JPH1091199A - Recording and reproducing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable reducing memory quantity required for recording and reproducing a signal by converting a frequency of an input signal, eliminating the prescribed region, storing it, and returning it to time series data assuming that a frequency component of a eliminated region is zero at the time of reproducing. SOLUTION: This device has an AD converter 21, a DA converter 22, a DSP 23, and a memory 24. A frequency component of a digitized input signal is analyzed. The prescribed region in a frequency component of the analyzed input signal is extracted and eliminated. A residual signal component of the input signal after extraction is stored as a recording signal with its region value. This stored recording signal is read out assuming that the recording signal is zero in the prescribed region, and converted to time series data. Also, data of a frequency region of which a frequency component is the prescribed level or less is eliminated at the time of extraction and elimination. Data of a frequency range previously specified is eliminated at the time of extraction and elimination. Further, the prescribed region is eliminated based on priority decided separately at the time of extraction and elimination.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a recording / reproducing apparatus for recording / reproducing a signal having a constant frequency spectrum such as an audio signal with a small storage capacity.

[0002]

2. Description of the Related Art FIG. 17 is "Toshiba 1991 Edition Semiconductor Data Book" Sound LSI "", pp. 230-233.
1 is a diagram illustrating an outline of a system configuration of a recording / reproducing apparatus configured based on the voice recording / reproducing LSI shown in FIG. In FIG. 17, 2 is an SRAM (static RAM) for recording an audio signal, 3 is a microphone for inputting an audio signal, 4 is an amplifier for amplifying an output audio signal, and 5 is a speaker for outputting an audio signal.

Further, reference numeral 6 denotes a microphone amplifier for amplifying an audio signal inputted from the microphone 3, reference numeral 7 denotes an ADM analysis / synthesis circuit for converting an analog signal through the microphone amplifier 6 into digital data and data compression, and reference numeral 8 denotes an input / output to and from the SRAM 2. Do S
A RAM interface 9 converts a digital signal output from the ADM analysis / synthesis circuit 7 into an analog signal.
A / A converter 10, a band-pass filter for passing a signal of a predetermined frequency band from a signal output from the D / A converter 9, and 11 an SRAM interface 8 or A
This is a timing generator control circuit that sends a control signal to the DM analysis / synthesis circuit 7.

The operation of the conventional LSI for a voice recording / reproducing apparatus having the above-described configuration will be described. First, the operation at the time of recording an audio signal will be described. An analog input signal input from a speaker via the microphone 3 is amplified by the microphone amplifier 6. Then, the signal is converted into a digital signal by the ADM analysis / synthesis circuit 7 and further compressed according to a predetermined rule. Then, the compressed signal is
SRA chronologically as it is via the M interface 8
Recorded in M2. These operations are performed according to a control signal from the timing generator control circuit 11.

Next, the operation at the time of reproducing a recording signal will be described. At the time of reproduction, digital data recorded from the SRAM 2 is read out via the SRAM interface 8 according to a control signal from the timing generator control circuit 11. Then, the read signal is A
The signal is converted into an analog signal by a D / A converter 9 via a DM analysis / synthesis circuit 7, a predetermined frequency component is extracted by a bandpass filter 10, and output from a speaker 5 amplified by an amplifier 4.

[0006]

Since the conventional audio recording / reproducing apparatus is configured as described above, the audio signal to be recorded is subjected to ADM analysis as it is and is somewhat compressed. Since the portion is recorded in the SRAM 2, there is a problem that the SRAM 2 is required in proportion to the recording time.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and has as its object to provide a recording / reproducing apparatus which requires a small amount of memory for recording signals.

[0008]

According to the present invention, there is provided a recording / reproducing apparatus for analyzing a frequency component of a digitized input signal, and extracting a specific area in the frequency component of the analyzed input signal. Extracting and removing means for removing the residual signal component of the input signal after the extraction as a recording signal together with its area value; and determining from the stored recording signal that the recording signal is 0 in the specific area. And inverse frequency conversion means for supplementarily reading and converting the data into time series data.

Further, the extraction and removal means is low-level removal means for removing data in a frequency region whose frequency component is lower than a predetermined level.

Furthermore, the extraction and removal means is a designated frequency range removal means for removing data in a frequency range designated in advance.

Further, the extraction and removal means is a recording warning removal means for receiving memory information as input, outputting a warning as to whether or not recording is possible based on the memory amount, and extracting and removing a specific area with a recordable signal.

Further, the extraction and removal means is a priority removal means for storing separately determined priorities and removing a specific area based on the priorities.

[0013] Furthermore, the extraction and removal means determines the removal area so that the sum of the data numbers of the remaining signal components is constant.

Further, the extraction and removal means is configured to further define the removal area so that the sum of the powers of the remaining signal components becomes constant.

Further, the extraction and removal means passes the frequency components around the local maximum detected in response to the result of the frequency component analysis and removes components in other regions.

According to the present invention, there is provided a recording / reproducing apparatus for analyzing a frequency component of a digitized input signal, and extracting and removing a specific area in a frequency component of the analyzed input signal. Means, a first inverse frequency transforming means for informing the memory of the removal area, and inversely frequency transforming a virtual short-circuit frequency area signal obtained by adjoining the remaining frequency area adjacent to the low frequency side, and the inverse transformation A memory for storing the virtual time series signal and the elimination area, a frequency conversion unit for reading and converting the frequency of the virtual time series signal in the memory, a virtual frequency signal having the frequency converted, A second inverse frequency conversion means for obtaining a frequency domain value from the specific domain value and restoring the time-series data is provided.

Further, a data compression algorithm for compressing the time series signal and a data expansion algorithm for expanding and restoring the compressed time series data are added, and the virtual time series data is compressed and stored using the data compression algorithm. In the case of reproduction, the compressed and stored virtual time series data is reproduced using a data decompression algorithm.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1 FIG. FIG. 1 shows a recording / reproducing apparatus according to Embodiment 1 of the present invention. In FIG. 1A, reference numeral 21 denotes an AD converter, 2
2 is a DA converter, 23 is a DSP, and 24 is a memory.
FIG. 1B is a flowchart showing a data processing operation at the time of recording and reproducing an input signal performed by the apparatus having the configuration shown in FIG. FIG. 2 shows an example of a signal waveform when an audio signal is taken as an example of the analog input signal A1 to be recorded / reproduced in the present invention. In FIG.
t _s is a sampling interval when this signal is AD-converted and taken in as a digital signal. Also, FIG.
3 illustrates an example of a frequency spectrum when the analog input signal A1 in FIG. 2 is frequency-converted. Further, FIG. 3 is also a diagram for explaining the operation of the device of the present embodiment, in which a frequency region below the level of C in FIG. 3 is removed and recorded. FIG.
In, f _s is related to f _s = 1 / t _s is the aforementioned t _s, the spectrum distribution of the input signal, the sampling theorem, it is necessary that are limited to f _s / 2 or less.
In addition, the following f ₁ to a level below the frequency components of C in FIG. 3, between f _l1 and f _h1, between f ₂ and f _3, f ₄ or more frequencies is not recorded. FIG. 4 is an explanatory diagram of the state of recording in the memory.

The operation of the recording / reproducing apparatus according to the first embodiment configured as described above will be described. First, the case of recording will be described. The analog input signal A1 is sampled at a sampling interval t _s ,
Is converted into a digital data series D1. DSP23
Then, the frequency spectrum from D1 to A1 is analyzed.
This is, for example, an arithmetic operation such as a discrete Fourier transform (DFT). Here, _assuming that DFT is used, the result is obtained as a Fourier coefficient for a frequency of 1 / ts, and when the digital data sequence D1 is composed of n data (the number of sample data n), f ₁ = i / T
When _s, for each frequency to f ₀ ~f n / _2-1,
Each of the two data _ai and bi _gives a total of n data.

Here, when the frequency spectrum of the input signal S1 is, for example, between f _{11 and} f _h1 shown in FIG. 3 and the intensity is equal to or lower than the set level C, Fourier coefficient data in this range is extracted and deleted. I do. Then, m = h _i −l _i −1
When, as shown in FIG. 4, the address 0 of the address of the memory 24 to _{a 0 ~a 11, b 0 ~b} 11 until l _i addresses, from l _i +1 address to n-m address a _h1 ~
Write a _{n / 2-1} and b _{h1 to} b _{n / 2-1} . That is, the data after the removal is packed and written into the memory. This is shown in FIG.
Steps S3 to S7 of (b) are repeated. For example, assuming that a human voice signal is used as the input signal S1, it is known that the input signal S1 has two or three spectrum peaks generally called a fault mant. The component strength is not very large. The removed frequency domain values, that is, the values of _fl1 to _fh1 themselves are separately stored in a memory.

Next, the case of reproduction will be described. During playback, the data that has been stored in the memory 24, the Fourier coefficients a ₀ ~a content from address 0 to l ₁ address
₁₁ , b _{0 to} b ₁₁ , l ₁ +1 to the mn address are read as the Fourier coefficients a _{h1 to an / 2-1} and b _{h1 to} b _{n / 2-1} , while the DSP 23 performs discrete inverse. Perform Fourier transform (inverse DFT) operation. At this time, the Fourier coefficients corresponding to the frequency domain removed at the time of recording have _{a11 + 1 to ah1-1.}
And b _{11 + 1 to} b _h1-1 are set to 0 to supplement the data. This means that steps S13 to S17 in FIG. 1B are repeated. These obtained results are referred to in step S18.
To perform the inverse frequency conversion to obtain time-series data. The result is
It becomes a digital data series D2 of data number n,
The signal is reproduced as an analog signal S2 by the A converter 22.

As described above, the capacity of the memory 24 can be reduced by using the low-level removing means for removing m data of frequency components whose intensity (level) is not so strong. In addition, since the number of data to be calculated is reduced, the number of calculation processes is reduced, and the calculation time and current consumption of the DSP can be reduced. In the above description, only the operation of the partial region f _l1 in Figure 3 to f _h1, the same operation applies to the area to be other removal.

Embodiment 2 FIG. FIG. 5 is a configuration diagram of a recording / reproducing apparatus according to Embodiment 2. In FIG.
Reference numerals 11 and 312 denote setting registers for setting the frequency regions l ₁ and h ₁ in the _first embodiment from the outside, respectively. That is, in the present embodiment, unnecessary frequency domain values are specified in advance. As described above, the specific region may be designated by the level, or may be designated by the frequency using the fact that the spectrum of the input signal is generally biased.

The recording / reproducing apparatus having the specified frequency range removing means configured as shown in FIG. 5 performs the same operation as the apparatus of the first embodiment. That is, at the time of recording, after obtaining the Fourier coefficients, the data is removed by referring to the registers 311 and 312 and stored in the memory 24. At this time, the contents of the registers 311 and 312 may also be stored in the memory 24. Registers 311 and 312 set a range in which the frequency component is considered to be small beforehand in accordance with the characteristics of input signal S1 at that time before data deletion.

At the time of reproduction, the operation is the same as that of the first embodiment. However, when data is deleted in the memory 24, the range of the address is referred to by referring to the contents of the address. If it is left in the registers 311 and 312, it is referred to.

As described above, since the range to be deleted can be set from the outside, even if the analog input signal A1 can be any of several types of signals having different characteristics (frequency spectrum tendencies), the signal at that time is If the type is known, recording and reproduction can be performed without deteriorating the quality without deteriorating the frequency component data of the necessary band. For example, assuming a human voice signal, if it is known whether the voice of the speaker is high or low, a designated frequency removing unit that sets a range to be deleted can be used.

Embodiment 3 A recording / reproducing apparatus that issues a warning as to whether or not recording is possible based on the memory capacity will be described. The apparatus according to the present embodiment has the hardware configuration shown in FIG. 1, but performs the operation shown in the flowchart of FIG. Note that the processing at the time of reproduction is the same as that of the apparatus of the first embodiment, and the description is omitted.

The processing flow of an apparatus having a removing means for issuing a recording warning will be described below. At the time of recording, step S2
3, the remaining amount of the memory is checked. If the remaining amount of the memory 24 is small, if the frequency component data is to be deleted so as to fit in the remaining amount, the deletion range extends to a place where the frequency component is large. When you play
This is far from the original signal S1.

Therefore, in such a case, step S
At 28, the user is notified that the remaining memory capacity is too small. Then, when there is a signal indicating that the user can input in step S30, the data is deleted and recorded in step S27. As described above, by using the recording warning removing unit, when the remaining amount of the memory 24 is reduced without the user knowing, the recording quality is extremely reduced,
It is possible to avoid noticing that the recording is not normal only at the time of reproduction.

Embodiment 4 A recording / reproducing apparatus in which the amount of memory for recording is reduced by changing the deletion range depending on the amount of memory or externally designated information will be described.
The hardware configuration of the recording / reproducing apparatus of the present embodiment is the same as that of FIG. However, the deletion range specification register 311,
The number of 312 sets is increased and priorities are assigned. FIG. 7 is a diagram for explaining the flow of processing in the case of recording by the apparatus of the present embodiment and the state of recording in the memory. In the figure, D3i indicates that it is the i-th data series when a plurality of input signals are recorded. Here, k kinds of signals have already been recorded, and the number of remaining signals has been reduced.

Next, the flow of the process shown in FIG. 7 will be described. First, at the time of recording, the frequency component is analyzed, and thereafter, the remaining capacity of the memory 24 is checked. That is, every time one signal (a series of digitally changed and data deleted) is recorded, the remaining amount of the memory 24 is recorded in the control circuit or stored at a specific address of the memory 24. The remaining amount can be known from the above.

If the remaining capacity of the memory 24 is large, the deletion range of the frequency component data is reduced in step S34 in combination with the priority information so as not to lower the recording quality, and the remaining capacity of the memory 24 is small. In such a case, the deletion range is determined so as to increase the deletion range. The determined deletion range is stored in the memory 24 as in the other embodiments.

Since the reproducing operation is the same as that of the other embodiments, the description is omitted. As described above, the recording quality can be kept optimal according to the remaining memory capacity.

A description will be given of a case where the priority order is determined based on a frequency level, that is, a part to be removed as a specific area is determined based on the component intensity. FIG. 8 is a flowchart showing the flow of processing of another recording / reproducing apparatus of the present embodiment. Hereinafter, only the operation at the time of recording having a different operation will be described.

The frequency component of the analog input signal A1 is analyzed.
That is, find the valley of the spectrum. Then, the control circuit 25 or the DSP 23 determines the deletion range of the frequency component data based on the result, based on the priority in step S34a. Thereafter, data is deleted according to the determined deletion range, and the memory 2 is deleted together with the deletion range.
4 is stored.

As described above, in the apparatus having the priority removing unit, the deletion range is determined according to the spectrum of the analog input signal A1, so that the deterioration of the recording quality due to the data deletion can be kept small.

Embodiment 5 A recording / reproducing apparatus will be described which focuses on a frequency spectrum of an input signal, narrows a recording range by using a band-pass filter that passes only a predetermined range, and further effectively uses a memory by using another algorithm. FIG. 9 is a diagram for explaining the operation of the recording / reproducing apparatus of the present embodiment. FIG. 10 is a processing flowchart based on the algorithm used in combination. In FIG.
The frequency range to be removed as the specific region is previously set to f _cHPF or less, between f _l1 to f _h1 , and f _l2 to f corresponding to the input signal.
_The interval between _h2 and _fcLPF or more is defined.

The spectrum 41 shown in FIG.
Represents the spectrum of the analog input signal A1. On the other hand, some band rejection filters (BE
F), a high-pass filter (HPF), and a low-pass filter (LPF) that have passed through are the spectrum 42 in FIG. 9B. F _li and f _hi are the i-th B
This represents the cutoff frequency on the low frequency side and high frequency side of the EF. Further, f
_cHPF indicates the cutoff frequency of the HPF, and _fcLPF indicates the cutoff frequency of the LPF.

Passing these filters after the frequency components have been analyzed is equivalent to zeroing the intensity of the frequency components in the non-pass band (rejection band) of the filter. Then, the data in this range is deleted and recorded in the memory 24 together with each parameter (cutoff frequency of each filter) indicating the deletion range.

Although the above range may be kept as limited data, here, while keeping the sum of the frequency components to be deleted constant, the cutoff frequency of each filter is changed, and the sum of the intensities of the remaining frequency components is minimized. To be bigger,
Determine the cutoff frequency. For example, when the frequency component is analyzed using DFT, the cutoff frequency of the filter is a discrete value of the order, so that the combination of the cutoff frequencies of each filter is valid, and among all the combinations, A method of selecting a combination that maximizes the intensity of the residual component is conceivable.

Alternatively, a method shown in FIG. 10 may be used. That is, first, the cut-off frequency of the LPF is lowered, and the component intensity in the pass band is set to a constant ratio with respect to the original component intensity, and then, the cut-off frequency of the HPF is increased. The setting is performed in the same manner, and subsequently, the BEF blocking area is sequentially set. Note that if the number of frequency components remaining as a result of deletion during the setting of the cutoff frequencies of the LPF and HPF falls below a specified value determined from the remaining memory capacity and the like, the setting of the cutoff frequency ends there. By doing so, it is possible to prevent data from being deleted more than necessary. Of course, it is not limited to these methods.

As described above, since the deletion range is determined according to the spectrum of the analog input signal A1, the deterioration of the recording quality due to the data deletion can be kept small. In addition, by keeping the number of frequency components constant, the memory consumption can be kept constant regardless of the input data, and management becomes easy.

In the above embodiment, as a result of the frequency component data deletion, the cutoff frequency of each filter is determined so that the sum of the remaining data numbers becomes constant. Is also good. FIG. 11 shows an example of the flow of a process of determining the cutoff frequency of the filter when recording is performed so that the total of the stored power components, that is, the remaining frequency and the area of the component are constant. The cutoff frequencies of the LPF and HPF are determined, and if the number of data after passing through each filter becomes equal to or less than a specified value determined from the remaining amount of memory or the like, further deletion of data is stopped. Here, every time the cutoff frequency of each filter is changed, the total intensity of the remaining components is obtained as a result of the component intensity data deletion, and the LPF, LPF,
The cutoff frequencies of HPF and BEF are sequentially determined. Of course, the method is not limited to this method as long as the total intensity after data deletion is kept constant. This is effective for an input signal whose recording quality is degraded unless data whose intensity sum is equal to or more than a certain value is secured.

Embodiment 6 FIG. Another recording / reproducing apparatus from which the amount of memory has been deleted will be described. If the input signal has a certain frequency spectrum, only the frequency range of the characteristic spectral range may be recorded. FIG. 12 is a diagram illustrating the operation of the recording / reproducing apparatus according to the present embodiment. In the figure, spectra 41 and 42 are the same as those shown in FIG. F _oi representing each parameter
And f _wi represent the center frequency and pass bandwidth of the i-th band-pass filter (BPF).

FIG. 13A is an operation flowchart showing a processing flow of the apparatus of this embodiment. After analyzing the component intensities, first, the peak of the spectrum is obtained. For example, there is a method of obtaining a smooth envelope by calculating a moving average on the frequency axis for each component intensity, and obtaining a peak thereof. Further, a method of estimating a spectrum by an AR method or the like to obtain a peak may be adopted, or various other methods are conceivable.

Next, the peaks of the spectrum obtained in this way are set as the center frequency of the BPF having a pass band of a fixed width in order from the one with the highest intensity, and the components of the pass band of all the filters are set to the center frequency of each filter. With memory 24
To memorize. As described above, since the deletion range is determined according to the spectrum of the analog input signal A1, the deterioration of the recording quality due to the data deletion can be kept small.

The deletion range may be determined by another procedure. That is, FIG. 13B is an operation flowchart illustrating a flow of another process according to the present embodiment. First,
The peak of the spectrum is obtained, and this is set to the center frequency of each BPF. Here, the pass band width f _wi of each BPF is further determined from the analysis result of the frequency components, and only the component intensity data within those pass bands is stored in the memory 24 together with the center frequency and the pass band width. This is shown in FIG. The method of determining f _wi is as follows.
A method of keeping the sum of the components to be stored or the total intensity (power) of the components constant may be used together.

Embodiment 7 FIG. Another unique recording / reproducing apparatus from which the amount of memory has been deleted will be described. That is, the specific area to be removed is deleted, the remaining frequencies are collected on the low frequency side, a virtual short-circuit frequency signal is generated, and this signal is converted back to a time-series signal and recorded. Reproduction follows this reverse process. FIG.
FIG. 5 is a processing explanatory diagram of a frequency spectrum for explaining an operation of the recording / reproducing apparatus in the present embodiment. FIG.
FIG. 5 is a flowchart showing a processing operation at the time of recording of the apparatus of the present embodiment.

The operation of the above device will be described. At the time of recording, the frequency component data is deleted in step S63 as in the case of the apparatus of the other embodiments. Thereafter, the part of the spectrum that has been skipped in step S64 (the part with zero power) is extracted, and the remaining frequency regions A, B, and C of FIG. Generate a signal with

Subsequently, in step S65, a virtual time-series signal is obtained from the spectrum 43 in FIG. 14B (if the DFT is performed first, the first inverse DFT is performed). Then, this virtual time-series signal has a bandwidth f
_Since the signal is _s ′ / 2, the sampling interval t _s ′ =
If Shiteyare sampled at 1 / f _s', does not any more data is lost. Therefore, in step S66, if time-series data is thinned out or interpolated so as to perform sampling at an interval corresponding to t _s ′, and sampling is performed again, data smaller than the sampling number of the input signal S1 is obtained. It becomes a number. This data is stored in the memory 24 together with the parameters when the frequency component was deleted.
To memorize.

At the time of reproduction, the spectrum 43 is first converted from the time series data stored in the memory 24 by frequency conversion.
To reproduce. This virtual frequency signal is returned to the position of the spectrum 41 using the parameter at the time of deleting the component intensity data. Further, original time-series data is obtained from the frequency component data by the second inverse frequency conversion, and is sequentially output from the DA converter.

As described above, the capacity of the memory 24 can be reduced by deleting (not storing) the data of the frequency component whose intensity is not so strong. In addition, since the number of data to be calculated is reduced, the number of calculation processes is reduced, and the calculation time and current consumption of the DSP can be reduced.

In the above-described deletion of the memory amount by generating the virtual time-series signal, the memory amount can be further reduced by using another data compression algorithm. FIG.
FIG. 7 is a flowchart of processing of the recording / reproducing apparatus of the present embodiment.

The operation is as follows. After the data of the frequency components whose intensity is not so strong is deleted, they are gathered to the lower frequency side, returned to the time series, and sampled again according to the narrowed band. It is set as series data. Here, in step S67, a general data compression algorithm is used for this time-series data,
Further, data compression is performed. For example, a method such as ADPCM can be considered. Then, the compressed data is stored in the memory 24 in the same manner as the parameter at the time of deleting the frequency component.

At the time of reproduction, time-series data whose bandwidth has been narrowed is reproduced by an algorithm corresponding to that used at the time of recording, and the same processing as that of the above-described apparatus may be performed thereafter.

[0056]

According to the present invention, the input signal is frequency-converted to remove and store a specific area, and at the time of reproduction, the frequency component of the removed area is compensated for as 0 and returned to time-series data. This has the effect of reducing the amount of memory required for recording and reproducing signals.

Further, since the designated frequency region is removed, even if the input signal is a signal of several types having characteristics different in frequency spectrum, if the type is known, the necessary band of the required band is removed. There is an effect that recording and reproduction can be performed without damaging the frequency component data.

Further, since a recording warning is issued according to the amount of memory, there is an effect of avoiding the danger of poor quality recording even when the remaining amount of memory is reduced.

Further, since the removal area of the input signal is determined based on the specified priority order, there is an effect that the deterioration of the recording quality due to the data deletion is kept small.

Further, since the record collection defines the number range so that the sum of the remaining data numbers is constant, deterioration of the recording quality due to data deletion is reduced, and the memory consumption is constant regardless of the input data. This also has the effect of facilitating management.

Further, since the recording frequency range is determined so that the residual power sum is constant, deterioration of the recording quality due to data deletion is reduced, and the recording quality is kept constant to some extent regardless of input data. There is also an effect that can be done.

Further, the input signal is subjected to frequency conversion, and after removing a specific region, a virtual time-series signal obtained by converting a virtual signal collected on the low frequency side back to time-series data is stored. Since the reproduction is performed according to the procedure described above, there is an effect that the amount of memory can be reduced.

[Brief description of the drawings]

FIG. 1 is a flowchart of a hardware configuration and a data processing operation of a recording / reproducing apparatus according to Embodiment 1 of the present invention.

FIG. 2 is a diagram illustrating an example of an analog input signal.

FIG. 3 is a diagram illustrating an example of a frequency spectrum of an input signal and an operation of the device according to the first embodiment.

FIG. 4 is an explanatory diagram of memory writing of the device of FIG. 1;

FIG. 5 is a hardware configuration diagram of a recording and reproducing device according to a second embodiment of the present invention.

FIG. 6 is a flowchart showing a processing flow of the recording / reproducing apparatus according to Embodiment 3 of the present invention.

FIG. 7 is a flowchart illustrating a processing flow of a recording / reproducing apparatus according to Embodiment 4 of the present invention and a diagram illustrating a memory usage state.

FIG. 8 is a flowchart showing a flow of processing of another recording / reproducing apparatus according to the fourth embodiment.

FIG. 9 is a diagram illustrating a spectrum and parameters for explaining an operation of the recording / reproducing apparatus according to the fifth embodiment of the present invention.

FIG. 10 is a flowchart showing a processing flow of a recording / reproducing device in a fifth embodiment.

FIG. 11 is a flowchart showing a processing flow of another recording / reproducing apparatus in the fifth embodiment.

FIG. 12 is a diagram illustrating the spectrum and operation of an input signal of a recording / reproducing device according to a sixth embodiment of the present invention.

FIG. 13 is a flowchart showing a processing flow of the recording / reproducing apparatus in the sixth embodiment.

FIG. 14 is a frequency spectrum diagram illustrating an operation of the recording / reproducing apparatus according to Embodiment 7 of the present invention.

FIG. 15 is a flowchart showing a processing flow of the recording / reproducing apparatus according to Embodiment 7 of the present invention.

FIG. 16 is a flowchart showing a flow of another recording / reproducing device process in the seventh embodiment.

FIG. 17 is a system configuration diagram of a conventional recording / reproducing apparatus including a voice recording / reproducing LSI.

[Explanation of symbols]

21 AD converter, 22 DA converter, 23 DSP,
24 memory, 25 control circuits, 311 and 312 registers, S1 frequency conversion step, S3 specific area detection step, S4 frequency domain value and data value storage step,
S11 Step of reading data from memory, S13
Specific area detection step, S14 Frequency value and data setting step, S15 Data 0 value complementation step, S18
Inverse frequency conversion step, S23 memory remaining amount checking step, S24 warning step to user, S34
Data removal range determination step based on priority, S52
Spectral peak value detection step, S53 storage bandwidth determination step, S64 low frequency continuous short frequency spectrum generation step, S65 time series data restoration step, S67 data compression algorithm application step.

Claims (10)

[Claims] 1. A frequency component analyzing means for analyzing a frequency component of a digitized input signal; an extracting and removing means for extracting and removing a specific region in a frequency component of the analyzed input signal; A memory for storing the remaining signal component after the extraction as a recording signal together with its area value, and an inverse frequency for reading out the recording signal from the stored recording signal assuming that it is 0 in the specific area and converting it into time-series data A recording / reproducing apparatus provided with a conversion unit. 2. The recording / reproducing apparatus according to claim 1, wherein the extraction and removal means is a low-level removal means for removing data in a frequency region whose frequency component is lower than a predetermined level. 3. The recording / reproducing apparatus according to claim 1, wherein the extraction and removal means is a designated frequency range removing means for removing data in a frequency range designated in advance. 4. The extraction and removal means is a recording warning removal means which receives memory information as input, outputs a warning on whether or not recording is possible based on the memory amount, and extracts and removes a specific area with a recordable signal. The recording / reproducing apparatus according to claim 1, wherein 5. The recording / reproducing apparatus according to claim 1, wherein the extraction and removal means is a priority removal means for storing a separately determined priority and removing a specific area based on the priority. . 6. The recording / reproducing apparatus according to claim 2, wherein said extraction and removal means further defines a removal area so that the sum of the data numbers of the remaining signal components is constant. 7. The recording / reproducing apparatus according to claim 2, wherein the extraction and removal means further defines a removal area so that the sum of the powers of the remaining signal components becomes constant. 8. The method according to claim 2, wherein the extraction and removal means passes the frequency components around the local maximum detected in response to the frequency component analysis result and removes components in other regions. Recording and playback device. 9. A frequency component analyzing means for analyzing a frequency component of a digitized input signal; an extracting and removing means for extracting and removing a specific area in a frequency component of the analyzed input signal; First inverse frequency conversion means for informing the memory and inversely frequency-converting a virtual short-circuit frequency domain signal obtained by making the remaining frequency domain adjacent and continuous on the low frequency side; and the inverse-transformed virtual time-series signal And a memory for storing the removal area, frequency conversion means for reading and converting the frequency of the virtual time-series signal in the memory, the frequency-converted virtual frequency signal, and the stored specific area value to be removed. A recording / reproducing apparatus comprising second inverse frequency conversion means for obtaining a frequency domain value from and recovering time-series data. 10. A data compression algorithm for compressing a time-series signal and a data decompression algorithm for decompressing and decompressing the time-series data after compression, and compressing and storing virtual time-series data using the data compression algorithm. 10. The recording / reproducing apparatus according to claim 9, wherein in the case of reproduction, the compressed and stored virtual time-series data is reproduced using the data expansion algorithm.