN 1170925A
Authority
CN
China
Prior art keywords
voice signal
standard
singer
performance
accompaniment
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.)
Granted
Application number
CN97112227A
Other languages
Chinese (zh)
Other versions
CN1165890C (en
Inventor
铃木俊彦
大圣寺健
山本裕介
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yamaha Corp
Original Assignee
Yamaha Corp
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Yamaha Corp filed Critical Yamaha Corp
Publication of CN1170925A publication Critical patent/CN1170925A/en
Application granted granted Critical
Publication of CN1165890C publication Critical patent/CN1165890C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H1/00—Details of electrophonic musical instruments
    • G10H1/36—Accompaniment arrangements
    • G10H1/361—Recording/reproducing of accompaniment for use with an external source, e.g. karaoke systems
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H2210/00—Aspects or methods of musical processing having intrinsic musical character, i.e. involving musical theory or musical parameters or relying on musical knowledge, as applied in electrophonic musical tools or instruments
    • G10H2210/031—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal
    • G10H2210/066—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal for pitch analysis as part of wider processing for musical purposes, e.g. transcription, musical performance evaluation; Pitch recognition, e.g. in polyphonic sounds; Estimation or use of missing fundamental
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H2210/00—Aspects or methods of musical processing having intrinsic musical character, i.e. involving musical theory or musical parameters or relying on musical knowledge, as applied in electrophonic musical tools or instruments
    • G10H2210/031—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal
    • G10H2210/076—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal for extraction of timing, tempo; Beat detection
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H2210/00—Aspects or methods of musical processing having intrinsic musical character, i.e. involving musical theory or musical parameters or relying on musical knowledge, as applied in electrophonic musical tools or instruments
    • G10H2210/031—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal
    • G10H2210/091—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal for performance evaluation, i.e. judging, grading or scoring the musical qualities or faithfulness of a performance, e.g. with respect to pitch, tempo or other timings of a reference performance

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Reverberation, Karaoke And Other Acoustics (AREA)
  • Electrophonic Musical Instruments (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The present invention provides a device and a method for the evaluation of Karaoke singers, as well as a storage medium storing the executable program of a computer. Wavelet transformation is performed on the singing representing a standard singer and the standard sound signals of a first accompaniment, the singing representing the singing ability of a singer waiting to be evaluated and the sound-evaluating signals of a second accompaniment similar to the first accompaniment. On each same frequency band, the consistency between the frequency component of the standard sound signals and each corresponding frequency component of the sound-evaluating signals is determined. According to the determined consistency, the singing evaluation value of the singer whose singing ability needs to be evaluated is determined.

Description

Karaoke singer rating apparatus and method and medium thereof
The present invention relates to a kind of karaoke singer rating apparatus and method that is used to estimate the karaoke person's who sings with the accompaniment of a karaoke OK system performance ability, and medium.
Traditionally, in this type of designed karaoke singer rating apparatus, the representative of one signal extraction of the data reproduction that will from recording medium, store such as CD (LD) and compact-disc (CD) the standard voice signal sung of one standard under the accompaniment of karaoke OK system, sing karaoke person's the voice signal of microphone input of performance of reality of a first song in pitch (interval) with having represented by one, volume, make comparisons in the aspect such as shift gears of these parameters, this singer's performance ability just obtains estimating from the consistent degree between the signal that compares mutually at these two quilts.Extract in the voice signal that extracts this standard voice signal, standard and microphone input from then on during pitch and other handle, legacy system has used a digital signal processor (DSP)
More specifically, this traditional karaoke singer rating apparatus extracts the standard voice signal according to a preset program or data from a digital audio signal that offers DSP, and extracts the pitch of this standard voice signal through extracting and pitch of microphone input audio signal or the like.
Yet, this traditional karaoke singer rating apparatus comprises program and the complicated structure thereof that a large amount of whole thus algorithms are carried out, because will carry out the pitch that extracts this voice signal, extract this voice signal, detect its level or the like work in order to extract an audio signal characteristics, and, all to carry out these programs to each standard voice signal and microphone input audio signal.
In addition, this tradition Karaoke evaluating apparatus also has following inconvenience: when the extraction of voice signal is enough inaccurate, just can not accurately extract the pitch of standard voice signal.
An object of the present invention is to provide a kind of simple in structure and can be with near karaoke singer rating apparatus and the method for estimating a singer's performance ability according to people's the artificial evaluation method of feeling to carry out, and medium.
For achieving the above object, according to a first aspect of the invention, provide a kind of karaoke singer rating apparatus, comprising: first converting means is used for the performance and the first standard voice signal of accompanying of representing the standard singer are carried out wavelet transformation; Second converting means is used for the performance of representing its performance ability singer to be evaluated is carried out wavelet transformation with the second evaluation voice signal of accompanying identical with first accompaniment; And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance at it.
The wavelet transformation of sound or voice signal provides the inferior band signal with following frequency characteristic: along with frequency band is high more, it is wide more that the bandwidth of each inferior band signal also becomes, and therefore just can easily determine it according to its sound level.As a result, just can save the algorithm of the pitch that is used to extract sound or speech signal, therefore the layout of this device is also oversimplified.In addition, the mode of the bandwidth of division sound or speech signal is consistent with the human sense of hearing, and this just feasible evaluation to the singer proves effective in the anthropoid mode of class.
For achieving the above object, according to a second aspect of the invention, provide a kind of karaoke singer rating apparatus, comprising: first converting means is used for the standard voice signals execution wavelet transformation to the performance of representing the standard singer; Second converting means is used for the evaluation voice signal execution wavelet transformation to the performance of representing its performance ability singer to be evaluated; And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance at it.
For achieving the above object, according to a third aspect of the present invention, a kind of karaoke singer rating apparatus is provided, comprise: regenerating unit, be used for reading the standard voice signal and an accompaniment of representative standard singer's performance, with this standard voice signal of regenerating from a recording medium that has wherein write down the standard voice signal; First converting means is used for the standard voice signal of this regeneration is carried out wavelet transformation; The standard voice signals removal device is used for and will represents the standard voice signals of this standard singer's performance to remove only the voice signal of representing this accompaniment is taken out thus from the standard voice signal of having regenerated; Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation speech signal; Device sound mixing is used for the voice signal of the just representative accompaniment that will be taken out by the standard voice signals removal device and carries out audio mixing and estimate voice signal to form one by singing evaluation voice signal that sound conversion device obtains; The second wavelet transformation device is used for carrying out wavelet transformation to estimating voice signal; And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance at it.
For achieving the above object, according to a fourth aspect of the present invention, a kind of karaoke singer rating apparatus is provided, comprise: regenerating unit, be used for reading the standard voice signal and an accompaniment of representative standard singer's performance, with this standard voice signal of regenerating from a recording medium that has wherein write down the standard voice signal; Standard.The voice signal removal device is used to receive the standard voice signal after the regeneration, and is used for representing the standard voice signals of this standard singer's performance to remove with thus only the voice signal taking-up of represent this accompaniment from paid-in standard voice signal; The standard voice signals withdrawing device is used for deducting the only voice signal of representative accompaniment from the standard voice signal of having regenerated, to take out standard voice signals thus; First converting means is used for the standard voice signals of taking out is carried out wavelet transformation; Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal; The second wavelet transformation device is used for the evaluation voice signal that is obtained is carried out wavelet transformation; And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance at it.
For achieving the above object, according to a fifth aspect of the present invention, a kind of karaoke singer rating apparatus is provided, comprise: regenerating unit, be used for from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the signal of representative accompaniment only; First converting means is used for the standard voice signal of regeneration is carried out wavelet transformation; Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal; Estimate voice signal and form device, the voice signal of the only representative accompaniment that is used for to receive and add together from the evaluation voice signal of singing the sound conversion and to estimate voice signal to form one thus; The second wavelet transformation device is used for carrying out wavelet transformation to estimating voice signal; And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance at it.
For achieving the above object, according to a sixth aspect of the invention, a kind of karaoke singer rating apparatus is provided, comprise: regenerating unit, be used for from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the voice signal of representative accompaniment only; The standard voice signals withdrawing device is used for deducting the only voice signal of representative accompaniment from the standard voice signal of having regenerated, to take out standard voice signals thus; First converting means is used for the standard voice signals of taking out is carried out wavelet transformation; Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal; The second wavelet transformation device is used for carrying out wavelet transformation to estimating voice signal; And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance at it.
Preferably, wavelet transformation is an orthogonal wavelet transformation.
For achieving the above object, according to a seventh aspect of the present invention, provide a kind of karaoke singer rating method, may further comprise the steps: a) performance and the first standard voice signal of accompanying of representing the standard singer are carried out wavelet transformation; B) performance of representing its performance ability singer to be evaluated is carried out wavelet transformation with the second evaluation voice signal of accompanying identical with first accompaniment; C) determine the frequency component of the standard voice signal that wavelet transformation obtains separately on the frequency band corresponding and estimate consistance between each self-corresponding frequency component of voice signal at it, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance.
For achieving the above object, according to an eighth aspect of the present invention, provide a kind of karaoke singer rating method, may further comprise the steps: a) standard voice signals of the performance of representing the standard singer is carried out wavelet transformation; B) the evaluation voice signal of the performance of representing its performance ability singer to be evaluated is carried out wavelet transformation; C), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
For achieving the above object, according to a ninth aspect of the present invention, a kind of karaoke singer rating method is provided, may further comprise the steps: a) read the standard voice signal and an accompaniment of representative standard singer's performance, with this standard voice signal of regenerating from a recording medium that has wherein write down the standard voice signal; B) the standard voice signal of this regeneration is carried out wavelet transformation; C) will represent the standard voice signals of this standard singer's performance from the standard voice signal of having regenerated, to remove only the voice signal of representing this accompaniment is taken out thus; D) receive its performance ability singer's to be evaluated performance sound, this is sung sound change one into and estimate speech signal; E) voice signal of the just representative accompaniment that will take out by the standard voice signals removal device and carry out audio mixing and estimate voice signal to form one by singing evaluation voice signal that sound conversion device obtains; F) carry out wavelet transformation to estimating voice signal; G), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
For achieving the above object, according to a tenth aspect of the present invention, a kind of karaoke singer rating method is provided, may further comprise the steps: a) read the standard voice signal and an accompaniment of representative standard singer's performance, with this standard voice signal of regenerating from a recording medium that has wherein write down the standard voice signal; B) receive standard voice signal after the regeneration, and be used for represent the standard voice signals of this standard singer's performance to remove with thus only the voice signal taking-up of representing this accompaniment from paid-in standard voice signal; C) from the standard voice signal of having regenerated, deduct the only voice signal of representative accompaniment, to take out standard voice signals thus; D) standard voice signals of taking out is carried out the small form conversion; E) receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal; F) the evaluation voice signal that is obtained is carried out wavelet transformation; G), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
For achieving the above object, according to an eleventh aspect of the present invention, a kind of karaoke singer rating method is provided, may further comprise the steps: a) from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the signal of representative accompaniment only; B) the standard voice signal of regeneration is carried out wavelet transformation; C) receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal; The voice signal of the only representative accompaniment that d) will receive and add together from the evaluation voice signal of singing the sound conversion and to estimate voice signal to form one thus; E) carry out wavelet transformation to estimating voice signal; F), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
For achieving the above object, according to a twelfth aspect of the present invention, a kind of karaoke singer rating method is provided, may further comprise the steps: a) from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the voice signal of representative accompaniment only; B) from the standard voice signal of having regenerated, deduct the only voice signal of representative accompaniment, to take out standard voice signals thus; C) standard voice signals of taking out is carried out wavelet transformation; D) receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal; E) the evaluation voice signal that is obtained is carried out wavelet transformation; F), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
For achieving the above object,, provide respectively the medium stored programme of carrying out by a computing machine accordingly with the method for the 7th to the 12 aspect of the present invention according to others of the present invention.
By the following detailed that interrelates with accompanying drawing, above and other purpose of the present invention, characteristics and advantage will become more apparent.
Fig. 1 is the block scheme of layout of the OK system of the principle ground explanation karaoke singer rating apparatus of realizing first embodiment according to the invention;
Fig. 2 is a block scheme, has shown the structure of the sound blanking piece shown in Fig. 1;
Fig. 3 is a block scheme, has shown the structure of one of two wavelet transformation pieces shown in Fig. 1;
Fig. 4 is the figure of an output signal frequency characteristic of the wavelet transformation piece shown in the displayed map 3;
Fig. 5 shows that one estimates the block scheme of the structure of piece;
Fig. 6 is the block scheme that shows the layout of karaoke singer rating apparatus according to a second embodiment of the present invention;
Fig. 7 is the block scheme of layout that shows the karaoke singer rating apparatus of a third embodiment in accordance with the invention;
Fig. 8 is the block scheme of layout that has shown the karaoke singer rating apparatus of a fourth embodiment in accordance with the invention.
Now, will be explained in detail the present invention with reference to the accompanying drawing that shows the embodiment of the invention.
At first with reference to figure 1, it has shown the karaoke OK system of realizing the karaoke singer rating apparatus of first embodiment according to the invention, and this system comprises according to a karaoke singer rating apparatus 1A and a Karaoke processing block 2 by first embodiment.
The karaoke singer rating apparatus 1A of this karaoke OK system can receive with digital form and be recorded in standard voice signal (V on compact-disc (CD) the 3 neutral body tracks r+ A), this signal is by the standard voice signals V of the performance of representing a standard singer rWith represent a voice signal A of singer's accompaniment for this reason.Standard voice signal number (V from CD3 regeneration r+ A) left channel signals is input to L channel numeral entry terminal 4a, and its right-channel signals 12 is imported into R channel digital terminal 4b.
This karaoke singer rating apparatus 1A comprises: the first mixer circuit 5a and 6, two of sound blanking pieces all are connected to two entry terminal 4a, 4b; Analog output terminal 8 is by its evaluation voice signal V from its performance ability of representative of microphone 7 inputs one analog form singer's to be evaluated performance mA/D converter 9 is used for the evaluation voice signal V of analog form that will be by analog output terminal 8 input mBe converted to a digital signal; The second mixer circuit 5b links to each other with A/D converter 9 with sound blanking piece 6; The first wavelet transformation piece 10a links to each other with the first mixer circuit 5a; The second wavelet transformation piece 10b links to each other with the first mixer circuit 5b; Estimate piece 11, link to each other, in order to the value of determining according to the output of the first and second wavelet transformation piece 10a, 10b to estimate with the first and second wavelet transformation piece 10a, 10b; And display unit 12, such as CRT monitor, link to each other with evaluation piece 11, estimate the evaluation of estimate that piece 11 is determined in order to show.
Karaoke processing block 2 comprises: Echo transfer circuit 20 is used for Echo is sent to the output of A/D converter 9; Karaoke signal Processing piece 21, it is provided the output of sound blanking piece 6; The 3rd mixer circuit 22 is used for the output of Echo transfer circuit 20 and the output audio mixing of Karaoke signal Processing piece 21; And D/A converter 23, being used for changing the output of the 3rd mixer circuit 22 into a simulating signal, this simulating signal is sent to an audio system (not shown).
The first mixer circuit 5a of karaoke singer rating apparatus 1A will be from the standard voice signal (V of CD3 regeneration r+ A) a left side and right-channel signals L, R carries out audio mixing, and synthetic signal is reached the first wavelet transformation piece 10a.
Fig. 2 has shown the structure of the sound blanking piece 6 of karaoke singer rating apparatus 1A.
Provide left channel signals L and right-channel signals R by entry terminal 4a, 4b to this sound blanking piece 6, sound blanking piece 6 is with standard voice signals V rFrom standard voice signal (V r+ A) the voice signal A of accompaniment is only represented in middle cancellation with recovery.Sound blanking piece 6 comprises: each is provided from Hi-pass filter (HPF) 60a, low-pass filter (LPF) 61a and the multiplier 62a of the left channel signals L of CD3 regeneration; Each is provided from Hi-pass filter (HPF) 60b, low-pass filter (LPF) 61b and the multiplier 62b of the right-channel signals R of CD3 regeneration; Totalizer 63 is used for the output of multiplier 62a, 62b is added together; Left channel signals totalizer 64a is used for the output for the output of the H PF60a of left channel signals L and LPF61a and totalizer 63 added and is in the same place; And right-channel signals totalizer 64b, be used for output with the output of the HPF60b of right-channel signals R and LPF61b and totalizer 63 and add and be in the same place.
Hi-pass filter (HPF) 60a, 60b allow to pass through from it from the left channel signals L of CD3 regeneration and the high fdrequency component of right-channel signals R, and this high fdrequency component for example is higher than 7KHZ.Low-pass filter 9 (LPF) 61a, 61b allow to pass through from it from the left channel signals L of CD3 regeneration and the low frequency component of right-channel signals R, and this low frequency component for example is lower than 100HZ.The frequency characteristic of these wave filters is not limited to above-mentioned, but they will be set to any value that voice (voice) signal is filtered as far as possible effectively.
Multiplier 62a will multiply by coefficient+1 from the left channel signals L of CD3 regeneration and multiplier 62b multiply by coefficient-1 with right-channel signals R.
In the sound blanking piece 6 of structure like this, totalizer 63 will multiply by the left channel signals L of coefficient+1 at multiplier 62a and add with the right-channel signals R that multiply by coefficient-1 at multiplier 62b and be in the same place, and final sum delivered to totalizer 64a and 64b, eliminate the standard voice signals V at the center amount place of sound map between a left side and right-channel signals thus rComponent, and only represented the voice signal A that accompanies in the language frequency wave band to be removed and be sent to next stage.
The second mixer circuit 5b will be from the voice signal A of the digital form of the representative of sound blanking piece 6 accompaniment with from the evaluation voice signal V of the digital form of A/D converter 9 mCarry out audio mixing and estimate voice signal (V to form m+ A), and this is estimated voice signal reach the second wavelet transformation piece 10b.
Fig. 3 has shown the structure of one of the first and second wavelet transformation piece 10a, 10b.Because the structure of the first and second wavelet transformation piece 10a, 10b is identical, one of them demonstration and explanation have then been represented the two.
Term " wavelet transformation " is defined as on mathematics according to by the following one group of signal transformation that basis function carries out that forms: (1) expansion or compression, and (2) move the basis function (female small echo) that is positioned on time or the frequency.Specifically, when by expansion or compression and to move this group basis function that this basis function forms whole when mutually orthogonal, this conversion then is called as " a kind of orthogonal wavelet transformation ".For realizing that with hardware this wavelet transformation, wave filter can be used to realize moving of (2), recurrence and repeated use have the wave filter of same algorithm then can realize the expansion or the compression of (1).
Wavelet transformation piece 10 shown in Figure 3 (10a, 10b) has been realized this wavelet transformation.More specifically, wavelet transformation piece 10 by as two-part sub-band wave filter 100a, the 100b of elementary cell ..., 100g forms, and is used to finish orthogonal wavelet transformation.These sub-band wave filters recursively connect each other so that the wave filter 100 of second or following stages is linked to the low frequency signal outgoing side of its previous stage wave filter, frequency band to the input signal of each wave filter is divided into a high-frequency wavelet section and a lower frequency wavelet section thus, this lower frequency wavelet section quilt again is sampled to a high frequency filter section and a lower frequency wavelet section, so that export a plurality of frequency sub-band signal W1 to W7 and E7 at last.Forming of each two-part sub-band wave filter 100 by mutual sub-sampling filter 101 of a completely orthogonal high pass and the sub-sampling filter 102 of a low pass.Wave filter 101 and 102 can be the linear filter (QMF) that satisfies orthogonality condition basically.Fig. 3 has shown the layout of this kind of example wavelet transformation piece 10, and wherein the bandwidth of an original input signal finally is divided into 8 frequency sub-band by 7 two parts sub-band wave filter 100a to 100g that use serial connection.First to the 7th grade of sub-sampling filter of high pass 101a to 101g delivers to high communication number (sub-band signals) 101 to 107 and estimates piece 11, and at the sub-sampling filter 102 of the 7th grade low pass low-pass signal (sub-band signals) E7 delivered to evaluation piece 11.In this orthogonal wavelet transformation, can recursively utilize identical structure (algorithm), can make that thus structure (algorithm) is quite simple.
Fig. 4 has shown the output signal frequency feature of wavelet transformation piece 10.
As shown in the figure, from first to the 7th grade of sub-sampling filter of high pass 101a to 101g passes in the frequency band separately that main frequency component that the sub-band signals W1 to W7 that comes had drops on 2KHZ or higher, 1KHZ to 2KHZ, 500HZ to 1KHZ, 250HZ to 500HZ, 125HZ to 250HZ, 62HZ to 125HZ and 31HZ to 62HZ.The sub-band signals E7 that sends here from the 7th grade of sub-decimation filter 102g of low pass has the main frequency component that is lower than 31HZ.The output signal frequency of first and second wavelet transformation piece 10a, the 10b that is obtained by orthogonal wavelet transformation is so so that each output signal has a plurality of frequency spans of being determined by sub-band signals in other words, the frequency band of sub-band signals is high more, and the bandwidth of each sub-band signals is just wide more.
As described below, in the present embodiment, consistent degree between standard signal and the microphone input signal is to determine in each identical frequency band at it, and above the output signal frequency characteristic of the first and second wavelet transformation piece 10a, the 10b of description " bandwidth of each sub-band signals broadens with the rising of this sub-band signals on frequency " is consistent with the interval frequency characteristic of note signal.In other words, comparing not the just comparison between frequency between the same frequency band that is obtained by orthogonal wavelet transformation, but be equivalent to compare between the interval of music, this characteristic with the human auditory is consistent.
Fig. 5 shows the structure of estimating piece 11.Estimate piece 11 and determining degree consistent between two outputs on each same frequency band of the output of the first and second wavelet transformation piece 10a, 10b, and determine the evaluation of estimate that this system sings according to the consistent degree of so determining.In other words, estimate piece 11 and comprise: absolute value transducer (ABS) 110a to 110h is used to detect each sub-band signals W1 to W7 that sends here from the first wavelet transformation piece 10a and the absolute value of E7; Absolute value transducer (ABS) 111a to 111h is used to detect from each corresponding sub-band signals W1 to W7 of second wavelet transformation piece 10b transmission and the absolute value of E7; Smoothing circuit 112a is used for the output of level and smooth each ABS110a to 110h to 112h; Smoothing circuit 113a to 113h is used for the output of level and smooth each ABS111a to 111h; Consistency detector 114a to 114h is used for determining whether each sub-band signals from the first wavelet transformation piece 10a is with consistent with each other on corresponding same frequency band from each corresponding sub-band signals of the second wavelet transformation piece 10b; Counter (CNT) 115a to 115h is used for increasing when each consistency detector 114a to 114h detects consistent between above-mentioned sub-band signals; Normalization circuit 116a to 116h is used for the output normalization with each CNT115a to 115h; Randomizer 200 is used to produce any random number in 0 to 1 scope; Multiplier 117a to 117h is used for separately coefficient a to h is multiply by in the output of normalization circuit 116a to 116h; And multiplier 117i, be used for parameter i is multiply by in the output of randomizer 200; Totalizer 118 is used for the output addition with multiplier 117a to 117i; Reach counter (CNT) 119, be used to count the number of times K of executed comparison.Estimating piece 111 also comprises: increment/decrement circuit 201 is used for according to determining an increment/decrement value from totalizer 118 is that send here with x; Reach totalizer 202, be used for the output y of sigma-delta/decrement circuit 201.The value that is added up by totalizer 202 is used as the final evaluation of estimate that this karaoke person sings and is presented on the display unit 12.
Each normalization circuit 116 according to the number of times K of the comparison that provides by CNT119 will a corresponding CNT115 output usefulness coefficient 1/K normalization.
Parameter a can be provided with or changes to i and h with a parameter setting piece (not shown), if itself and equal 1, the criterion of Ping Jiaing just can be changed thus.
Below, with the work of describing as the karaoke singer rating apparatus 1A of above-mentioned structure.
At first, will be from entry terminal 4a, 4b and 8 to the first and second wavelet transformation piece 10a, the 10b etc. that are used to handle the standard voice signal (Vr+A) that inputs to it, the work of each parts part of this karaoke singer rating apparatus 1A is described.
At first, will import karaoke singer rating apparatus 1A so far from two sound channel signals of the standard voice signal (Vr+A) of the digital form of CD3 regeneration by entry terminal 4a and 4b.In addition, the evaluation voice signal Vm of the analog form that will be produced by its performance ability singer's to be evaluated actual performance by analog output terminal 8 imports karaoke singer rating apparatus 1A so far with microphone 7.
To carry out audio mixing through the left channel signals L and the right-channel signals R of the standard voice signal (Vr+A) of entry terminal 4a and 4b input by the first mixer circuit 5a, the signal (Vr+A) after mixing is input to the first wavelet transformation piece 10a.The first wavelet transformation piece 10a carries out wavelet transformation to standard voice signal (Vr+A) and provides sub-band signals W1 to W7 and E7 to give evaluation piece 11 thus.
Voice blanking piece 6 with standard voice signals Vr from through two sound channel signals of the standard voice signal (Vr+A) of entry terminal 4a and 4b input, removing the voice signal that provides only representative to accompany to the second mixer circuit 5b thus.On the other hand, the evaluation voice signal Vm through analog output terminal 8 inputs is converted to a digital signal by A/D converter 9.This digital signal is provided for the second mixer circuit 5b.The second mixer circuit 5b will only represent the voice signal A of accompaniment and carry out audio mixing from the next evaluation voice signal Vm of A/D converter, and the evaluation voice signal (Vm+A) that will synthesize is delivered to the second wavelet transformation piece 10b.The second wavelet transformation piece 10b estimates voice signal (Vm+A) and carries out wavelet transformation to this, and final sub-band signals W1 to W7 and E7 are delivered to evaluation piece 11.
Below, will work that estimate piece 11 and back be described.
Absolute value transducer (ABS) 110a to 110h that sub-band signals W1 to W7 from the first wavelet transformation piece 10a and E7 are estimated piece 11 is transformed into each signal of the absolute value of expression sub-band signals, and the final smoothed respectively circuit 112a to 112h of absolute value signal carries out smoothly.Delivered to consistency detector 114a to 114h respectively through level and smooth signal.
In addition, absolute value transducer (ABS) 111a to 111h that sub-band signals W1 to W7 from the second wavelet transformation piece 10b and E7 are estimated piece 11 is transformed into each signal of the absolute value of expression sub-band signals, and the final smoothed respectively circuit 113a to 113h of absolute value signal carries out smoothly.Level and smooth signal is delivered to consistency detector 114a to 114h respectively.Similar to the smooth signal that comes from the first wavelet transformation piece 10a.
Consistency detector 114a to 114h determines on its identical separately frequency band whether the signal that comes from the first wavelet transformation piece 10a is consistent with each other with each corresponding signal that comes from the second wavelet transformation piece 10b.When among the consistency detector 114a to 114h corresponding one on a same frequency band, detect between signal consistently, a corresponding number of times that consistance between signal is taken place increases by 1 among the counter on this same frequency band (CNT) 115a to 115h.According to number of comparisons K by counter 119 counting, the counting of the renewal of each CNT115 by among the normalization circuit 116a to 116h corresponding one carry out normalization by coefficient 1/K.Value after normalization hypothesis for example, is a value in 0 to 1, and this value is sent among the multiplier 117a to 117h corresponding one, multiply by among the parameter a to h corresponding one at this.Each finant product of coming out is sent to totalizer 118 from multiplier 117a to 117h.
On the other hand, any random number in 0 to 1 scope that is produced by randomizer 200 is multiplied by parameter i in multiplier 117i, and final product is sent to totalizer 118, and is similar to each the output among the multiplier 117a to 117h.The random number that is produced by randomizer 200 is in order to add the anthropoid factor of class for the end value of estimating, thus, compare with the traditional karaoke singer rating apparatus that mechanically provides evaluation, the evaluation that karaoke singer rating apparatus of the present invention is done more approaches the human evaluation of being done.In addition, at its value i with equal 1 (a+b+c ... under+i=1) the situation, parameter a to h and i can arbitrarily change, and thus, the standard of evaluation just can be changed.
The product value of multiplier 117a to 117h and multiplier 117i is added together by totalizer 118, and final i and x (suppose for example be 0 to 1 a value) are sent to increment/decrement circuit 201.Increment/decrement circuit 201 has a plurality of conversion tables or calculation element, thus, has determined the increment/decrement value corresponding to this and X, also this value is delivered to totalizer 202.Can select a plurality of conversion tables of increment/decrement circuit 201 or calculation element to be used to and to be scaled to the increment/decrement ratio of the increment/decrement value among the increment/decrement value y with X, and this make that also changing evaluation criterion becomes possibility with change.
The increment/decrement value y that sends here from increment/decrement circuit 201 adds a predefined initial value (for example 50 minutes) by totalizer 202, and the value of this " initial value+increment/decrement value y " is presented on the display unit 12 as final evaluation result.The evaluation of estimate of delivering to display unit 12 does not need to be limited on the value of a final appraisal results value, and can show the nonce of a changing evaluation yet.This makes this singer can know given interim scoring during it is sung.
According to the karaoke singer rating apparatus 1A of first embodiment, on its identical separately frequency band, determine from the result of (forming sub-band signals) orthogonal wavelet transformation of the standard voice signal (Vr+A) of CD3 regeneration and the consistance between the corresponding results separately from (the formation sub-band signals) orthogonal wavelet transformation of the evaluation voice signal (Vm+A) of microphone 7 inputs.Therefore, can save the algorithm that extracts pitch, the layout of this device is also just simple.And orthogonal wavelet transformation might be finished this evaluation in the mode of the evaluation done according to people's the sense of hearing near the people.
Now, will the second embodiment of the present invention be described with reference to figure 6.
In Fig. 6, shown karaoke singer rating apparatus 1B according to second embodiment, it is similar to first embodiment, comprising: L channel numeral entry terminal 4a, R channel numeral entry terminal 4b, the first and second mixer circuit 5a, 5b, sound blanking piece 6, analog output terminal 8, A/D converter 9, the first wavelet transformation piece 10a, the second wavelet transformation piece 10b, evaluation piece 11 and display unit 12.Yet this embodiment is different from the connection that the first embodiment part is said elements.
More specifically, this karaoke singer rating apparatus 1B comprises: totalizer 13, being used for output signal with the second mixer circuit 5b of the output signal of the first mixer circuit 5a and opposite in sign adds and is in the same place, and the output of totalizer 13 is input to the first wavelet transformation piece 10a, and the output of A/D converter 9 is input to the second wavelet transformation piece 10b.
Below, description is had the work of karaoke singer rating apparatus 1B of second embodiment of said structure.In being described below, the overlapping operation identical with the description of being done with reference to first embodiment will repeat no more.
At first, similar with first embodiment, by digital entry terminal 4a, 4b input standard voice signal (Vr+A), and by analog output terminal 8 input Simulation evaluation voice signal Vm.
Through the left channel signals L of standard voice signal (Vr+A) of entry terminal 4a, 4b input and right-channel signals R by the first mixer circuit 5a audio mixing.Simultaneously, sound blanking piece 6 is from removal standard voice signals Vr through the standard voice signal (Vr+A) of entry terminal 4a, 4b input, and only the voice signal A of representative accompaniment is provided for the second mixer circuit 5b.A left side and the right-channel signals component audio mixing of the voice signal A of the only representative accompaniment that the second mixer circuit 5b will be provided by sound blanking piece 6, and send the composite signal A of this sign-inverted to totalizer 13.Standard voice signal (Vr+A) that totalizer 13 will be come from the first mixer circuit 5a and the counter-rotating output signal A of the second mixer circuit 5b add together, and final standard voice signals Vr is offered the first wavelet transformation piece 10a.The first wavelet transformation piece 10a carries out wavelet transformation to send final sub-band signals W1 to W7 and E7 to evaluation piece 11 to standard voice signals Vr.
On the other hand, converted to a digital signal through the evaluation voice signal Vm of analog output terminal and input by A/D converter 9, thereupon, final digital signal is provided for the second wavelet transformation piece 10b.The second wavelet transformation piece 10b estimates speech signal Vm to this and carries out wavelet transformation, to send final sub-band signals W1 to W7 and E7 to evaluation piece 11.
Estimating piece 11 determines from the sub-band signals W1 to W7 of the first wavelet transformation piece 10a and E7 with from each the corresponding signal W1 to W7 of the 2nd wavelet transformation piece 10b and the consistance between the E7, also determine an evaluation of estimate of this performance according to fixed consistance number, this evaluation of estimate is displayed on the display unit 12.
According to the karaoke singer rating apparatus 1B of second embodiment, be determined in its identical separately frequency range from the result of the orthogonal wavelet transformation of the standard voice signals Vr of CD3 regeneration with from the consistance between each accordingly result of the evaluation voice signal Vm of microphone 7 inputs.Therefore can obtain identical effect basically with first embodiment acquisition.
Fig. 7 has shown the layout of the karaoke singer rating apparatus of a third embodiment in accordance with the invention.
Karaoke singer rating apparatus 1C according to the 3rd embodiment comprises: L channel analog output terminal 15a and R channel analog output terminal 15b, import by them respectively from a left side and the R channel simulating signal of CD (LD) 14 regeneration; A/D converter 16a, being used for the analog-signal transitions from L channel analog output terminal 15a is a digital signal; Analog output terminal 8, from the evaluation voice signal Vm of language tube 7 by its input; Totalizer 17, the evaluation voice signal V that is used for the analog form of self simulation entry terminal 8 in the future adds with right-channel signals from the analog form of R channel analog output terminal 15b and is in the same place; A/D converter 16b, being used for the analog signal conversion from totalizer 17 is digital signal; The first wavelet transformation piece 10a, 16a links with A/D converter; The second wavelet transformation piece 10b, 16b links with A/D converter; And estimate piece 11 and display unit 12, similar among these two and first embodiment.
CD (LD) 15 is stored the standard voice signal (Vr+A) according to the accompaniment on the analogue audio frequency channel of a standard singer performance and on the left side, and the voice signal A that only represents accompaniment on another simulation half frequency channel.
Below, description is had the work of the karaoke singer rating apparatus 1C of said structure.
At first, the standard voice signal (Vr+A) of analog form is through L channel analog output terminal 15a input, and only the voice signal A of the analog form of representative accompaniment imports through L channel entry terminal 15a.In addition, import through analog output terminal 8 from microphone 7 by the actual evaluation voice signal Vm that produces that sings of its performance ability singer to be evaluated.
Standard voice signal (Vr+A) through the analog form of L channel analog output terminal input converts digital signal to by A/D converter 17a, and also that this is final digital signal offers the first wavelet transformation piece 10a.The first wavelet transformation piece 10a carries out wavelet transformation to transmit final sub-band signals W1 to W7 and E7 to standard voice signal (Vr+A), and these signals are sent to estimates piece 11.
On the other hand, through only having represented of R channel analog output terminal 15b input accompaniment voice signal A and through the evaluation voice signal Vm of analog output terminal 8 inputs by totalizer 17 additions, and synthetic signal is as estimating voice signal (Vm+A).A/D converter 16b estimates voice signal (Vm+A) with this and converts digital signal to, and final digital signal is provided for the second wavelet transformation piece 10b.The second wavelet transformation piece 10b carries out wavelet transformation so that final sub-band signals W1 to W7 and E7 are transmitted evaluation piece 11 to estimating voice signal (Vm+A).Estimate piece 11 determine from the sub-band signals W1 to W7 of the first wavelet transformation piece 10a and E7 with from the second wavelet transformation piece 10b separately accordingly signal W1 to the consistance between W7 and the E7, and determine the evaluation of this performance according to fixed consistance number, this evaluation is displayed on the display unit 12.
According to the karaoke singer rating apparatus 1C of the 3rd above-mentioned embodiment, determining on its corresponding same frequency band from the result of the orthogonal wavelet transformation of the standard voice signal (Vr+A) of LD14 regeneration with from the consistance between each accordingly result of the wavelet transformation of the evaluation voice signal (Vm+A) of microphone 7 inputs.Therefore, can obtain the identical effect that obtains with first embodiment basically.
Fig. 8 has shown the layout of the karaoke singer rating apparatus of a fourth embodiment in accordance with the invention.
Similar with the 3rd embodiment, comprise according to the karaoke singer rating apparatus 1D of the 4th embodiment: L channel analog output terminal 15a, R channel analog output terminal 15b, A/ D converter 16a, 16b, analog output terminal 8, the first and second wavelet transformation piece 10a, 10b, evaluation piece 11 and display unit 12.Yet this embodiment is different from the connection that is above-mentioned each parts of the 3rd embodiment.
More specifically, karaoke singer rating apparatus 1D comprises: totalizer 18, being used for will be through the standard voice signal (Vr+A) of L channel analog output terminal 15a input and the voice signal A addition of importing through R channel analog output terminal 15b of only representing the contrary sign of accompaniment, and totalizer 18 links to each other with the first wavelet transformation piece 10a through A/D converter 16a, and analog output terminal 8 is connected to the second wavelet transformation piece 10b through A/D converter 16b.
Below, description is had the work of evaluating apparatus 1D in the Karaoke of the 4th embodiment of said structure.
At first, similar with the 3rd embodiment, standard voice signal (Vr+A) is through L channel analog output terminal 15a input, and only the voice signal of the analog form of representative accompaniment is imported through R channel analog output terminal 15b.And the evaluation voice signal Vm of analog form is through analog output terminal 8 inputs.
Totalizer 18 is through the standard voice signal (Vr+A) of L channel analog output terminal 15a input with through the voice signal A addition of the only representative accompaniment of the contrary sign of R channel analog output terminal 15b input, and composite signal is offered A/D converter as standard voice signals Vr.A/D converter 16a is converted to digital signal with standard voice signals Vf, and final digital signal is offered the first wavelet transformation piece 10a.The first wavelet transformation piece 10a carries out wavelet transformation to transmit final sub-band signals W1 to W7 and E7 to the standard voice signals Vr of digital form, and these signals are provided for estimates piece 11.
On the other hand, be converted to digital signal through the evaluation voice signal Vm of analog output terminal 8 inputs by A/D converter 16b, and final digital signal is sent to the second wavelet transformation piece 10b.The second wavelet transformation piece 10b carries out wavelet transformation to transmit final sub-band signals W1 to W7 and E7 to the evaluation voice signal Vm of digital form, and these signals are provided for estimates piece 11.
Estimating piece 11 determines from the sub-band signals W1 to W7 of the first wavelet transformation piece 10a and E7 and the consistance between signal W1 to W7 and the E7 accordingly separately from the second wavelet transformation piece 10b, and determine the evaluation of this performance according to fixed consistance number, this evaluation is displayed on the display unit 12.
According to the evaluating apparatus 1D of this Karaoke, determining on its corresponding same frequency band from the result of the orthogonal wavelet transformation of the standard voice signals Vr of LD14 regeneration with from the consistance between each accordingly result of the wavelet transformation of the evaluation voice signal Vm of microphone 7 inputs.Therefore, can obtain the identical effect that obtains with first embodiment basically.
The invention is not restricted to above-mentioned first to fourth embodiment, but and accomplished in various ways.
For example, the wavelet hop count that is used for wavelet transformation is not limited to 8, and can be set as greater or lesser number.
Though in above-mentioned first to fourth embodiment, the present invention is realized by the karaoke singer rating apparatus as special-purpose hardware, is not limited to this, but also available general purpose equipment or device realization the present invention such as personal computer.In addition, not only easily realize the present invention with hardware also available software.
Needless to say, realize the storage that can provide to a system or a device of the present invention be used for carrying out the function that realizes any one the foregoing description software program medium and carry out the computing machine (CPU or MPU) of the program that is stored in medium.
In this case, the program itself of reading from this medium has realized invention function of the present invention, and the medium of storing this program has been formed the present invention.
Conducts such as available floppy disk, hard disk, CD, magneto-optic disk, CD-ROM, CD-R, tape, Nonvolatile memory card, ROM are used to store the medium of this program.Can provide this program from a MI DI equipment or a server computer with the mode of communication network.
Self-evident, the present invention not only comprises the computer implemented situation of above-mentioned functions by execution fetch program from memory storage, also comprise a kind of situation:, carry out the part or all of practical programs of the functional requirement that realizes any one the foregoing description at an OS who calculates hands-operation according to the instruction of this program.
In addition, from needless to say, the present invention includes following situation: the program that reads from memory storage is written into and is included in an expansion board or the storer in the expanding element that links to each other of computing machine therewith that inserts the computing machine, according to the instruction of this program, be included in the needed part or all of practical programs of function of any one the foregoing descriptions of execution realization such as CPU in this expansion board or this expanding element then.

Claims (21)

1. karaoke singer rating apparatus comprises:
Comprise: first converting means is used for the performance and the first standard voice signal of accompanying of representing the standard singer are carried out wavelet transformation;
Second converting means is used for the performance of representing its performance ability singer to be evaluated is carried out wavelet transformation with the second evaluation voice signal of accompanying identical with first accompaniment;
And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance at it.
2. karaoke singer rating apparatus comprises:
First converting means is used for the standard voice signals execution wavelet transformation to the performance of representing the standard singer;
Second converting means is used for the evaluation voice signal execution wavelet transformation to the performance of representing its performance ability singer to be evaluated;
And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance at it.
3. karaoke singer rating apparatus comprises:
Regenerating unit is used for reading from the recording medium that has wherein write down the standard voice signal standard voice signal and an accompaniment of representative standard singer's performance, with the described standard voice signal of regenerating;
First converting means is used for the standard voice signal of described regeneration is carried out wavelet transformation;
The standard voice signals removal device is used for and will represents the standard voice signals of this standard singer's performance to remove to represent the voice signal of this accompaniment to take out by described from the standard voice signal of having regenerated;
Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing described performance sound into an evaluation speech signal;
Device sound mixing is used for the voice signal of the just representative accompaniment that will be taken out by the standard voice signals removal device and carries out audio mixing and estimate voice signal to form one by singing evaluation voice signal that sound conversion device obtains;
The second wavelet transformation device is used for carrying out wavelet transformation to estimating voice signal;
And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance at it.
4. karaoke singer rating apparatus comprises:
Regenerating unit is used for reading from a recording medium that has wherein write down the standard voice signal standard voice signal and an accompaniment of representative standard singer's performance, with the described standard voice signal of regenerating;
The standard voice signals removal device is used to receive the standard voice signal after the regeneration, and is used for representing the standard voice signals of this standard singer's performance to remove the voice signal taking-up of represent this accompaniment with by described from the standard voice signal of receiving;
The standard voice signals withdrawing device is used for deducting the only voice signal of representative accompaniment from the standard voice signal of having regenerated, to take out described standard voice signals;
First converting means is used for the standard voice signals of taking out is carried out wavelet transformation;
Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing described performance sound into an evaluation voice signal;
The second wavelet transformation device is used for the evaluation voice signal that is obtained is carried out wavelet transformation;
And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance at it.
5. karaoke singer rating apparatus comprises:
Regenerating unit, be used for from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the signal of representative accompaniment only;
First converting means is used for the standard voice signal of regeneration is carried out wavelet transformation;
Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing described performance sound into an evaluation voice signal;
Estimate voice signal and form device, the voice signal of the only representative accompaniment that is used for to receive and add together from the evaluation voice signal of singing the sound conversion and to estimate voice signal to form one;
The second wavelet transformation device is used for carrying out wavelet transformation to estimating voice signal;
And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance at it.
6. karaoke singer rating apparatus comprises:
Regenerating unit, be used for from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the voice signal of representative accompaniment only;
The standard voice signals withdrawing device is used for deducting the only voice signal of representative accompaniment from the standard voice signal of having regenerated, to take out standard voice signals;
First converting means is used for the standard voice signals of taking out is carried out wavelet transformation;
Sing sound conversion device, be used to receive its performance ability singer's to be evaluated performance sound, and be used for changing described performance sound into an evaluation voice signal;
The second wavelet transformation device is used for the evaluation voice signal that is obtained is carried out wavelet transformation;
And evaluating apparatus, be used for determining consistance between the corresponding separately output of the output of first converting means and second converting means separately on the frequency band corresponding, and be used for determining the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance at it.
7. according to any one a kind of karaoke singer rating apparatus in the claim 1 to 6, wherein said wavelet transformation is an orthogonal wavelet transformation.
8. a karaoke singer rating method comprises the following steps:
A) performance and the first standard voice signal of accompanying of representing the standard singer are carried out wavelet transformation;
B) performance of representing its performance ability singer to be evaluated is carried out wavelet transformation with the second evaluation voice signal of accompanying identical with first accompaniment; And
C) determine the frequency component of the standard voice signal that wavelet transformation obtains separately on the frequency band corresponding and estimate consistance between each self-corresponding frequency component of voice signal at it, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance.
9. a karaoke singer rating method comprises the following steps:
A) standard voice signals of the performance of representing the standard singer is carried out wavelet transformation;
B) the evaluation speech signal of the performance of representing its performance ability singer to be evaluated is carried out wavelet transformation; And
C), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
10. a karaoke singer rating method comprises the following steps:
A) read the standard voice signal and an accompaniment of representative standard singer's performance from a recording medium that has wherein write down the standard voice signal, with this standard voice signal of regenerating;
B) the standard voice signal of this regeneration is carried out wavelet transformation;
C) will represent the standard voice signals of this standard singer's performance from the standard voice signal of having regenerated, to remove only the voice signal of representing this accompaniment is taken out thus;
D) receive its performance ability singer's to be evaluated performance sound, this is sung sound change one into and estimate speech signal;
E) voice signal of the just representative accompaniment that will take out by the standard voice signals removal device and carry out audio mixing and estimate voice signal to form one by singing evaluation voice signal that sound conversion device obtains;
F) carry out wavelet transformation to estimating voice signal;
G), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
11. a karaoke singer rating method comprises the following steps:
A) read the standard voice signal and an accompaniment of representative standard singer's performance from a recording medium that has wherein write down the standard voice signal, with this standard voice signal of regenerating;
B) receive standard voice signal after the regeneration, and be used for represent the standard voice signals of this standard singer's performance to remove with thus only the voice signal taking-up of representing this accompaniment from paid-in standard voice signal;
C) from the standard voice signal of having regenerated, deduct the only voice signal of representative accompaniment, to take out standard voice signals thus;
D) standard voice signals of taking out is carried out the small form conversion;
E) receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal;
F) the evaluation voice signal that is obtained is carried out wavelet transformation;
G), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
12. a karaoke singer rating method comprises the following steps:
A) from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the signal of representative accompaniment only;
B) the standard voice signal of regeneration is carried out wavelet transformation;
C) receive its performance ability singer's to be evaluated performance sound, and be used for changing this performance sound into an evaluation voice signal;
The voice signal of the only representative accompaniment that d) will receive and add together from the evaluation voice signal of singing the sound conversion and to estimate voice signal to form one thus;
E) carry out wavelet transformation to estimating voice signal;
F), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this consistance of determining in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
13. a karaoke singer rating method comprises the following steps:
A) from standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the voice signal of representative accompaniment only;
B) from the standard voice signal of having regenerated, deduct the only voice signal of representative accompaniment, to take out standard voice signals;
C) standard voice signals of taking out is carried out wavelet transformation;
D) receive its performance ability singer's to be evaluated performance sound, and be used for changing described performance sound into an evaluation voice signal;
E) the evaluation voice signal that is obtained is carried out wavelet transformation;
F), and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance in the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal.
14. any one a kind of karaoke singer rating method in 13 according to Claim 8, wherein said wavelet transformation is an orthogonal wavelet transformation.
15. the medium of the program that a storage is carried out by computing machine, described program comprises:
Be used for the standard voice signal of the performance of representing the standard singer and first accompaniment is carried out a module of wavelet transformation;
Be used for the evaluation voice signal of the performance of representing its performance ability singer to be evaluated and second accompaniment identical with first accompaniment is carried out the module of wavelet transformation;
Be used for determining the frequency component of the standard voice signal that wavelet transformation obtains separately on the frequency band corresponding and estimating consistance between each self-corresponding frequency component of voice signal, and determine the module of evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described fixed consistance at it.
16. the medium of the program that a storage is carried out by computing machine, described program comprises:
Be used for the standard voice signals of the performance of representing the standard singer is carried out the module of wavelet transformation;
Be used for the evaluation speech signal of the performance of representing its performance ability singer to be evaluated is carried out the module of wavelet transformation;
One module, be used in the frequency component of its voice signal that settles the standard on frequency band corresponding separately and estimate consistance between each self-corresponding frequency component of voice signal, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance.
17. the medium of the program that a storage is carried out by computing machine, described program comprises:
Be used for reading the standard voice signal of representative standard singer's performance and accompaniment so that the module of the described standard voice signal of regenerating from the medium that wherein stores described standard voice signal;
Be used for the described standard voice signal of regeneration is done the module of wavelet transformation;
Be used for removing expression standard singer's the standard voice signals of performance with the module of the voice signal that takes out the described accompaniment of representative from the described standard voice signal of regeneration;
Be used to receive its ability singer's to be evaluated singing voice and convert thereof into the module of estimating voice signal;
Be used for the voice signal of only representing accompaniment that described standard voice signals removal device is taken out and the evaluation voice signal audio mixing that described performance sound conversion device obtains so that form the module of estimating voice signal; And
One module, be used for the consistance between each self-corresponding frequency component of the frequency component of its voice signal that settles the standard on frequency band corresponding separately and scoring voice signal, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance.
18. the medium of the program that a storage is carried out by a computing machine, described program comprises:
Read the standard voice signal and an accompaniment of representative standard singer's performance from a recording medium that has wherein write down the standard voice signal, with the module of this standard voice signal of regenerating;
Receive the standard voice signal after the regeneration, and be used for represent the standard voice signals of this standard singer's performance to remove with thus only the module of the voice signal taking-up of representing this accompaniment from paid-in standard voice signal;
From the standard voice signal of having regenerated, deduct the only voice signal of representative accompaniment, to take out the module of standard voice signals thus;
The standard voice signals of taking out is carried out the module of small form conversion;
Receive its performance ability singer's to be evaluated performance sound, and be used for this is sung the module that sound changes an evaluation voice signal into;
The evaluation voice signal that is obtained is carried out the module of wavelet transformation;
One module, be used in the frequency component of its voice signal that settles the standard on frequency band corresponding separately and estimate consistance between each self-corresponding frequency component of voice signal, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance.
19. the medium of the program that a storage is carried out by computing machine, described program comprises:
From standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the module of the signal of representative accompaniment only;
The standard voice signal of regeneration is carried out the module of wavelet transformation;
Receive its performance ability singer's to be evaluated performance sound, and be used for this is sung the module that sound changes an evaluation voice signal into;
Add together to form a module of estimating voice signal thus with the voice signal of the only representative accompaniment of receiving with from the evaluation voice signal of singing the sound conversion;
Carry out the module of wavelet transformation to estimating voice signal;
One module, be used in the frequency component of its voice signal that settles the standard on frequency band corresponding separately and estimate consistance between each self-corresponding frequency component of voice signal, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance.
20. the medium of the program that a storage is carried out by computing machine, described program comprises:
From standard voice signal wherein be recorded on the respective carter and only the voice signal of the representative accompaniment respective carter that is recorded in the recording medium on another respective carter read the standard voice signal of representative standard singer's performance and accompaniment and the voice signal of representative accompaniment only, with the standards for recycling voice signal and the module of the voice signal of representative accompaniment only;
From the standard voice signal of having regenerated, deduct the only voice signal of representative accompaniment, to take out the module of standard voice signals thus;
The standard voice signals of taking out is carried out the module of small form conversion;
Receive its performance ability singer's to be evaluated performance sound, and be used for this is sung the module that sound changes an evaluation voice signal into;
The evaluation voice signal that is obtained is carried out the module of wavelet transformation;
In the frequency component of its voice signal that settles the standard on frequency band corresponding separately with estimate consistance between each self-corresponding frequency component of voice signal, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to this fixed consistance.
One module, be used for the frequency component of its voice signal that settles the standard on frequency band corresponding separately and estimate consistance between each self-corresponding frequency component of voice signal, and determine the evaluation of estimate of the singer's that its performance ability is to be evaluated performance according to described definite consistance.
21. according to any one a kind of medium in the claim 15 to 20, wherein said wavelet transformation is an orthogonal wavelet transformation.
CNB97112227XA 1996-07-12 1997-07-14 Karaoke singer rating apparatus and method, and storage medium therefor Expired - Fee Related CN1165890C (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP183375/96 1996-07-12
JP183375/1996 1996-07-12
JP18337596A JP3707135B2 (en) 1996-07-12 1996-07-12 Karaoke scoring device

Publications (2)

Publication Number Publication Date
CN1170925A true CN1170925A (en) 1998-01-21
CN1165890C CN1165890C (en) 2004-09-08

Family

ID=16134676

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB97112227XA Expired - Fee Related CN1165890C (en) 1996-07-12 1997-07-14 Karaoke singer rating apparatus and method, and storage medium therefor

Country Status (4)

Country Link
JP (1) JP3707135B2 (en)
KR (1) KR100517887B1 (en)
CN (1) CN1165890C (en)
TW (1) TW333641B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100359559C (en) * 2004-04-01 2008-01-02 凌阳科技股份有限公司 Automatic sound-source scoring method and device
CN1763840B (en) * 2004-10-19 2010-06-02 凌阳科技股份有限公司 Singing grading method and system
CN107103912A (en) * 2017-04-24 2017-08-29 行知技术有限公司 A kind of student for imparting knowledge to students and checking and rating sings performance points-scoring system

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6225547B1 (en) 1998-10-30 2001-05-01 Konami Co., Ltd. Rhythm game apparatus, rhythm game method, computer-readable storage medium and instrumental device
JP4091892B2 (en) * 2003-09-30 2008-05-28 ヤマハ株式会社 Singing voice evaluation device, karaoke scoring device and program thereof
CN112331227B (en) * 2020-09-17 2022-05-24 武汉大学 Method for simulating human voice by laser

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS62134695A (en) * 1985-12-09 1987-06-17 松下電器産業株式会社 Scorer
JP2864827B2 (en) * 1991-12-27 1999-03-08 日本電気株式会社 Wavelet converter
KR0127218Y1 (en) * 1992-08-13 1998-10-15 강진구 Score evaluation displaying apparatus of karaoke system
KR950009692A (en) * 1993-09-28 1995-04-24 이헌조 Automatic scoring system of laser disc player (LDP)
KR950012422A (en) * 1993-10-08 1995-05-16 배순훈 Voice automatic control circuit of karaoke system
KR970010947B1 (en) * 1993-11-09 1997-07-02 대우전자 주식회사 Apparatus on scoring point in karaoke
KR0154991B1 (en) * 1993-11-15 1998-11-16 김광호 Voice state display method and apparatus of music player
KR0136740B1 (en) * 1993-12-27 1998-05-15 구자홍 Method of score examination in karaoke system

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100359559C (en) * 2004-04-01 2008-01-02 凌阳科技股份有限公司 Automatic sound-source scoring method and device
CN1763840B (en) * 2004-10-19 2010-06-02 凌阳科技股份有限公司 Singing grading method and system
CN107103912A (en) * 2017-04-24 2017-08-29 行知技术有限公司 A kind of student for imparting knowledge to students and checking and rating sings performance points-scoring system

Also Published As

Publication number Publication date
JP3707135B2 (en) 2005-10-19
CN1165890C (en) 2004-09-08
KR100517887B1 (en) 2005-12-01
KR980011000A (en) 1998-04-30
JPH1026994A (en) 1998-01-27
TW333641B (en) 1998-06-11

Similar Documents

Publication Publication Date Title
Pampalk Islands of music: Analysis, organization, and visualization of music archives
JP7243052B2 (en) Audio extraction device, audio playback device, audio extraction method, audio playback method, machine learning method and program
CN103559876B (en) Sound effect treatment method and system
CN1111382A (en) Performance elavuator for use in a karaoke apparatus
DE102012103553A1 (en) AUDIO SYSTEM AND METHOD FOR USING ADAPTIVE INTELLIGENCE TO DISTINCT THE INFORMATION CONTENT OF AUDIOSIGNALS IN CONSUMER AUDIO AND TO CONTROL A SIGNAL PROCESSING FUNCTION
JP2009510658A (en) Method and apparatus for processing audio for playback
CN110459196A (en) A kind of method, apparatus and system adjusting singing songs difficulty
JP2004530153A (en) Method and apparatus for characterizing a signal and method and apparatus for generating an index signal
CN1151077A (en) Method for reproducing audio signals and apparatus therefor
CN101667422A (en) Method and device for adjusting mode of song accompaniment
CN1717716A (en) Musical composition data creation device and method
CN108257609A (en) The modified method of audio content and its intelligent apparatus
JP2002215195A (en) Music signal processor
CN1105998C (en) Karaoke marking apparatus
CN105321526A (en) Audio processing method and electronic device
US6673995B2 (en) Musical signal processing apparatus
CN1165890C (en) Karaoke singer rating apparatus and method, and storage medium therefor
CN101930732B (en) Music producing method and device based on user input voice and intelligent terminal
CN101116135A (en) Sound synthesis
JP3776196B2 (en) Audio signal encoding method and audio recording / reproducing apparatus
CN1763841A (en) Tone data generation method and tone synthesis method, and apparatus therefor
CN103035250A (en) Audio encoding device
CN115171632A (en) Audio processing method, computer device and computer program product
JP4202964B2 (en) Device for adding music data to video data
JP2001265330A (en) Device and method for extracting melody

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C17 Cessation of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20040908

Termination date: 20120714