JPS63126309A - Octave multiplex filter - Google Patents

Abstract

PURPOSE:To designate the characteristic at a logarithmic interval over a wide band of a frequency region by providing a couple of digital filters having band pass and low-pass characteristics in parallel between input and output. CONSTITUTION:An input signal is given from an input terminal IN to an A-D converter A/D. Digital filters S0-S4 of band-pass characteristics have respectively a delay element operated at one time interval among an integral number of time intervals T, 2T, 3T... (T is a reference time interval) of the sampling time interval of the converter A/D. Digital filters F0-F4 having the low-pass characteristic have respectively a delay element operated by a delay time being one over an integral number of the delay time per one stage of the delay element of the corresponding filters S0-S4 in the delay time per one stage. Moreover, each filter stage is coupled via adder circuits D0-D5. Thus, the characteristic is designated at a logarithmic interval over a wide band of the frequency region.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an octave multiplex filter constructed by regularly combining digital filters with low-pass characteristics and band-pass characteristics.

[Summary of the Invention] The octave multiplex filter of the present invention has, for example, a plurality of filter stages each consisting of a pair of digital filters (FIR type filters) each having a band-pass characteristic and a low-pass characteristic. The octave multiple filters installed in parallel make it possible to specify characteristics at logarithmically uniform intervals over a wide band in the frequency domain.

[Prior art] Conventional digital filters operate at a predetermined clock frequency in order to process discrete sample values at equal time intervals, so characteristics in the frequency domain are specified at equal frequency intervals. I was worried.

For this reason, conventional filters that specify characteristics at logarithmic intervals over a wide band have only been realized by combining a plurality of analog filters.

[Problems to be Solved by the Invention] As mentioned above, the former digital filter uses a method of specifying characteristics at equal frequency intervals, so the resolution in the low range is too coarse, while the resolution in the high range is too fine than necessary. , it is extremely difficult to widen the bandwidth.

Generally, the characteristics of audio equipment are evaluated or specified in the frequency domain using logarithmic intervals. Therefore, when using conventional digital filters, such as FIR type (finite impulse response) filters, the characteristics are accurate in the low frequency range for the reasons mentioned above. could not be specified, and in the high range, the characteristics had to be specified with higher precision than necessary, which was extremely inefficient in terms of circuit implementation.

In particular, when attempting to realize something with complex frequency characteristics such as a graphic equalizer using a conventional FIR type filter, such restrictions can be avoided; in fact, a graphic equalizer using an FIR type filter has not yet been put to practical use. .

On the other hand, since the latter filter uses an addition or multiplication method of analog active filters, there is interference between the filters, and it is difficult to specify the phase and amplitude independently. In particular, it has been extremely difficult to specify the phase characteristics arbitrarily.

SUMMARY OF THE INVENTION In order to solve the problems of the prior art, it is an object of the present invention to provide an octave multiple filter that can specify characteristics at logarithmically uniform intervals over a wide band in one frequency domain.

[Means for Solving the Problems] In order to achieve the above object, the first invention of the present application provides a bandpass characteristic and a lowpass characteristic that operate using the sampling time interval of the input signal as a clock. Each filter stage is constructed from a pair of digital filters having the above sampling characteristics, and the inputs of each filter stage are connected in parallel. The digital filter with low-pass characteristics is configured with a delay element that operates at one time interval out of a group of time intervals that are integral multiples of the reference time interval, with the time interval as a reference time interval, and the digital filter with low-pass characteristics has a delay time per stage. , has a delay element that operates with a delay time that is an integer fraction of the delay time per stage of the delay element of the digital filter having the band-pass characteristic, and each of the filter stages has a delay element that operates with a longer delay time. If a digital filter stage with a bandpass characteristic and/or a lowpass characteristic is defined as an upper stage, and the output of the digital filter with a bandpass characteristic of each filter stage is added to the output of the digital filter with a lowpass characteristic one stage above that stage. The gist is that the signal is connected so as to be input to a digital filter with a low-pass characteristic in its own stage.

Further, the second invention of the present application provides a plurality of digital filter groups having band-pass characteristics and low-pass characteristics that operate with the sampling time interval of an input signal as a clock, and a digital filter group having low-pass characteristics. are connected in series, and digital filter groups with band-pass characteristics are connected in parallel. An input signal is given to each of the digital filter groups with band-pass characteristics, and each output is applied to each of the digital filter groups with low-pass characteristics. If the digital filters are numbered in the order of connection in each digital filter group, the delay time per stage of the unit delay element of the digital filter with the highest number is less than or equal to the sampling time interval of the input signal, and each digital In the filter group, each time the number decreases by one, the delay time per stage of the digital filter increases by AB (A>1.B>0), and the delay time per stage of the digital filter with a certain number has a low-pass characteristic. The gist is that the delay time per unit delay time is an integral multiple of the delay time per stage of a unit delay element of a digital filter having the same number and bandpass characteristic.

Furthermore, the third invention of the present application is an output parallel type, whereas the second invention is an input parallel type.

The gist of the present invention is that the outputs of the digital filters with low-pass characteristics are applied to the digital filters with band-pass characteristics, and the outputs are added and extracted.

[Operation] In the first invention of the present application described above, as the digital filter with low-pass and band-pass characteristics, for example, an FIR type filter is used, and the digital filter with low-pass characteristics is an odd-numbered digital filter with band-pass characteristics. Operates to eliminate the passband, and sets both to 1.
Each filter stage is constructed as a pair, and the delay elements constituting the digital filter of each stage are used for multiplexing by changing the unit delay time.

The second and third inventions of the present application are basically equivalent to the first invention, and therefore operate in exactly the same way.

[Embodiments of the Invention] The present invention will be described below with reference to embodiments shown in the drawings.
FIG. 1 shows the implementation of an octave multiplex filter (5 octaves) according to the first invention of the present application. In the same figure, IN is an input terminal, OUT is an 8-output terminal, and A/D is an A/D terminal.
-D converter, D/A is a DA converter, LPF is a normal analog low-pass filter, S, ~S4 are FtR, for example
F0 to F4 are digital filters having low-pass characteristics using similar filters, D0 to D
4 is an adder circuit.

As shown in FIG.
The coefficient multiplier M
The coefficients a-, ~a, of L are defined.

Figure 3 (a) shows the desired frequency characteristics (low-pass characteristics)
, and FIG. 3(b) shows a time-domain function obtained by inverse Fourier transform.

Each pair of digital filters (so, F
, ) to (S4.F4), each filter stage is configured in parallel between input and output.

In each filter stage, each of the digital filters 80 to S4 having the bandpass characteristic has a delay time nT per stage with the sampling time interval of the input signal in the A/D converter A/D being the reference time interval T. Time interval groups T, 2T, 3 that are integral multiples of the reference time interval
A delay element S that operates at one time interval of T...
, -1 to S, -5. Further, the digital filters F0 to F4 with low-pass characteristics are the digital filters 8 with band-pass characteristics corresponding to the delay time per stage.
It has delay elements Fll-1 to F4-5 that operate with a delay time that is one integer fraction of the delay time per stage of delay elements 0 to S4.

Further, each filter stage is coupled as shown through the adder circuits D0 to D. That is, if each filter stage is defined as an upper stage (in the order of 84 to S, , F4 to F), which has a band-pass characteristic or low-pass characteristic digital filter having a delay element that operates with a longer delay time, each The output of a digital filter with a band-pass characteristic in a filter stage is added to the output of a digital filter with a low-pass characteristic in one stage above that stage, and the sum is input to the digital filter with a low-pass characteristic in its own stage.

Now, in the octave multiple filter configured as described above, the input signal applied to the input terminal IN is sent to the A-D converter A.
/D performs A-D conversion at a sampling frequency f, (sampling rate l/f osec), and is applied to a digital filter 5Il-84 with bandpass characteristics. This filter, for example S4, has a delay time of 1/f, 2'''sec per stage of delay element constituting it.Usually, a digital filter with this configuration has a clock frequency of f.
, 2'''Hz, and the sampling frequency of the input signal is also used as f,2-'Hz. Therefore, the clock frequency and the above sampling frequency are different.

In this state, the desired processing cannot be performed. Therefore, in the present invention, first, as mentioned above, the unit delay time is 1/f, 2-'
Determine the coefficient of the FIR type filter as sac, then for the input signal with actual sampling interval 1/f and see,
Set the clock frequency to fI and HL, and leave the coefficients unchanged as the delay time per stage of the delay element 54-1 nT = 1/
f, 2-'sec. The digital filters of other stages are set in the same manner.

The output characteristics of the digital filter S4 are as shown in FIG.
The signal is input to a digital filter F4 with a 3-sec low-pass characteristic. The output characteristics of F4 are as shown in FIG. 4(b). The output of F4 is added to the output of a digital filter S having a bandpass characteristic with a delay time of l/f, 2-'see per delay element stage, and an adder circuit. FIG. 4(c) shows this added characteristic.

Furthermore, to eliminate the odd-numbered passband of S3.

The delay time per stage of delay element is input to a digital filter F having a low-pass characteristic of 1/f and sM''sac.The output characteristic of F is as shown in FIG. 4(d).

The same process is repeated and the final digital filter F
0, an output with the desired characteristics as shown in FIG. 5 can be obtained.

6 and 7 are respective embodiments of octave multiplex filters according to the second and third inventions of the present application,
The embodiment shown in FIG. 6 is an input parallel type, that is, a digital filter BPI with bandpass characteristics in which input signals are connected in parallel.
BP5, and each output is applied to the amplitude phase coefficient unit C
It is designed to be applied to each of the digital filters LPI to LP4 with low-pass characteristics connected in series via 0El to C0E5, but in the embodiment shown in FIG. 7, the output is parallel type,
That is, the respective outputs of the series-connected digital filters LP4 to LPI with low-pass characteristics are given to the parallel-connected digital filters BP5 to BP1 with band-pass characteristics, and the respective outputs are applied to the amplitude phase coefficient multiplier C0.
The signals are added and taken out by an adder D5 via E5 to C0E1.

In the embodiments shown in FIGS. 6 and 7, the delay time per stage of the digital filters LP4 to LP5 having the highest number in the connection order is less than or equal to the sampling time interval of the input signal, and each digital filter group LPI to LP
4. In BPI to BP5, each time the number decreases by one, the delay time per unit delay element stage of the digital filter becomes AB (A>1.B>0) times, and the digital filter with a certain number has a low-pass characteristic. The delay time per stage of unit delay element is n times (n is an integer) the delay time per stage of unit delay element of a digital filter having the same number and bandpass characteristics.

The embodiments shown in FIGS. 6 and 7 are basically equivalent to the embodiment shown in FIG. 1 and operate in the same way, and the embodiment shown in FIG. The unnecessary band components of the band-pass output signal are sequentially eliminated by digital filter groups LPI to LP4 having low-pass characteristics, whereas the embodiment shown in FIG. A method is adopted in which the low-pass signal output is multiplied by each band-pass characteristic output of the filters BPI to BP5.

FIG. 8 shows an example of FIG. 7 in which AB=2'.

The output characteristics of each digital filter in the case of n=2 are shown.

In the embodiment shown in FIG. 6, random access memory (RAM) can be shared as a delay element for the group of digital filters with bandpass characteristics in the input stage, so memory is saved.
The design becomes easier as well.

Furthermore, in the embodiment shown in FIG. 7, the components of each band can be extracted separately, so it is suitable for applications that add other functions.

[Effects of the Invention] As is clear from the above explanation, according to the present invention, by having the above-mentioned configuration, the frequency domain (on the frequency axis)
The characteristics can be specified at logarithmically uniform intervals, and the resolution can be improved because the band can be widened in the low frequency direction.Moreover, the amplitude and phase can be specified independently, and the interference between each filter can be arbitrarily reduced. can.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing an example of the configuration of an FIR type filter used in the embodiment, FIG. 3 is a characteristic diagram of the filter, and FIG. Output characteristic diagrams of each digital filter in the above embodiments, FIG. 5 is a diagram of the output of the target characteristics, FIGS. 6 and 7 are block diagrams showing other embodiments of the present invention, and FIG. 8 is a diagram of the output characteristics. It is an output characteristic diagram of each digital filter in an example. S0-S4...Digital filters having band-pass characteristics F0-F4...Digital filters having low-pass characteristics D0-D...Addition circuits. A/D...A-D converter, D/A...D-A converter. Figure 2 Clock Figure 3 (α) (b) Date on February Δ, 1986

Claims (3)

[Claims] (1) Each filter stage is composed of a pair of digital filters having band-pass characteristics and low-pass characteristics that operate using the sampling time interval of the input signal as a clock, and each filter stage has inputs in parallel. In each filter stage, the digital filter with the above-mentioned bandpass characteristic has a delay time per stage that is selected from a group of time intervals that are integral multiples of the above-mentioned sampling time interval as a reference time interval. The digital filter with low-pass characteristics has a delay time per stage that is an integer fraction of the delay time per stage of the delay element of the digital filter with band-pass characteristics. Each filter stage has a delay element that operates with a longer delay time, and each filter stage is defined as an upper stage of a digital filter having a band-pass characteristic and a low-pass characteristic that has a delay element that operates with a longer delay time. It is characterized in that the output of a digital filter with a band-pass characteristic is added to the output of a digital filter with a low-pass characteristic one stage above that stage, and the result is connected so that the result is input to the digital filter with a low-pass characteristic of its own stage. Octave multiple filter. (2) A plurality of digital filter groups having band-pass characteristics and low-pass characteristics are provided that operate with the time interval as a clock for the sampling time interval of the input signal, and the digital filter groups with low-pass characteristics are connected in series, and The digital filters with band-pass characteristics are connected in parallel so that an input signal is given to each of the digital filters with band-pass characteristics, and each output is given to each input of the digital filter group with low-pass characteristics. When numbering each digital filter group in the order of connection, the delay time per unit delay element stage of the digital filter with the highest number is less than or equal to the sampling time interval of the input signal, and each digital filter group has one number. Each time the delay time per stage of the digital filter decreases, A^B (A>1,
B>0) times, and the delay time per unit delay element stage of a digital filter with a low-pass characteristic of a certain number is an integer multiple of the delay time per stage of a unit delay element of a digital filter with the same number and band-pass characteristic. An octave multiplex filter characterized by: (3) A plurality of digital filter groups having band-pass characteristics and low-pass characteristics are provided that operate using the sampling time interval of the input signal as a clock, and the digital filter groups with low-pass characteristics are connected in series, and The digital filters with band-pass characteristics are connected in parallel, and the output of each digital filter group with low-pass characteristics is given to each digital filter group with band-pass characteristics, and the output of each digital filter group with band-pass characteristics is applied to each digital filter group with band-pass characteristics. are added and taken out, and if each digital filter group is numbered in the order of connection, the delay time per unit delay element stage of the digital filter with the highest number is less than or equal to the sampling time interval of the input signal, and each In a group of digital filters, each time the number decreases by one, the delay time per stage of that digital filter increases A^B (A>1,
B>0) times, and the delay time per unit delay element stage of a digital filter with a low-pass characteristic of a certain number is an integer multiple of the delay time per stage of a unit delay element of a digital filter with the same number and band-pass characteristic. An octave multiplex filter characterized by: