(19)
(11)EP 2 615 736 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
08.08.2018 Bulletin 2018/32

(21)Application number: 11823574.6

(22)Date of filing:  06.09.2011
(51)International Patent Classification (IPC): 
H03G 7/00(2006.01)
(86)International application number:
PCT/JP2011/070283
(87)International publication number:
WO 2012/033099 (15.03.2012 Gazette  2012/11)

(54)

SIGNAL PROCESSING DEVICE AND METHOD, PROGRAM, AND DATA RECORDING MEDIUM

SIGNALVERARBEITUNGSVORRICHTUNG, -VERFAHREN UND -PROGRAMM SOWIE DATENAUFZEICHNUNGSMEDIUM

DISPOSITIF ET PROCÉDÉ DE TRAITEMENT DU SIGNAL, AINSI QUE PROGRAMME ET SUPPORT D'ENREGISTREMENT DE DONNÉES


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 08.09.2010 JP 2010201078

(43)Date of publication of application:
17.07.2013 Bulletin 2013/29

(73)Proprietor: Sony Corporation
Tokyo 108-0075 (JP)

(72)Inventors:
  • TSUJI Minoru
    Tokyo 108-0075 (JP)
  • CHINEN Toru
    Tokyo 108-0075 (JP)

(74)Representative: D Young & Co LLP 
120 Holborn
London EC1N 2DY
London EC1N 2DY (GB)


(56)References cited: : 
EP-A1- 2 209 326
WO-A1-2010/005033
JP-A- 2002 504 279
US-A1- 2005 147 262
EP-A2- 1 251 715
JP-A- 5 234 249
JP-A- 2007 067 868
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    Technical Field



    [0001] The present invention relates to a signal processing apparatus and method, a program, and a data recording medium, and more particularly, relates to a signal processing apparatus and method, a program, and a data recording medium configured such that the playback level of an audio signal can be easily and effectively enhanced without requiring additional information given by prior analysis.

    Background Art



    [0002] For example, in the case where movie content or music content with a large dynamic range in audio volume is played back on a portable device with compact built-in speakers, not only does the audio volume become lower overall, but audio such as low-volume dialogue in particular becomes difficult to hear.

    [0003] Thus, although normalization and automatic gain control technology does exist as technology for making the audio of such content easier to hear, volume control becomes audibly unstable unless data is read sufficiently far enough ahead.

    [0004] There also exists technology that boosts low-volume portions and compresses high-volume portions of audio by means of a dynamic range compression process for volume. However, with a compression process, it is difficult to obtain large audio enhancement effects if generalized volume boost and compression settings are used. In order to obtain large effects, it is necessary to vary settings on a per-content basis.

    [0005] For example, there exists technology that takes a sound pressure level specified by dialogue normalization as a basis, boosting signals with a lower sound pressure level and compressing signals with a higher sound pressure level. With this technology, however, it is necessary to specify boost and compression settings and a sound pressure level for dialogue normalization at the time of encoding the audio signal in order to obtain sufficient effects.

    [0006] Furthermore, technology has also been proposed in which, in the case of compressing the dynamic range of audio volume, faint sounds in an audio signal are made easier to hear by multiplying the audio signal by coefficients determined by the average of the absolute values of the audio signal (see PTL 1, for example). PLT 2 teaches a multi-channel hearing instrument with inter-channel communication, PLT 3 discloses an auditory sensibility correction device capable of adding favourable loudness characteristics to an output sound without depending on sound source characteristics of an audio signal even when a high-frequency band or a low-frequency band of the audio signal has already been subjected to loudness correction. PLT 4 is directed to an acoustic processing apparatus which allows for reproduction of an audio signal with an acoustically satisfactory quality by being capability of adding an acoustic process corresponding to a listener's sense to the audio signal independently of the properties of an audio source. PLT 5 describes a method of decreasing the dynamic range of a signal.

    Citation List


    Patent Literature



    [0007] 

    PTL 1: Japanese Unexamined Patent Application Publication No. 5-275950

    PLT 2: EP 1251715 A2

    PLT 3: EP 2209326 A1

    PLT 4: WO 2010/005033 A1

    PLT 5: US 2005/147262 A1


    Summary of Invention


    Technical Problem



    [0008] Meanwhile, various types of content such as movies, music, and self-recorded content are coming to be played back on portable devices with compact built-in speakers. However, much of such content lacks additional information for effective volume control given by prior analysis done at the time of encoding, etc. as discussed above. For this reason, there is a need for technology that effectively conducts volume control even in cases where additional information obtained by prior analysis is not included in the audio signal of given content.

    [0009] Accordingly, if the technology described in PTL 1 discussed above is used, it becomes possible to suppress sudden changes in loudness while making faint sounds in an audio signal easier to hear by means of a compression process, without requiring prior analysis of the audio signal. However, the playback level of an audio signal cannot be sufficiently enhanced with this technology.

    [0010] For example, with the technology described in PTL 1, since amplitude is simply damped by multiplying an audio signal by a constant, there is little freedom in amplitude conversion settings, and it cannot be said that the playback level of an audio signal can be effectively enhanced. Also, this technology can only be used in the case of narrowing the dynamic range of volume by amplitude conversion of an audio signal. Converting amplitude without changing the dynamic range of volume or widening the dynamic range of volume cannot be conducted.

    [0011] The present invention, being devised in light of such circumstances, is configured such that the playback level of an audio signal can be easily and effectively enhanced without requiring additional information given by prior analysis. The invention is described by the contents of the independent claims.

    Solution to Problem



    [0012] A signal processing apparatus of a first aspect of the present invention comprises analyzing means analyzing input signal characteristics, mapping processing means conducting amplitude conversion of the input signal on the basis of a predetermined linear function or nonlinear function, weighting controlling means respectively multiplying a plurality of the input signals, which have been respectively amplitude-converted on the basis of mutually different functions by a plurality of the mapping processing means, by weights determined by the analysis result for the input signal characteristics, and adding means generating an output signal by adding together the plurality of input signals which have been multiplied by the weights.

    [0013] The analyzing means may be made to compute a value expressing the average sample value of samples included in a given segment of the input signal as the analysis result.

    [0014] The analysis result may be taken to be the root mean square or a moving average of sample values of samples included in the given segment.

    [0015] In the case where amplitude conversion is conducted on the input signal for each of a plurality of channels to generate an output signal for each channel, the analyzing means may be made to compute one analysis result shared by all channels.

    [0016] It may be configured such that the weights are determined by the analysis result for every single sample of the input signal.

    [0017] It may be configured such that the weights are determined by the analysis result for every given number of two or more consecutive samples of the input signal.

    [0018] A signal processing method or program of a first aspect of the present invention includes the steps of analyzing input signal characteristics, conducting amplitude conversion of the input signal on the basis of a predetermined linear function or nonlinear function, respectively multiplying a plurality of the input signals, which have been amplitude-converted on the basis of a plurality of mutually different functions, by weights determined by the analysis result for the input signal characteristics, and generating an output signal by adding together the plurality of input signals which have been multiplied by the weights.

    [0019] In a first aspect of the present invention, input signal characteristics are analyzed, amplitude conversion of the input signal is conducted on the basis of a predetermined linear function or nonlinear function, a plurality of the input signals, which have been respectively amplitude-converted on the basis of mutually different functions by a plurality of the mapping processing means, are respectively multiplied by weights determined by the analysis result for the input signal characteristics, and an output signal is generated by adding together the plurality of input signals which have been multiplied by the weights.

    [0020] A data recording medium of a second aspect of the present invention has recorded thereon an output signal obtained by analyzing input signal characteristics, conducting amplitude conversion of the input signal on the basis of a predetermined linear function or nonlinear function, respectively multiplying a plurality of the input signals, which have been amplitude-converted on the basis of a plurality of mutually different functions, by weights determined by the analysis result for the input signal characteristics, and adding together the plurality of input signals which have been multiplied by the weights.

    [0021] A signal processing apparatus of a third aspect of the present invention comprises analyzing means analyzing input signal characteristics, and mapping processing means generating an output signal by conducting amplitude conversion of the input signal on the basis of a nonlinear function determined by the analysis result for the input signal characteristics.

    [0022] The analyzing means may be made to compute a value expressing the average sample value of samples included in a given segment of the input signal as the analysis result.

    [0023] The analysis result may be taken to be the root mean square or a moving average of sample values of samples included in the given segment.

    [0024] In the case where amplitude conversion is conducted on the input signal for each of a plurality of channels to generate an output signal for each channel, the analyzing means may be made to compute one analysis result shared by all channels on the basis of the input signals in the plurality of channels.

    [0025] The nonlinear function may be determined by the analysis result for every single sample of the input signal.

    [0026] The nonlinear function may be determined by the analysis result for every given number of two or more consecutive samples of the input signal.

    [0027] A signal processing method or program of a third aspect of the present invention includes the steps of analyzing input signal characteristics, and generating an output signal by conducting amplitude conversion of the input signal on the basis of a nonlinear function determined by the analysis result for the input signal characteristics.

    [0028] In a third aspect of the present invention, input signal characteristics are analyzed, and an output signal is generated by conducting amplitude conversion of the input signal on the basis of a nonlinear function determined by the analysis result for the input signal characteristics.

    [0029] A data recording medium of a fourth aspect of the present invention has recorded thereon an output signal obtained by analyzing input signal characteristics, and conducting amplitude conversion of the input signal on the basis of a nonlinear function determined by the analysis result for the input signal characteristics. Advantageous Effects of Invention

    [0030] According to the first through fourth aspects of the present invention, the playback level of an audio signal can be easily and effectively enhanced without requiring additional information given by prior analysis.

    Brief Description of Drawings



    [0031] 

    [Fig. 1] Fig. 1 is a diagram illustrating an exemplary configuration of a first embodiment of an audio signal processing apparatus applying the present invention.

    [Fig. 2] Fig. 2 is a flowchart explaining a conversion process.

    [Fig. 3] Fig. 3 is a diagram illustrating exemplary mapping functions.

    [Fig. 4] Fig. 4 is a diagram illustrating exemplary mapping functions.

    [Fig. 5] Fig. 5 is a diagram illustrating exemplary mapping functions.

    [Fig. 6] Fig. 6 is a diagram illustrating another exemplary configuration of an audio signal processing apparatus.

    [Fig. 7] Fig. 7 is a flowchart explaining a conversion process.

    [Fig. 8] Fig. 8 is a diagram illustrating another exemplary configuration of an audio signal processing apparatus.

    [Fig. 9] Fig. 9 is a flowchart explaining a conversion process.

    [Fig. 10] Fig. 10 is a diagram illustrating exemplary mapping functions.

    [Fig. 11] Fig. 11 is a diagram illustrating another exemplary configuration of an audio signal processing apparatus.

    [Fig. 12] Fig. 12 is a flowchart explaining a conversion process.

    [Fig. 13] Fig. 13 is a block diagram illustrating an exemplary configuration of a computer.


    Description of Embodiments



    [0032] Hereinafter, embodiments applying the present invention will be described with reference to the drawings.

    <First embodiment>


    [Configuration of audio signal processing apparatus]



    [0033] Fig. 1 is a diagram illustrating an exemplary configuration of an embodiment of an audio signal processing apparatus applying the present invention.

    [0034] The audio signal processing apparatus 11 is provided in a portable playback apparatus that plays back content consisting of video signals and audio signals, for example, and conducts amplitude conversion on an input audio signal such that the playback level is enhanced, and outputs the amplitude-converted audio signal. Note that hereinafter, an audio signal input specifically input into the audio signal processing apparatus 11 is designated an input signal, while an audio signal obtained by amplitude-converting an input signal is designated an output signal.

    [0035] The audio signal processing apparatus 11 is composed of an analyzer 21, a mapping processor 22, an output unit 23, and a drive 24.

    [0036] The analyzer 21 analyzes the characteristics of a supplied input signal, and supplies mapping control information indicating the analysis results to the mapping processor 22.

    [0037] The mapping processor 22 uses mapping control information supplied from the analyzer 21 to conduct a mapping process on a supplied input signal and enhance the playback level of the input signal. In the mapping process, linear or non-linear amplitude conversion is conducted on an input signal. The mapping processor 22 supplies the output unit 23 with an output signal obtained by the mapping process.

    [0038] The output unit 23 may output an output signal supplied from the mapping processor 22 to a subsequent audio output unit, etc. or supply it to the drive 24. The drive 24 records an output signal supplied from the output unit 23 to a removable medium 25, which is a recording medium able to be freely inserted into and removed from the drive 24.

    [Description of conversion process]



    [0039] Next, operation of the audio signal processing apparatus 11 in Fig. 1 will be described.

    [0040] The audio signal processing apparatus 11 conducts a conversion process upon being supplied with an input signal, and generates and outputs an output signal. Hereinafter, a conversion process conducted by the audio signal processing apparatus 11 will be described with reference to the flowchart in Fig. 2.

    [0041] In step S11, the analyzer 21 analyzes the characteristics of a supplied input signal, and generates mapping control information.

    [0042] Specifically, the analyzer 21 may, for example, perform the computation in the following Eq. 1, and compute the root mean square RMS(n) for the nth sample of the input signal as mapping control information for the nth sample.
    [Expression 1]



    [0043] In Eq. 1 herein, x(m) represents the sample value of the mth sample of the input signal (the input signal value). Also, in Eq. 1, the input signal values, or in other words the sample values of each input signal sample, are taken to be normalized such that -1≤x(m)≤1.

    [0044] Consequently, the root mean square RMS(n) is obtained by taking the logarithm of the square root of the mean square value of the sample values for samples included in a segment consisting of N consecutive samples centered about the nth sample, and multiplying the value thus obtained by the constant 20.

    [0045] Values of the root mean square RMS(n) obtained in this way decrease with decreasing absolute values of the sample values for the samples in a specific segment centered about the nth sample of the input signal being processed. In other words, the root mean square RMS(n) decreases as the overall audio volume decreases in a specific segment that includes the input signal sample being processed.

    [0046]  Once the analyzer 21 analyzes an input signal and mapping control information is obtained, the mapping control information is supplied to the mapping processor 22.

    [0047] In step S12, the mapping processor 22 uses mapping control information from the analyzer 21 to conduct a mapping process on a supplied input signal, and generates an output signal.

    [0048] Specifically, the mapping processor 22 converts amplitude by substituting the sample value x of the nth sample of an input signal with the nonlinear mapping function f(x) expressed in the following Eq. 2. In other words, the value obtained by substituting the sample value x with the mapping function f(x) is taken to be the sample value of the nth sample of the output signal.
    [Expression 2]



    [0049] In Eq. 2 herein, the sample value x of an input signal is taken to be normalized to a value from -1 to 1. Also, in Eq. 2, α represents a control factor, with the control factor α being a numerical value determined by the value of the mapping control information.

    [0050] According to Eq. 2, in the case where the sample value x is -1 or greater but less than -0.5, the mapping function f(x) becomes (2-α)x-(α-1), whereas in the case where the sample value x is between -0.5 and 0.5 inclusive, the mapping function f(x) becomes αx. Also, in the case where the sample value x is greater than 0.5 but 1 or less, the mapping function f(x) becomes (2-α)x+(α-1).

    [0051] Additionally, the control factor α is taken to be a value determined by the following Eq. 3, for example.
    [Expression 3]



    [0052] In other words, in the case where the value of the root mean square RMS(n) given to be the mapping control information is greater than -30 but 0 or less, the value of the control factor α is taken to be (RMS(n)/-30)+1. At this point, the range of values that the control factor α may take becomes 1≤α<2, with the value of the control factor α decreasing as the root mean square RMS(n) increases. Also, in the case where the value of the root mean square RMS(n) is -30 or less, the value of the control factor α is taken to be 2.

    [0053] Consequently, the control factor α increases as the volume of audio based on an input signal decreases overall, or in other words as the root mean square RMS(n) decreases. As a result, the slope of the mapping function f(x) increases as illustrated in Fig. 3.

    [0054] In Fig. 3 herein, the horizontal axis represents the sample value x of an input signal, while the vertical axis represents the value of the mapping function f(x). Also, the line f11, the curve f12, and the curve f13 represent the mapping function f(x) for a control factor α of 1.0, 1.4, and 1.8, respectively.

    [0055] As Fig. 3 demonstrates, as the volume of audio based on an input signal decreases overall, amplitude conversion of the input signal is conducted using a steeper mapping function f(x). In other words, as the audio volume decreases overall, the rate of change in f(x) versus change in the sample value x increases in the mapping function f(x).

    [0056]  For example, for a control factor α of 1.0 and a mapping function f(x) expressed by the line f11, a sample value in the input signal is taken without change as a sample value in the output signal. In contrast, with a control factor α of 1.8 and a mapping function f(x) expressed by the curve f13, the slope of the mapping function becomes greater than the slope of the mapping function expressed by the line f11 in the segment where the sample value x of an input signal is between -0.5 and 0.5 inclusive.

    [0057] In this way, as the volume of audio in an input signal decreases overall, amplitude conversion of the input signal is conducted using a mapping function f(x) with steeper characteristics in the segment for most sample values x, including sample values x equal to 0.

    [0058] Consequently, in segments where the audio of an input signal is at low volume overall, the input signal is amplitude-converted such that low-volume audio is converted into higher-volume audio, enhancing the playback level of the input signal. Thus, even faint sounds that have been conventionally difficult to hear can be made easier to hear by a mapping process conducted on an input signal, even in cases where movies or other content with a large dynamic range in volume is played back on a portable device housing compact speakers.

    [0059] Also, even in segments where the audio of an input signal is at high volume overall, amplitude conversion of the input signal is conducted using a mapping function f(x) with appropriately steep characteristics on the parts of the signal with small sample values x therein.

    [0060] Consequently, even in segments where the audio of an input signal is at high volume overall, the input signal is amplitude-converted such that faint audio therein is converted into loud audio, enhancing the playback level of the input signal. Thus, even sounds that conventionally have been played back comparatively loudly become heard even more loudly.

    [0061] Moreover, with the audio signal processing apparatus, 11, it is not necessary to analyze an input signal in advance and add additional information for amplitude conversion to the input signal, or to read ahead long segments of an input signal and analyze the read input signal before conducting amplitude conversion.

    [0062] Also, by varying a nonlinear mapping function f(x) according to a control factor α, amplitude conversion with a higher degree of freedom can be realized. In other words, by taking the nonlinear function with the most effective characteristics as the mapping function depending on the overall characteristics of a specific segment of an input signal, amplitude conversion can be conducted while taking into account not only the characteristics of the segment containing the sample being processed, but also the magnitude of the sample value for that sample.

    [0063] For example, with the technology described in PTL 1 discussed earlier, an audio signal is multiplied by a constant determined by the average of the absolute values of the audio signal, irrespectively of the audio signal values. In other words, a sample of an audio signal would be always a constant multiple, regardless of the magnitude of the amplitude of that sample.

    [0064] For this reason, if a computation method that multiplies an audio signal by a constant is determined with respect to comparatively low-volume audio in order to increase the volume of that audio, high-volume audio may not reach a suitable volume even if the audio signal is multiplied by a constant.

    [0065]  In contrast, with the audio signal processing apparatus 11, if the mapping function is taken to be a nonlinear function, amplitude conversion with a high degree of freedom becomes possible, such as greatly increasing the amplitude in the case where the sample amplitude (sample value) is small, but not greatly changing the amplitude in the case where the sample amplitude is large. Thus, the playback level of an audio signal can be effectively enhanced, such as by converting low-volume audio into higher-volume audio while also not significantly changing the volume of audio that was at high volume to begin with.

    [0066] Furthermore, by appropriately setting a mapping function f(x), the dynamic range in audio volume can be widened, kept the same, or the dynamic range can be narrowed by amplitude conversion.

    [0067] In this way, with the audio signal processing apparatus 11, the playback level of an audio signal can be easily and effectively enhanced.

    [0068] Returning to the description of the flowchart in Fig. 2, upon generating an output signal by means of a mapping process, the mapping processor 22 supplies the obtained output signal to the output unit 23.

    [0069] In step S13, the output unit 23 outputs an output signal supplied from the mapping processor 22 to a subsequent unit, and the conversion process ends. The output unit 23 may also supply the output signal to the drive 24 as necessary, with the drive 24 recording the supplied output signal to the removable medium 25.

    [0070] As above, the audio signal processing apparatus 11 analyzes the characteristics of an input signal and uses a mapping function that varies according to the analysis results to conduct a mapping process on the input signal and generate an output signal.

    [0071] The root mean square RMS(n) taken as the mapping control information obtained by analyzing an input signal expresses the average magnitude of sample values in a given segment of an input signal, or in other words, an amplitude distribution of samples in a given segment. For example, in the case where the root mean square RMS(n) is small, many samples of low amplitude are contained in the input signal, whereas in the opposite case where the root mean square RMS(n) is large, many samples of high amplitude are contained in the input signal.

    [0072]  In the audio signal processing apparatus 11, by using the root mean square RMS(n) to generate a mapping function with more effective characteristics and then conduct a mapping process, an input signal can be easily converted into an output signal having an ideal amplitude distribution.

    <Modification 1>



    [0073] Although the foregoing describes using a function which is linear in each segment as the mapping function, a function with smoother curves may also be used as the mapping function. In such cases, the nonlinear function expressed in the following Eq. 4 may be taken to be the mapping function, for example.
    [Expression 4]



    [0074] In Eq. 4 herein, x represents the sample value of an input signal sample. The input signal sample value x is taken to be normalized to a value from -1 to 1. Also, the control factor α in Eq. 4 is determined by the following Eq. 5.
    [Expression 5]



    [0075] In other words, in the case where the value of the root mean square RMS(n) given to be the mapping control information is greater than -30 but 0 or less, the value of the control factor α is taken to be (RMS(n)/-30)x5+5. At this point, the range of values that the control factor α may take becomes 5≤α<10, with the value of the control factor α decreasing as the root mean square RMS(n) increases. Also, in the case where the value of the root mean square RMS(n) is -30 or less, the value of the control factor α is taken to be 10.

    [0076] Consequently, the control factor α increases as the volume of audio based on an input signal decreases overall, or in other words as the root mean square RMS(n) decreases. As a result, the slope of the mapping function f(x) varies as illustrated in Fig. 4.

    [0077] In Fig. 4 herein, the horizontal axis represents the sample value x of an input signal, while the vertical axis represents the value of the mapping function f(x). Also, the curves f21 to f23 represent the mapping function f(x) for a control factor α of 5, 7, and 10, respectively.

    [0078] As Fig. 4 demonstrates, as the volume of audio based on an input signal decreases overall, amplitude conversion of the input signal is conducted using a mapping function f(x) with a greater overall rate of change versus change in the sample value x. In other words, as the volume of audio based on an input signal decreases overall, a steeper mapping function f(x) is used in the segment for most sample values x, including sample values x equal to 0.

    [0079] In this way, in the case of using the mapping function expressed in Eq. 4 and the control factor α expressed in Eq. 5, the mapping function takes on steeper characteristics as the root mean square RMS(n) decreases, similarly to the case of using the mapping function in Eq. 2 and the control factor α in Eq. 3. Likewise in this case, the playback level of an audio signal can be more easily and effectively enhanced without requiring prior analysis or reading ahead of the input signal. Furthermore, faint sounds that have been conventionally difficult to hear can be made easier to hear, while in addition, even sounds that have been comparatively loud conventionally become even easier to hear.

    <Modification 2>



    [0080] Meanwhile, since the mapping function in Eq. 4 discussed above is an exponential function, if computation of the mapping process is performed by a computer or DSP (Digital Signal Processor), etc., the computational load typically increases. Thus, by adopting the cubic function expressed in the following Eq. 6 as the mapping function, for example, the computational load can be decreased, and the mapping process can be conducted more rapidly.
    [Expression 6]



    [0081] In Eq. 6 herein, x represents the sample value of an input signal sample. The input signal sample value x is taken to be normalized to a value from -1 to 1. Also, the control factor α in Eq. 6 is determined by the following Eq. 7.
    [Expression 7]



    [0082] In other words, in the case where the value of the root mean square RMS(n) given to be the mapping control information is between -0.9 and 0 inclusive, the control factor α is taken to be 100, whereas in the case where the value of the root mean square RMS(n) is greater than -30 but less than -0.9, the value of the control factor α is taken to be -30/RMS(n)×3. At this point, the range of values that the control factor α may take becomes 3<α<100, with the value of the control factor α increasing as the root mean square RMS(n) increases. Also, in the case where the value of the root mean square RMS(n) is -30 or less, the value of the control factor α is taken to be 3.0.

    [0083] Consequently, the control factor α decreases as the volume of audio based on an input signal decreases overall, or in other words as the root mean square RMS(n) decreases. As a result, the mapping function f(x) varies steeply as illustrated in Fig. 5.

    [0084] In Fig. 5 herein, the horizontal axis represents the sample value x of an input signal, while the vertical axis represents the value of the mapping function f(x). Also, the curves f31 to f33 represent the mapping function f(x) for a control factor α of 100, 5, and 3, respectively.

    [0085] As Fig. 5 demonstrates, as the volume of audio based on an input signal decreases overall, amplitude conversion of the input signal is conducted using a mapping function f(x) with a greater overall rate of change versus change in the sample value x.

    [0086] In this way, in the case of using the mapping function expressed in Eq. 6 and the control factor α expressed in Eq. 7, the mapping function takes on steeper characteristics as the root mean square RMS(n) decreases, similarly to the case of using the mapping function in Eq. 2 and the control factor α in Eq. 3. Likewise in this case, the playback level of an audio signal can be more easily and effectively enhanced without requiring prior analysis or reading ahead of the input signal. Furthermore, faint sounds that have been conventionally difficult to hear can be made easier to hear, while in addition, even sounds that have been comparatively loud conventionally become even easier to hear.

    [0087] Note that the mapping function f(x) may be any function insofar as the function becomes -1≤f(x)≤1 for sample values x where -1≤x≤1.

    [0088] Also, although an example of using the root mean square RMS(n) as mapping control information for determining a control factor α is described in the foregoing, the mapping control information may be another element, or may be taken to be mapping control information combining multiple elements. For example, the moving average of sample values for samples in a given segment of an input signal, the number of zero-crossings in samples in a given segment, or a value expressing the tonality of an input signal may also be taken to be mapping control information. In other words, it may be configured such that elements with large processing effects, whereby audio suitable for listening is obtained by the mapping process, may be used for the mapping function, the control factor α, and the mapping control information.

    [0089] Also, although the foregoing describes conducting a mapping process by computing mapping control information and a control factor α for the mapping function for every single sample of an input signal, the mapping process may also be conducted by computing mapping control information and a control factor α for every two or more consecutive samples. For example, in such cases, the mapping control information and control factor computed for one sample may be successively used for a given number of consecutive samples.

    [0090] Furthermore, it may also be configured such that the method for computing mapping control information or a control factor is varied according to the output device for the output signal, etc., so as to adjust the degree of enhancement to the playback level of an audio signal. Specifically, the relationship between the control factor α and the mapping control information may be varied, such as by using different formulas for computing the control factor α, for example.

    <Second embodiment>


    [Configuration of audio signal processing apparatus]



    [0091] Meanwhile, in the case where an audio signal given as an input signal has two or more channels, inter-channel volume balance of audio based on the output signal may change if analysis of input signal characteristics or a mapping process is conducted independently on each channel.

    [0092] For this reason, it is desirable to conduct identical analyses and mapping processes on the input signals in all channels. Thus, it may also be configured such that characteristics are analyzed for the input signals in all channels, and a mapping process is conducted using one set of mapping control information obtained from the analysis results. In such cases, the audio signal processing apparatus may be configured as illustrated in Fig. 6, for example.

    [0093] The audio signal processing apparatus 51 is composed of an analyzer 21, a mapping processor 22, a mapping processor 61, an output unit 23, and a drive 24. In Fig. 6 herein, like numerals are given to portions corresponding to the case in Fig. 1, and description of such portions will be reduced or omitted.

    [0094] Input signals are supplied to the audio signal processing apparatus 51 as a left-channel audio signal and a right-channel audio signal constituting a movie or other content, for example. In other words, the left-channel input signal is supplied to the analyzer 21 and the mapping processor 22, while the right-channel input signal is supplied to the analyzer 21 and the mapping processor 61.

    [0095] The analyzer 21 analyzes the respective characteristics of the supplied left- and right-channel input signals, generates mapping control information on the basis of the two sets of analysis results thus obtained, and supplies the mapping control information to the mapping processor 22 and the mapping processor 61.

    [0096] The mapping processor 61 uses the mapping control information supplied from the analyzer 21 to conduct a mapping process on the supplied right-channel input signal and generates a right-channel output signal. At this point, a process similar to that of the mapping processor 22 is conducted in the mapping processor 61. The mapping processor 61 supplies the right-channel output signal obtained by the mapping process to the output unit 23.

    [0097] In this way, in the audio signal processing apparatus 51, shared mapping control information is used to conduct mapping processes in the mapping processor 22 and the mapping processor 61.

    [0098] The output unit 23 may output the left- and right-channel output signals supplied from the mapping processor 22 and the mapping processor 61 to a subsequent unit or to the drive 24 for recording to the removable medium 25.

    [Description of conversion process]



    [0099] Next, a conversion process conducted by the audio signal processing apparatus 51 will be described with reference to the flowchart in Fig. 7.

    [0100] In step S41, the analyzer 21 analyzes the characteristics of supplied left- and right-channel input signals. For example, the analyzer 21 may perform the computation in Eq. 1 discussed earlier, and compute a left-channel root mean square RMS(n) and a right-channel root mean square RMS(n).

    [0101] In step S42, the analyzer 21 generates mapping control information on the basis of the input signal characteristics analysis results, and supplies the mapping control information to the mapping processor 22 and the mapping processor 61. For example, the analyzer 21 may compute the average of the left-channel root mean square RMS(n) and the right-channel root mean square RMS(n), and take the obtained average to be the mapping control information.

    [0102] However, it may also be configured such that the larger value of the left-channel root mean square RMS(n) and the right-channel root mean square RMS(n) may be taken without change as the mapping control information. Also, samples from the left-channel input signal and samples from the right-channel input signal may be used to compute a single root mean square RMS(n), etc. as the mapping control information.

    [0103] Once the operation in step S42 is conducted and mapping control information is generated, the operations in steps S43 and S44 are subsequently conducted and the conversion process ends. However, since these processing operations are similar to the operations in steps S12 and S13 of Fig. 2, their description will be reduced or omitted.

    [0104] However, in step S43, the mapping control information is used in the mapping processor 22 and the mapping processor 61, and a left-channel output signal and a right-channel output signal are respectively generated with identical mapping functions and control factors.

    [0105] In so doing, the audio signal processing apparatus 51 analyzes the characteristics of left- and right-channel input signals, generates common mapping control information for the left and right channels, and uses the obtained mapping control information to conduct an identical mapping process on each channel. By using common mapping control information for the left and right channels to conduct an identical mapping process on per-channel input signals in this way, the playback level of an audio signal can be enhanced without changing the inter-channel volume balance.

    [0106] Although the foregoing describes a case where two left-and right-channel input signals are input, an input signal may also be composed of three or more channels. Even in such cases, common mapping control information is generated for all channels.

    <Third embodiment>


    [Configuration of audio signal processing apparatus]



    [0107] Also, although the foregoing describes using a single mapping function to generating an output signal, a plurality of linear or nonlinear mapping functions may be prepared, and it may be configured such that an output signal is generated by selectively using those mapping functions according to the mapping control information. In such cases, changes in the output that occur due to switching the mapping function used to generate the output signal can be made smoother by taking the output signal to be a weighted sum of outputs from plural mapping functions depending on the mapping control information.

    [0108] In this way, in the case of generating an output signal by using a plurality of mapping functions, an audio signal processing apparatus may take the configuration illustrated in Fig. 8, for example.

    [0109] Namely, the audio signal processing apparatus 91 is composed of an analyzer 21, mapping processors 101-1 to 101-M, a weighting controller 102, an adder 103, an output unit 23, and a drive 24. In Fig. 8 herein, like numerals are given to portions corresponding to the case in Fig. 1, and description of such portions will be reduced or omitted.

    [0110] The mapping processors 101-1 to 101-M each conduct a mapping process on a supplied input signal using respectively different mapping functions, and supply the output signals obtained as a result to the weighting controller 102. Note that hereinafter, the mapping processors 101-1 to 101-M will also be simply designated the mapping processors 101 in cases where it is not necessary to individually distinguish them.

    [0111] The weighting controller 102 multiplies output signals supplied from the mapping processors 101 by distribution ratios, which are weights determined by mapping control information supplied from the analyzer 21, and supplies the result to the adder 103. In other words, the weighting controller 102 is provided with multipliers 111-1 to 111-M. The multipliers 111-1 to 111-M multiply output signals supplied from the mapping processors 101-1 to 101-M by distribution ratios α1 to αM determined by mapping control information, and supply the results to the adder 103.

    [0112] Note that hereinafter, the multipliers 111-1 to 111-M will also be simply designated the multipliers 111 in cases where it is not necessary to individually distinguish them.

    [0113] The adder 103 adds together M output signals supplied from the multipliers 111, and supplies the final output signal obtained as a result to the output unit 23.

    [Description of conversion process]



    [0114] Next, a conversion process conducted by the audio signal processing apparatus 91 will be described with reference to the flowchart in Fig. 9.

    [0115] In step S71, the analyzer 21 analyzes the characteristics of a supplied input signal. For example, the analyzer 21 may perform the computation in Eq. 1 discussed earlier, compute the root mean square RMS(n) for the nth sample of the input signal as mapping control information, and supply the mapping control information to the weighting controller 102.

    [0116] In step S72, the mapping processors 101 conduct a mapping process on supplied input signals, and supply the obtained output signals to the multipliers 111.

    [0117] For example, assume that the audio signal processing apparatus 91 is provided with four mapping processors 101-1 to 101-4. In this case, the mapping processors 101-1 to 101-4 conduct a mapping process on supplied input signals by using the mapping functions f1(x) to f4(x) expressed in the following Eqs. 8 to 11. In other words, the value obtained by substituting in a mapping function for the sample value x of the nth sample of an input signal is taken to be the sample value of the nth sample of an output signal.
    [Expression 8]









    [0118] In Eqs. 8 to 11 herein, x represents the sample value of an input signal sample. In Eqs. 8 to 11, the input signal sample value x is taken to be normalized to a value from -1 to 1.

    [0119] These mapping functions f1(x) to f4(x) are functions ordered from the mapping function f4(x) to the mapping function f1(x) in order of steepest characteristics.

    [0120] In Fig. 10 herein, the mapping functions f1(x) to f4(x) are illustrated in the upper-left, the upper-right, the lower-left, and the lower-right, respectively. Also, in Fig. 10, the horizontal axis represents the sample value x of an input signal, while the vertical axis represents the value of a mapping function.

    [0121] For example, in the drawing, the mapping function f1(x)=x illustrated in the upper-left is a linear first-order function, in which a sample value x in an input signal is taken without change as a sample value in an output signal. Also, the mapping functions f2(x) to f4(x) are nonlinear exponential functions ordered from the mapping function f4(x) to the mapping function f2(x) in order of steepest characteristics in the segment for most sample values x, including sample values x equal to 0. In other words, the mapping functions are ordered from the mapping function f4(x) to the mapping function f2(x) in order of the highest rate of change in the mapping function output versus change in the sample value x.

    [0122] Returning to description of the flowchart in Fig. 9, in step S73, the weighting controller 102 multiplies output signals supplied from the mapping processors 101 by distribution ratios acting as weights, on the basis of mapping control information supplied from the analyzer 21.

    [0123] For example, assume that the audio signal processing apparatus 91 is provided with four mapping processors 101-1 to 101-4. In this case, the weighting controller 102 performs the computation expressed in the following Eq. 12 and computes distribution ratios α1 to α4 for the mapping functions f1(x) to f4(x), on the basis of the root mean square RMS(n) given as mapping control information.
    [Expression 9]



    [0124] In other words, in the case where the root mean square RMS(n) is greater than -12 but 0 or less, the distribution ratios are taken to be α1=1-RMS(n)/-12, α2=1-α1, and α34=0. Herein, α1 in the distribution ratio α2 represents the value of the distribution ratio α1.

    [0125] Also, in the case where the root mean square RMS(n) is greater than -24 but -12 or less, the distribution ratios are taken to be α14=0, α2=1-(RMS(n)+12)/-12, and α3=1-α2. Herein, α2 in the distribution ratio α3 represents the value of the distribution ratio α2.

    [0126] Additionally, in the case where the root mean square RMS(n) is greater than -32 but -24 or less, the distribution ratios are taken to be α12=0, α3=1-(RMS(n)+24)/-12, and α4=1-α3. Herein, α3 in the distribution ratio α4 represents the value of the distribution ratio α3. In the case where the root mean square RMS(n) is -32 or less, the distribution ratios are taken to be α123=0 and α4=1.

    [0127] In this way, the weighting controller 102 computes distribution ratios α1 to α4 such that the weights on steeper mapping functions increase as the root mean square RMS(n) decreases, or in other words, as the volume of audio based on an input signal decreases overall. In contrast, distribution ratios are computed such that the weights increase on mapping functions with gentler characteristics in segments with high-volume audio based on an input signal.

    [0128] Consequently, amplitude conversion of an input signal is conducted such that weights are placed on steeper mapping functions in segments with low-volume audio in an input signal.

    [0129] In so doing, once distribution ratios are computed for the mapping functions, the multipliers 111-1 to 111-4 multiply output signals supplied from the mapping processors 101-1 to 101-4 by distribution ratios α1 to α4, and supply the results to the adder 103.

    [0130] In the foregoing, it is described that output signals are first generated in the mapping processors 101, and then distribution ratios by which to multiply those output signals are computed. However, it may also be configured such that mapping processes are conducted after computing distribution ratios.

    [0131] In such cases, distribution ratios α1 to αM may be computed on the basis of mapping control information, for example, and only output signals multiplied by non-zero distribution ratios from among these distribution ratios are generated by mapping processes. In other words, mapping processes are not conducted on input signals in mapping processors 101 which use mapping functions corresponding to a distribution ratio whose value is 0. By configuring it such that only output signals multiplied by non-zero distribution ratios are generated, the computation load can be further reduced, and a final output signal can be rapidly obtained.

    [0132] Also, although distribution ratios are described as being computed by arithmetic computation, it may also be configured such that a table is prepared in which mapping control information values and values of distribution ratios α1 to αM determined by those values are recorded in association with each other, with distribution ratios being acquired by table lookup.

    [0133] In step S74, the adder 103 adds together output signals supplied from the multipliers 111-1 to 111-M, and generates a single, final output signal.

    [0134] In this way, by using distribution ratios to perform weighted addition of output signals obtained using a plurality of mutually different mapping functions, amplitude conversion of an input signal using a single mapping function can be approximately realized. For example, if the distribution ratios α1 and α2 are taken to be 0 while the distribution ratios α3 and α4 are taken to be given non-zero values, amplitude conversion can be realized using a function having characteristics intermediate between the mapping function f3(x) and the mapping function f4(x).

    [0135] By dynamically varying an approximately generated mapping function according to the characteristics in each segment of an input signal, amplitude conversion with a higher degree of freedom can be realized with the audio signal processing apparatus 91. In other words, by approximately generating a mapping function according to per-segment characteristics in each segment of an input signal, amplitude conversion can be conducted while taking into account not only the characteristics of a segment, but also the sample value of the sample being processed, similarly to the case of the audio signal processing apparatus 11.

    [0136] Also, by suitably selecting a mapping function, the dynamic range in audio volume can be widened, kept the same, or the dynamic range can be narrowed by amplitude conversion.

    [0137] Once the operation in step S74, the operation in step S75 is subsequently conducted and the conversion process ends. However, since the operation in step S75 is similar to the operation in step S13 of Fig. 2, its description will be omitted.

    [0138] In so doing, the audio signal processing apparatus 91 conducts mapping processes on an input signal by using a plurality of mutually different mapping functions, performs weighted addition of the obtained plurality of output signals with distribution ratios obtained from the input signal characteristics analysis results, and takes the result as the final output signal.

    [0139] In this way, by selectively using mapping functions according to input signal characteristics analysis results to conduct mapping processes and generate an output signal, the playback level of an audio signal can be easily and effectively enhanced without requiring prior analysis or reading ahead of the input signal. Also, faint sounds that have been conventionally difficult to hear can be made easier to hear, while in addition, even sounds that have been comparatively loud conventionally become even easier to hear.

    [0140] Meanwhile, the audio signal processing apparatus 91 may likewise be configured such that the method for computing mapping control information or distribution ratios is varied according to the output device for the output signal, etc., so as to adjust the degree of enhancement to the playback level of an audio signal.

    [0141] Also, although the foregoing describes generating an output signal by computing mapping control information and distribution ratios for every single sample of an input signal, mapping control information and distribution ratios may also be computed for every two or more consecutive samples.

    <Fourth embodiment>


    [Configuration of audio signal processing apparatus]



    [0142] Also, although the case of a one-channel input signal as the audio signal was described with Fig. 8, an input signal may also be taken to have plural channels. For example, in the case where a two-channel input signal is input, the audio signal processing apparatus may take the configuration illustrated in Fig. 11.

    [0143] The audio signal processing apparatus 141 in Fig. 11 is composed of an analyzer 21, mapping processors 101-1 to 101-M, a weighting controller 102, an adder 103, mapping processors 151-1 to 151-M, a weighting controller 152, and adder 153, an output unit 23, and a drive 24. In Fig. 11 herein, like numerals are given to portions corresponding to the case in Fig. 8, and description of such portions will be reduced or omitted.

    [0144] Input signals are supplied to the audio signal processing apparatus 141 as a left-channel audio signal and a right-channel audio signal constituting a movie or other content, for example. In other words, the left-channel input signal is supplied to the analyzer 21 and the mapping processors 101-1 to 101-M, while the right-channel input signal is supplied to the analyzer 21 and the mapping processors 151-1 to 151-M.

    [0145] The analyzer 21 analyzes the respective characteristics of the supplied left- and right-channel input signals, generates mapping control information on the basis of the two sets of analysis results thus obtained, and supplies the mapping control information to the weighting controller 102 and the weighting controller 152.

    [0146] The mapping processors 151-1 to 151-M conduct mapping processes on a supplied input signal using the same respective mapping functions used by the mapping processors 101-1 to 101-M. Also, the mapping processors 151-1 to 151-M supply output signals obtained by the mapping processes to the weighting controller 152. Note that hereinafter, the mapping processors 151-1 to 151-M will also be simply designated the mapping processors 151 in cases where it is not necessary to individually distinguish them.

    [0147] The weighting controller 152 conducts the same operation as the weighting controller 102. In other words, multipliers 161-1 to 161-M constituting the weighting controller 152 correspond to the multipliers 111-1 to 111-M, multiplying output signals supplied from the mapping processors 151-1 to 151-M by distribution ratios α1 to αM and supplying the results to the adder 153. Note that hereinafter, the multipliers 161-1 to 161-M will also be simply designated the multipliers 161 in cases where it is not necessary to individually distinguish them.

    [0148] The adder 153 adds together M output signals supplied from the multipliers 161, and supplies the final output signal obtained as a result to the output unit 23.

    [Description of conversion process]



    [0149] Next, a conversion process conducted by the audio signal processing apparatus 141 will be described with reference to the flowchart in Fig. 12.

    [0150] In step S101, the analyzer 21 analyzes the characteristics of supplied left- and right-channel input signals. For example, the analyzer 21 may perform the computation in Eq. 1 discussed earlier, and compute a left-channel root mean square RMS(n) and a right-channel root mean square RMS(n).

    [0151] In step S102, the analyzer 21 generates mapping control information on the basis of the input signal characteristics analysis results, and supplies the mapping control information to the weighting controller 102 and the weighting controller 152. For example, the analyzer 21 may compute the average of the left-channel root mean square RMS(n) and the right-channel root mean square RMS(n), and take the obtained average to be the mapping control information.

    [0152] Once the operation in step S102 is conducted and mapping control information is generated, the operations in steps S103 to S106 are subsequently conducted and the conversion process ends. However, since these processing operations are similar to the operations in steps S72 to S75 of Fig. 9, their description will be reduced or omitted.

    [0153]  However, in steps S103 to S105, mapping processes are conducted in the mapping processors 101, the obtained output signals are multiplied by distribution ratios in the multipliers 111, the output signals which have been multiplied by the distribution ratios are added together in the adder 103, and the result is taken to be the final left-channel output signal. Similarly, mapping processes are conducted in the mapping processors 151, the obtained output signals are multiplied by distribution ratios in the multipliers 161, the output signals which have been multiplied by the distribution ratios are added together in the adder 153, and the result is taken to be the final right-channel output signal.

    [0154] In so doing, the audio signal processing apparatus 141 analyzes the characteristics of left- and right-channel input signals, generates common mapping control information for the left and right channels, and uses the obtained mapping control information to compute a common distribution ratio for the left and right channels for each mapping function. By using common mapping control information for the left and right channels to compute common distribution ratios for the left and right channels for each mapping function in this way, the playback level of an audio signal can be enhanced without changing the inter-channel volume balance.

    [0155] The series of processes discussed in the foregoing may be executed in hardware, but may also be executed in software. In the case of executing the series of processes in software, a program constituting such software is installed from a program recording medium to a computer built into special-purpose hardware, or a computer such as a general-purpose personal computer, for example, able to execute various functions by installing various programs thereon.

    [0156] Fig. 13 is a block diagram illustrating an exemplary hardware configuration of a computer that executes the series of processes discussed in the foregoing with a program.

    [0157] In a computer, a CPU (Central Processing Unit) 201, ROM (Read-Only Memory) 202, and RAM (Random Access Memory) 203 are connected to each other by a bus 204.

    [0158] An input/output interface 205 is additionally connected to the bus 204. Connected to the input/output interface 205 are an input unit 206 comprising a keyboard, mouse, microphone, etc., an output unit 207 comprising a display, speakers, etc., a recording unit 208 comprising a hard disk, non-volatile memory, etc., a communication unit 209 comprising a network interface, etc., and a drive 210 that drives a removable medium 211 such as a magnetic disk, an optical disc, a magneto-optical disc, or semiconductor memory.

    [0159] In a computer configured as above, the series of process discussed in the foregoing is conducted due to the CPU 201 loading a program recorded in the recording unit 208 into the RAM 203 via the input/output interface 205 and the bus 204, and executing the program, for example.

    [0160] A program executed by the computer (CPU 201) may for example be provided by being recorded onto a removable medium 211 as an instance of packaged media consisting of magnetic disks (including flexible disks), optical discs (including CD-ROMs (Compact Disc - Read-Only Memory), DVDs (Digital Versatile Disc), etc.), magneto-optical discs, or semiconductor memory. Alternatively, a program may be provided via a wired or wireless transmission medium such as a local area network, the Internet, or digital satellite broadcasting.

    [0161] Additionally, a program may be installed to the recording unit 208 via the input/output interface 205 by loading the removable medium 211 into the drive 210. Also, a program may be received by the communication unit 209 via a wired or wireless transmission medium and installed to the recording unit 208. Otherwise, a program may be installed in advance to the ROM 202 or the recording unit 208.

    [0162] Herein, a program executed by a computer may be a program in which operations are conducted in a time series following the order described in this specification, but may also be a program in which operations are executed in parallel or at required timings, such as upon being called.

    [0163] Furthermore, an embodiment of the present invention is not limited to the embodiments discussed in the foregoing, and various modifications are possible within a scope that does not depart from the principal matter of the present invention.

    Reference Signs List



    [0164] 

    11 audio signal processing apparatus

    21 analyzer

    22 mapping processor

    61 mapping processor

    101-1 to 101-M, 101 mapping processors

    102 weighting controller

    103 adder

    151-1 to 151-M, 151 mapping processors

    152 weighting controller

    153 adder




    Claims

    1. A signal processing apparatus (91) comprising:

    analyzing means (21) analyzing input signal characteristics;

    a plurality of mapping processing means (101) each conducting amplitude conversion of the input signal on the basis of one of a plurality of predetermined linear functions or one of a plurality of predetermined nonlinear functions, wherein the amplitude conversion comprises substitution of a sample value of each of a plurality of samples included of the input signal with an output of the one of the plurality of predetermined linear functions or the one of the plurality of predetermined nonlinear functions for which the each of the sample values is used as an input;

    weighting controlling means (102) respectively multiplying a plurality of the input signals, which have been respectively amplitude-converted on the basis of mutually different functions by a plurality of the mapping processing means, by weights determined by the analysis result for the input signal characteristics, wherein the weights are configured to increase for steeper ones of the plurality of predetermined linear functions or the plurality of predetermined nonlinear functions as a volume of the each of the samples of the input signal decreases; and

    adding means (103) generating an output signal by adding together the plurality of input signals which have been multiplied by the weights.


     
    2. The signal processing apparatus (91) according to Claim 1, wherein
    the analyzing means computes a value expressing the average sample value of samples included in a given segment of the input signal as the analysis result.
     
    3. The signal processing apparatus according to Claim 2, wherein
    the analysis result is the root mean square or a moving average of sample values of samples included in the given segment.
     
    4. The signal processing apparatus according to Claim 1, wherein
    in the case where amplitude conversion is conducted on the input signal for each of a plurality of channels to generate an output signal for each channel, the analyzing means computes one analysis result shared by all channels.
     
    5. The signal processing apparatus according to Claim 1, wherein
    the weights are determined by the analysis result for every single sample of the input signal.
     
    6. The signal processing apparatus according to Claim 1, wherein
    the weights are determined by the analysis result for every given number of two or more consecutive samples of the input signal.
     
    7. A signal processing method for a signal processing apparatus provided with
    analyzing means (21) analyzing input signal characteristics,
    a plurality of mapping processing means (101) each conducting amplitude conversion of the input signal on the basis of one of a plurality of predetermined linear functions or one of a plurality of predetermined nonlinear functions, wherein the amplitude conversion comprises substitution of a sample value of each of a plurality of samples included of the input signal with an output of the one of the plurality of predetermined linear functions or the one of the plurality of predetermined nonlinear functions for which the each of the sample values is used as an input,
    weighting controlling means (102) respectively multiplying a plurality of the input signals, which have been respectively amplitude-converted on the basis of mutually different functions by a plurality of the mapping processing means, by weights determined by the analysis result for the input signal characteristics, wherein the weights are configured to increase for steeper ones of the plurality of predetermined linear functions or the plurality of predetermined nonlinear functions as a volume of the each of the samples of the input signal decreases, and
    adding means (103) generating an output signal by adding together the plurality of input signals which have been multiplied by the weights,
    the signal processing method including the steps of:

    the analyzing means analyzing (S71) the input signal characteristics;

    the plurality of mapping processing means conducting (S72) amplitude conversion of the input signal;

    the weighting controlling means multiplying (S73) the amplitude-converted input signals by the weights determined by the analysis result; and

    the adding means generating the output signal by adding (S74) together the input signals which have been multiplied by the weights.


     
    8. A program causing a computer to execute processing including the steps of:

    analyzing (S71) input signal characteristics;

    conducting (S72) amplitude conversion of the input signal on the basis of one of a plurality of predetermined linear functions or one of a plurality of predetermined nonlinear functions, wherein the amplitude conversion comprises substitution of a sample value of each of a plurality of samples included of the input signal with an output of the one of the plurality of predetermined linear functions or the one of the plurality of predetermined nonlinear functions for which the each of the sample values is used as an input;

    respectively multiplying (S73) a plurality of the input signals, which have been amplitude-converted on the basis of a plurality of mutually different functions, by weights determined by the analysis result for the input signal characteristics, wherein the weights are configured to increase for steeper ones of the plurality of predetermined linear functions or the plurality of predetermined nonlinear functions as a volume of the each of the samples of the input signal decreases; and

    generating an output signal by adding (S74) together the plurality of input signals which have been multiplied by the weights.


     
    9. A data recording medium recording an output signal obtained by
    analyzing (S71) input signal characteristics,
    conducting (S72) amplitude conversion of the input signal on the basis of one of a plurality of predetermined linear functions or one of a plurality of predetermined nonlinear functions, wherein the amplitude conversion comprises substitution of a sample value of each of a plurality of samples included of the input signal with an output of the one of the plurality of predetermined linear functions or the one of the plurality of predetermined nonlinear functions for which the each of the sample values is used as an input,
    respectively multiplying (S73) a plurality of the input signals, which have been amplitude-converted on the basis of a plurality of mutually different functions, by weights determined by the analysis result for the input signal characteristics, wherein the weights are configured to increase for steeper ones of the plurality of predetermined linear functions or the plurality of predetermined nonlinear functions as a volume of the each of the samples of the input signal decreases, and
    adding (S74) together the plurality of input signals which have been multiplied by the weights.
     


    Ansprüche

    1. Signalverarbeitungsvorrichtung (91), umfassend:

    Analysemittel (21), die Eingangssignalcharakteristika analysieren;

    mehrere Abbildungsverarbeitungsmittel (101), die jeweils Amplitudenumwandlung des Eingangssignals auf Grundlage einer von mehreren vorbestimmten linearen Funktionen oder einer von mehreren vorbestimmten nichtlinearen Funktionen durchführt,

    wobei die Amplitudenumwandlung Ersetzung eines Abtastwerts von jedem von mehreren in dem Eingangssignal enthaltenen Abtastungen mit einer Ausgabe von der einen der mehreren vorbestimmten linearen Funktionen oder der einen der mehreren vorbestimmten nichtlinearen Funktionen, für welche der jeder der Abtastwerte als eine Eingabe verwendet wird, umfasst;

    ein Gewichtungssteuerungsmittel (102), das jeweils mehrere der Eingangssignale, welche jeweils auf Grundlage von gegenseitig anderen Funktionen durch mehrere der Abbildungsverarbeitungsmittel amplitudenumgewandelt wurden, mit durch das Analyseergebnis für die Eingangssignalcharakteristika bestimmten Gewichten multipliziert, wobei die Gewichte ausgelegt sind zum Erhöhen für steilere der mehreren vorbestimmten linearen Funktionen oder der mehreren vorbestimmten nichtlinearen Funktionen mit abnehmender Lautstärke von der jeder der Abtastungen des Eingangssignals; und

    ein Addiermittel (103), das ein Ausgangssignal erzeugt durch Aufaddieren der mehreren Eingangssignale, welche mit den Gewichten multipliziert wurden.


     
    2. Signalverarbeitungsvorrichtung (91) nach Anspruch 1, wobei
    das Analysemittel einen Wert berechnet, der den mittleren Abtastwert von in einem vorgegebenen Segment des Eingangssignals enthaltenen Abtastungen als das Analyseergebnis ausdrückt.
     
    3. Signalverarbeitungsvorrichtung nach Anspruch 2, wobei
    das Analyseergebnis der quadratische Mittelwert oder ein gleitender Mittelwert von Abtastwerten von in dem vorgegebenen Segment enthaltenen Abtastwerten ist.
     
    4. Signalverarbeitungsvorrichtung nach Anspruch 1, wobei
    in dem Fall, bei dem Amplitudenumwandlung an dem Eingangssignal für jeden von mehreren Kanälen durchgeführt wird, um ein Ausgangssignal für jeden Kanal zu erzeugen, das Analysemittel ein von allen Kanälen geteiltes Analyseergebnis berechnet.
     
    5. Signalverarbeitungsvorrichtung nach Anspruch 1, wobei
    die Gewichte durch das Analyseergebnis für jede einzelne Abtastung des Eingangssignals bestimmt werden.
     
    6. Signalverarbeitungsvorrichtung nach Anspruch 1, wobei
    die Gewichte durch das Analyseergebnis für jede vorgegebene Anzahl von zwei oder mehr aufeinander folgenden Abtastungen des Eingangssignals bestimmt werden.
     
    7. Signalverarbeitungsverfahren für eine Signalverarbeitungsvorrichtung, die versehen ist mit Analysemitteln (21), die Eingangssignalcharakteristika analysieren;
    mehreren Abbildungsverarbeitungsmitteln (101), die jeweils Amplitudenumwandlung des Eingangssignals auf Grundlage einer von mehreren vorbestimmten linearen Funktionen oder einer von mehreren vorbestimmten nichtlinearen Funktionen durchführt,
    wobei die Amplitudenumwandlung Ersetzung eines Abtastwerts von jedem von mehreren in dem Eingangssignal enthaltenen Abtastungen mit einer Ausgabe von der einen der mehreren vorbestimmten linearen Funktionen oder der einen der mehreren vorbestimmten nichtlinearen Funktionen, für welche der jeder der Abtastwerte als eine Eingabe verwendet wird, umfasst,
    einem Gewichtungssteuerungsmittel (102), das jeweils mehrere der Eingangssignale, welche jeweils auf Grundlage von gegenseitig anderen Funktionen durch mehrere der Abbildungsverarbeitungsmittel amplitudenumgewandelt wurden, mit durch das Analyseergebnis für die Eingangssignalcharakteristika bestimmten Gewichten multipliziert, wobei die Gewichte ausgelegt sind zum Erhöhen für steilere der mehreren vorbestimmten linearen Funktionen oder der mehreren vorbestimmten nichtlinearen Funktionen mit abnehmender Lautstärke von der jeder der Abtastungen des Eingangssignals, und
    einem Addiermittel (103), das ein Ausgangssignal erzeugt durch Aufaddieren der mehreren Eingangssignale, welche mit den Gewichten multipliziert wurden,
    wobei das Signalverarbeitungsverfahren die folgenden Schritte umfasst:

    das Analysemittel analysiert (S71) die Eingangssignalcharakteristika;

    die mehreren Abbildungsverarbeitungsmittel führen (S72) Amplitudenumwandlung des Eingangssignals durch;

    das Gewichtungssteuerungsmittel multipliziert (S73) die amplitudenumgewandelten Eingangssignale mit den durch das Analyseergebnis bestimmten Gewichten; und

    das Addiermittel erzeugt das Ausgangssignal durch Aufaddieren (S74) der Eingangssignale, welche mit den Gewichten multipliziert wurden.


     
    8. Programm, das einen Computer veranlasst zum Ausführen von Verarbeitung, beinhaltend die folgenden Schritte:

    Analysieren (S71) von Eingangssignalcharakteristika;

    Durchführen (S72) von Amplitudenumwandlung des Eingangssignals auf Grundlage einer von mehreren vorbestimmten linearen Funktionen oder einer von mehreren vorbestimmten nichtlinearen Funktionen, wobei die Amplitudenumwandlung Ersetzung eines Abtastwerts von jedem von mehreren in dem Eingangssignal enthaltenen Abtastungen mit einer Ausgabe von der einen der mehreren vorbestimmten linearen Funktionen oder der einen der mehreren vorbestimmten nichtlinearen Funktionen, für welche der jeder der Abtastwerte als eine Eingabe verwendet wird, umfasst;

    jeweiliges Multiplizieren (S73) von mehreren der Eingangssignale, welche auf Grundlage von mehreren gegenseitig verschiedenen Funktionen amplitudenumgewandelt wurden, mit durch das Analyseergebnis für die Eingangssignalcharakteristika bestimmten Gewichten, wobei die Gewichte ausgelegt sind zum Erhöhen für steilere der mehreren vorbestimmten linearen Funktionen oder der mehreren vorbestimmten nichtlinearen Funktionen mit abnehmender Lautstärke von der jeder der Abtastungen des Eingangssignals; und

    Erzeugen eines Ausgangssignals durch Aufaddieren (S74) der mehreren Eingangssignale, welche mit den Gewichten multipliziert wurden.


     
    9. Datenaufzeichnungsmedium, das ein Ausgangssignal aufzeichnet, das erhalten wurde durch
    Analysieren (S71) von Eingangssignalcharakteristika;
    Durchführen (S72) von Amplitudenumwandlung des Eingangssignals auf Grundlage einer von mehreren vorbestimmten linearen Funktionen oder einer von mehreren vorbestimmten nichtlinearen Funktionen, wobei die Amplitudenumwandlung Ersetzung eines Abtastwerts von jedem von mehreren in dem Eingangssignal enthaltenen Abtastungen mit einer Ausgabe von der einen der mehreren vorbestimmten linearen Funktionen oder der einen der mehreren vorbestimmten nichtlinearen Funktionen, für welche der jeder der Abtastwerte als eine Eingabe verwendet wird, umfasst,
    jeweiliges Multiplizieren (S73) von mehreren der Eingangssignale, welche auf Grundlage von mehreren gegenseitig verschiedenen Funktionen amplitudenumgewandelt wurden, mit durch das Analyseergebnis für die Eingangssignalcharakteristika bestimmten Gewichten, wobei die Gewichte ausgelegt sind zum Erhöhen für steilere der mehreren vorbestimmten linearen Funktionen oder der mehreren vorbestimmten nichtlinearen Funktionen mit einer abnehmenden Lautstärke von der jeder der Abtastungen des Eingangssignals, und
    Aufaddieren (S74) der mehreren Eingangssignale, welche mit den Gewichten multipliziert wurden.
     


    Revendications

    1. Appareil de traitement du signal (91) comprenant :

    des moyens d'analyse (21) analysant des caractéristiques d'un signal d'entrée ;

    une pluralité de moyens de traitement de mappage (101) effectuant chacun une conversion en amplitude du signal d'entrée sur la base de l'une d'une pluralité de fonctions linéaires prédéterminées ou de l'une d'une pluralité de fonctions non linéaires prédéterminées, la conversion en amplitude comprenant le remplacement d'une valeur d'échantillon de chacun d'une pluralité d'échantillons inclus dans le signal d'entrée par une sortie de l'une de la pluralité de fonctions linéaires prédéterminées ou de l'une de la pluralité de fonctions non linéaires prédéterminées pour laquelle chacune des valeurs d'échantillon est utilisée comme entrée ;

    des moyens de commande de pondération (102) multipliant respectivement une pluralité des signaux d'entrée, qui ont été respectivement convertis en amplitude sur la base de fonctions mutuellement différentes par une pluralité des moyens de traitement de mappage, par des poids déterminés par le résultat de l'analyse pour les caractéristiques du signal d'entrée, les poids étant configurés pour augmenter pour les plus fortes de la pluralité de fonctions linéaires prédéterminées ou de la pluralité de fonctions non linéaires prédéterminées lorsqu'un volume de chacun des échantillons du signal d'entrée diminue ; et

    des moyens d'addition (103) pour générer un signal de sortie par addition entre eux de la pluralité de signaux d'entrée qui ont été multipliés par les poids.


     
    2. Appareil de traitement du signal (91) selon la revendication 1, dans lequel
    les moyens d'analyse calculent une valeur exprimant la valeur d'échantillon moyenne des échantillons inclus dans un segment donné du signal d'entrée en tant que résultat d'analyse.
     
    3. Appareil de traitement du signal selon la revendication 2, dans lequel
    le résultat d'analyse est la moyenne quadratique ou une moyenne mobile des valeurs d'échantillon des échantillons inclus dans le segment donné.
     
    4. Appareil de traitement du signal selon la revendication 1, dans lequel
    dans le cas dans lequel une conversion en amplitude est effectuée sur le signal d'entrée pour chacun d'une pluralité de canaux pour générer un signal de sortie pour chaque canal, les moyens d'analyse calculent un résultat d'analyse partagé par tous les canaux.
     
    5. Appareil de traitement du signal selon la revendication 1, dans lequel
    les poids sont déterminés par le résultat d'analyse pour chaque échantillon individuel du signal d'entrée.
     
    6. Appareil de traitement du signal selon la revendication 1, dans lequel
    les poids sont déterminés par le résultat d'analyse pour chacun d'un nombre donné de deux échantillons consécutifs ou plus du signal d'entrée.
     
    7. Procédé de traitement du signal pour un appareil de traitement du signal avec
    des moyens d'analyse (21) analysant des caractéristiques d'un signal d'entrée,
    une pluralité de moyens de traitement de mappage (101) effectuant chacun une conversion en amplitude du signal d'entrée sur la base de l'une d'une pluralité de fonctions linéaires prédéterminées ou de l'une d'une pluralité de fonctions non linéaires prédéterminées, la conversion en amplitude comprenant le remplacement d'une valeur d'échantillon de chacun d'une pluralité d'échantillons inclus dans le signal d'entrée par une sortie de l'une de la pluralité de fonctions linéaires prédéterminées ou de l'une de la pluralité de fonctions non linéaires prédéterminées pour laquelle chacune des valeurs d'échantillon est utilisée comme entrée,
    des moyens de commande de pondération (102) multipliant respectivement une pluralité des signaux d'entrée, qui ont été respectivement convertis en amplitude sur la base de fonctions mutuellement différentes par une pluralité des moyens de traitement de mappage, par des poids déterminés par le résultat de l'analyse pour les caractéristiques du signal d'entrée, les poids étant configurés pour augmenter pour les plus fortes de la pluralité de fonctions linéaires prédéterminées ou de la pluralité de fonctions non linéaires prédéterminées lorsqu'un volume de chacun des échantillons du signal d'entrée diminue, et
    des moyens d'addition (103) pour générer un signal de sortie par addition entre eux de la pluralité de signaux d'entrée qui ont été multipliés par les poids,
    le procédé de traitement du signal incluant les étapes selon lesquelles :

    les moyens d'analyse analysent (S71) les caractéristiques du signal d'entrée ;

    la pluralité de moyens de traitement de mappage exécutent (S72) une conversion en amplitude du signal d'entrée ;

    les moyens de commande de pondération multiplient (S73) les signaux d'entrée convertis en amplitude par les poids déterminés par le résultat d'analyse ; et

    les moyens d'addition génèrent le signal de sortie en ajoutant (S74) entre eux les signaux d'entrée qui ont été multipliés par les poids.


     
    8. Programme amenant un ordinateur à exécuter un traitement incluant les étapes suivantes :

    analyse (S71) des caractéristiques du signal d'entrée ;

    exécution (S72) d'une conversion en amplitude du signal d'entrée sur la base de l'une d'une pluralité de fonctions linéaires prédéterminées ou de l'une d'une pluralité de fonctions non linéaires prédéterminées, la conversion en amplitude comprenant le remplacement d'une valeur d'échantillon de chacun d'une pluralité d'échantillons inclus dans le signal d'entrée par une sortie de l'une de la pluralité de fonctions linéaires prédéterminées ou de l'une de la pluralité de fonctions non linéaires prédéterminées pour laquelle chacune des valeurs d'échantillon est utilisée comme entrée ;

    multiplication respective (S73) d'une pluralité des signaux d'entrée, qui ont été convertis en amplitude sur la base d'une pluralité de fonctions mutuellement différentes, par des poids déterminés par le résultat de l'analyse pour les caractéristiques du signal d'entrée, les poids étant configurés pour augmenter pour les plus fortes de la pluralité de fonctions linéaires prédéterminées ou de la pluralité de fonctions non linéaires prédéterminées lorsqu'un volume de chacun des échantillons du signal d'entrée diminue ; et

    génération d'un signal de sortie par addition (S74) entre eux de la pluralité de signaux d'entrée qui ont été multipliés par les poids.


     
    9. Support d'enregistrement de données enregistrant un signal de sortie obtenu par
    analyse (S71) des caractéristiques du signal d'entrée ;
    exécution (S72) d'une conversion en amplitude du signal d'entrée sur la base de l'une d'une pluralité de fonctions linéaires prédéterminées ou de l'une d'une pluralité de fonctions non linéaires prédéterminées, la conversion en amplitude comprenant le remplacement d'une valeur d'échantillon de chacun d'une pluralité d'échantillons inclus dans le signal d'entrée par une sortie de l'une de la pluralité de fonctions linéaires prédéterminées ou de l'une de la pluralité de fonctions non linéaires prédéterminées pour laquelle chacune des valeurs d'échantillon est utilisée comme entrée ;
    multiplication respective (S73) d'une pluralité des signaux d'entrée, qui ont été convertis en amplitude sur la base d'une pluralité de fonctions mutuellement différentes, par des poids déterminés par le résultat de l'analyse pour les caractéristiques du signal d'entrée, les poids étant configurés pour augmenter pour les plus fortes de la pluralité de fonctions linéaires prédéterminées ou de la pluralité de fonctions non linéaires prédéterminées lorsqu'un volume de chacun des échantillons du signal d'entrée diminue ; et
    addition (S74) entre eux de la pluralité de signaux d'entrée qui ont été multipliés par les poids.
     




    Drawing












































    Cited references

    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description