JPH11202867A - Channel accumulator of sound source performing time-division channel operation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform channel accumulation of external input data without providing any adder. SOLUTION: A multiplication part 20-1 of a DSP 20 multiplies the external input data from an external input part 21 by level control data and the multiplication result is stored in an 8-stage FIFO(first-in first-out) 19. At the start of a 1DAC(digital-to-analog converter), a selection part 18 selects the output from the 8-stage FIFO 19 and then an adder 15 adds it to waveform data of a 1st time-division channel. Then a selection part 18 selects the delay output from an 8-stage delay 16 and then the channel accumulation of the external input data and waveform data of 32 time-division channels is performed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound source which performs a time-division channel operation for accumulating tone data of a plurality of time-division channels arranged so as to substantially equally divide a sampling period and external input data. Regarding the channel accumulator.

[0002]

2. Description of the Related Art In a conventional sound source operating in a time-division channel, waveform data of a plurality of time-division channels arranged so as to divide the sampling period equally are multiplied by envelope data and level control data for each time-division channel. In this way, the waveform data of a plurality of time-division channels on which the level control has been performed are accumulated by channel accumulation to generate waveform data for each sampling cycle. In this case, one time division channel is one sounding channel. In some cases, it is desired to input tone data from a microphone input or another sound source to a sound source, and a sound source that performs a time-division channel operation that enables input of such external data has been proposed.

[0003]

However, in order to channel-accumulate the external input data into the tone data inside the tone generator, an adder for adding the external input data to the tone data inside the tone generator is required. There is a problem that an adder must be provided. In addition, by allocating one time-division channel of the time-division channels to external input data, channel accumulation of external input data may be performed without newly providing an adder.
In this method, the number of sounding channels is reduced by one channel, and there is a problem that various performance modes may be affected.

Therefore, according to the present invention, even if external input data is accumulated in a channel, it is not necessary to provide a new adder, and the channel accumulation of a sound source which performs a time-division channel operation without decreasing the sounding channel. The purpose is to provide a vessel.

[0005]

In order to achieve the above-mentioned object, according to the present invention, a channel accumulator for a sound source which performs a time-division channel operation comprises a plurality of clock accumulators arranged so as to divide one sampling period substantially equally. An adding means having a first addition input to which the waveform data of the divided channels are sequentially input and a second addition input to which the output of the selection means is input; Delay means for delaying by a time corresponding to a divided channel; selection means for selecting one of delay output output from the delay means and external input data and inputting the selected data to a second addition input of the addition means; At the timing when the waveform data of the first time division channel, which is the first time division channel in one sampling period, is input to the first addition input of the addition means. When the selection means selects the external input data, the addition means adds the external input data to the waveform data of the first time-division channel. The addition means adds the external input data to the first addition input of the addition means. At the timing when the waveform data of the second time-division channel is input, the selection means selects the delay output output from the delay means until one sampling period ends and inputs the delay output to the second addition input of the addition means. Thus, channel accumulation of the external input data and the waveform data of all time division channels is performed.

In order to achieve the above object, another channel accumulator of a sound source which performs a time division channel operation according to the present invention is arranged so as to divide one sampling period substantially equally, N (n =
A first addition input in which waveform data of a plurality of time-division channels composed of waveform data of 1, 2, 3, 4,... Addition means having two addition inputs, n-stage delay means for delaying the addition output from the addition means by a time corresponding to the one time division channel, and time division output from the n-stage delay means Selecting means for selecting one of the selected n delayed outputs and time-divided external input data consisting of n systems and inputting the selected data to the second addition input of the adding means. The selecting means selects the external input data at the timing when the waveform data of the first time-division channel, which is the first time-division channel in the above, is input to the first addition input of the adding means. By the the addition of each system and the external input data and the first time-division channel of the waveform data in the adding means it is performed, the first of the adding means
At the timing when the waveform data of the second time-division channel is input to the addition input, the selection means selects the delay output output from the n-stage delay means until one sampling period ends, and the second output of the addition means is selected. By inputting to the addition input, channel accumulation for each system of the external input data and the waveform data of all time division channels is performed.

Further, in the channel accumulator of the sound source performing the time division channel operation, the waveform data for one time division channel stored in the delay means or the n-stage delay means at a timing when one sampling period is completed. Further comprising arithmetic processing means having a multiplier for performing mixing and effect processing on the waveform data, wherein the multiplier multiplies the external input data by a level control signal at a timing when the multiplier is empty. The level-controlled external input data output from the multiplier may be input to the selection unit via a first-in first-out unit.

According to the present invention, when the waveform data of the first time division channel, which is the first time division channel in one sampling period, is input to the first addition input of the adding means, the second addition input is performed. , The external input data and the first time division channel are added. Next, when the waveform data of the second time division channel is input to the first addition input of the addition means, the output of the delay means is input as the second addition input, so that the external input data and the first time division channel are input. The waveform data of the second time-division channel is added to the waveform data obtained by adding the above waveform data. Further, when the waveform data of the time division channels after the third time division channel is input to the first addition input of the addition means, the same processing as the processing of the second time division channel is performed. When the processing is completed, waveform data obtained by channel-accumulating the external input data and the waveform data of all time-division channels can be obtained. Accordingly, the present invention provides a channel accumulator for a sound source that performs a time-division channel operation that does not require a new adding unit and does not reduce the number of sounding channels when channel accumulating external input data. be able to.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram showing a configuration example of an embodiment of a sound source having a channel accumulator according to the present invention and performing a time division channel operation. In FIG. 1, reference numeral 1 denotes an arithmetic processing unit (Central Processing Unit: CPU) or an RO.
M (Read Only Memory), RAM (Random Access Memo)
ry), a one-chip microcomputer that controls the entire sound source,
Reference numeral 2 denotes a tone generator register for storing tone control data for controlling the tone generated in each tone generation channel which is a time-division channel, mixer control data for controlling the DSP 20, and DSP control data. 3 is MID
A MIDI input section for receiving a MIDI signal from an I device;
Reference numeral 4 denotes a panel display on which information of various sound sources is displayed. Reference numeral 5 denotes a panel operator for giving various instructions to the sound source.

[0011] Further, the waveform data of each of the 32 time-division sounding channels based on the frequency number parameter and the tone color parameter stored in the tone generator register 2 are equally divided into one DAC cycle as a sampling period. 32ch time division OS generated sequentially according to time division channel timing
C and 12 sequentially generate the envelope data of each of the 32 time-divided sound channels based on the envelope parameter of the sound channel in the tone generator register 2 in accordance with the time-sharing channel timing obtained by equally dividing one DAC cycle. 32c
h time division EG. Further, a multiplier 13 multiplies the waveform data of each time division channel output from the 32ch time division OSC 11 by the level-controlled envelope data of each time division channel output from the 32ch time division EG12, and 14 a 32ch This is a synthesizing unit that synthesizes level control data for level-controlling the envelope data of each time-division channel output from the time-division EG 12. Here, the synthesizing performed by the synthesizing unit is to add when the envelope data and the later-described level control data are of a decibel scale, and to multiply when the data is of a linear scale.

The waveform data in one time-division channel is composed of four systems (dry, reverb, chorus, and variation) in which four different effect processes are performed. That is, the waveform data of one time-division channel is controlled by eight levels of level control data, and the respective time-division eight levels of level control data are respectively allocated to each time slot obtained by equally dividing the one time-division channel into eight. Have been. Also, 3
From the 2ch time-division OSC 11, time-divided waveform data of 32 channels are sequentially generated for each time-division channel. From the 32ch time-division EG12, the time-divided 32-channel envelope data is sequentially extracted from each time-division channel. Generated for each channel. Then, from the level control data generating unit 2-1 in the tone generator register 2 which supplies the level control data to the synthesizing unit 14, eight time-divided level control data for 32 channels are sequentially generated for each time slot. You.

Returning to the description of FIG. 1, reference numeral 15 denotes waveform data in each of the eight time-division channels multiplied by the level-controlled envelope data from the multiplier 13 input to the addition input A and the addition input B. Selector 18 input
The adder 16 accumulates the channels by adding the data of the eight systems selected in the above. The adder 16 stores the accumulated waveform data of the eight time-divided channels output from the adder 15. An eight-stage delay unit consisting of eight unit delay stages, an eight-stage register 17 for storing waveform data of eight time-division channels output from the eight-stage delay unit 16, and a eight-stage delay unit 16
And an output from the eight-stage FIFO 19, and selects one of the outputs from the eight-stage FIFO 19 and inputs the result to the addition input B of the adder 15. It is an 8-stage FIFO (First-In First-Out) having a first-in first-out function for storing external input data.

Further, reference numeral 20 denotes 8 in which all the time-division channels read from the 8-stage register 17 are accumulated.
This DSP outputs stereo musical tone data obtained by performing arbitrary mixing by applying reverb, chorus, and variation effects to the waveform data of the system, for each DAC cycle. Note that the DSP 20 multiplies the external input data input from the external input unit 21 such as a microphone by eight level control data generated from the level control data generation unit 2-2 in the tone generator register 2 with the external input data. Multiplication means 20-1 for outputting eight sets of external input data. Further, 22L and 22R are DACs (Digital Analog Analog) which convert each of the stereo tone data output from the DSP 20 for each DAC cycle into analog tone signals.
Converter), 23L, 23R are DAC22L, DAC
This is a sound system in which analog sound signals of the L channel and the R channel output from the 22R are generated in stereo.

Next, the operation of the sound source shown in FIG. 1 relating to the channel accumulator of the present invention will be described with reference to the timing chart shown in FIG. When a MIDI event is received from the MIDI input unit 3, the one-chip microcomputer 1
Executes the sound source driver process and sends out the sound source parameters to the sound source register 2. At this time, when a note-on event is input, the one-chip microcomputer 1 first generates a new musical tone based on the note-on event.
A tone generation channel is assigned, and tone generator parameters such as note number, tone color information, velocity, envelope parameter, modulation parameter, and eight levels of level control data are set in the channel area assigned by the tone generator register 2.
When a note-off event is input, an envelope parameter or the like that can stop sound generation based on the note-off event is set in the tone generator register 2.

In the 32ch time-division OSC 11, a tone generation process is executed at each time-division channel timing. In this tone generation process, waveform data for 32 channels is generated according to the tone generator parameters of each tone generation channel set in the tone generator register 2 in accordance with the time division channel timing obtained by dividing one DAC cycle (one sampling period) into 32 equal parts. . The arrangement of the 32 time-division channels is the same as the arrangement shown in the column of EG multiplication output in FIG. That is, the start timing of one DAC cycle is set to time t0, and each time-division channel timing obtained by dividing one DAC cycle into 32 equals to time t0.
, T2, t3,..., The time t
Between 0 and time t1, waveform data of the first time division channel is generated and output, and between time t1 and time t2, waveform data of the second time division channel is generated and output. Waveform data of the 3rd to 31st time division channels is generated and output, and the waveform data of the 32nd time division channel is output from time t31 which is the last time division channel timing to time t0 of the next DAC cycle. Generated and output.

The structure of each time-division channel is, as exemplified in the sixteenth time-division channel starting from time t15, eight time-slots TS1, TS1
, TS3,..., TS8, and each level slot is assigned with 8 levels × 32 channels of level control data. That is, the level control data of the first system in the channel is allocated to the time slot TS1 of each time division channel, and the time slot TS2
, The second system level control data in the channel is assigned, and similarly, the time slots TS3 to TS8 are assigned the third system level control data in the channel.
Level control data of the system or the eighth system is assigned. Accordingly, the number of level control data for the eight time-divided 32 time-division channels is 8 × 32 = 25.
6, which means that 32 time-division channels consisting of eight systems time-divisionally divided into one DAC cycle are arranged in order.

In the 32ch time division EG12, 1
Envelope data for 32 channels corresponding to the envelope parameters of each sounding channel set in the tone generator register 2 is generated according to the time division channel timing obtained by dividing the DAC cycle into 32 equal parts. This 32
The envelope data for the channels is supplied to the level control data generator 2-1 in the tone generator register 2 in the synthesizer 14.
Are combined in order with the level control data for 8 channels × 32 channels generated from. As a result, the level control is performed on the envelope data of 32 channels, respectively, and each level control data can be set to an arbitrary value. The arrangement mode of the time-division channels of the envelope data for 8 channels × 32 channels whose level is output from the synthesizing unit 14 is the same as the arrangement mode shown in the column of the EG multiplication output in FIG.

Then, in the multiplier 13, each of the waveform data of the 32 channels generated from the 32 channel time-division OSC 11 is added to the 8 channels × 3 which are generated from the 32 channel time-division EG 12 and are respectively level-controlled by the synthesizing unit 14.
Envelope data for two channels is associated with each other and multiplied, and the time-divided 32 channels to which the envelopes are assigned are sequentially output from the multiplier 13. The arrangement of the 32 time-division channels is the arrangement shown in the column of EG multiplication output shown in FIG. 2, and the waveform data of each of the eight systems in the 32 time-division channels is sequentially added to the addition input A of the adder 15. Is input to

On the other hand, the data selected by the selection unit 18 is input to the addition input B of the adder 15, but from time t0 to time t1 shown in FIG. External input data VO is selected,
From time point t1 to the time point at which the 1 DAC cycle ends (the start time point of the next 1 DAC cycle) t0, delay outputs DL1 and DL composed of waveform data for one time division channel output from eight-stage delay section 16
2,..., DL31 are selected. That is, as shown in the 8-stage FIFO output column and the 8-stage delay output column shown as the B input in FIG. 2, the output is selected at a timing without hatching, and each output is the same as the time-division channel. Has eight time slots respectively assigned to eight sets of waveform data.

As a result, the adder 15 outputs the time t0
From time t1 to time t1, eight systems of waveform data of the first time-division channel and eight systems of external input data read from the eight-stage FIFO 19 are added at each time slot timing, and the eight systems have additional outputs. Is an 8-stage delay unit 16
Are sequentially stored in each stage. The external input data is data from a microphone or an external sound source input from the external input unit 21, and is output from the multiplication unit 2 in the DSP 20.
At 0-1, the timing is used by the microprogram in the immediately preceding DAC cycle to use the multiplying unit 20-1 so that the multiplication unit 20-1 is not used for each of the eight systems output from the level control data generating unit 2-2 in the tone generator register 2. The data is multiplied by the level control data and stored in the 8-stage FIFO 19. Here, the level control data generator 2-2 and the above-described level control data generator 2-1 are registers for storing level control data, respectively, and supply each level control data at a necessary timing. I do.

Next, at time t1, the selector 18
The delay output from the step delay unit 16 is selected. Then, from time t1 to time t2, the waveform data of eight systems of the second time division channel output from the multiplier 13;
Eight lines of waveform data for one time division channel sequentially output from the eight-stage delay unit 16 are added at each time slot timing, and the added output of the eight systems is stored in each stage of the eight-stage delay unit 16. Become so. As a result,
The eight-stage delay section 16 is added with eight levels of external input data, eight systems of waveform data of the first time division channel, and eight systems of waveform data of the second time division channel for each system. As a result, waveform data obtained by channel-accumulating the three data is stored. Further, the selection unit 18 selects the delay output from the eight-stage delay unit 16 at time t2. Then, from time t2 to time t3, eight sets of waveform data of the third time division channel output from the multiplier 13 and eight sets of waveform data of one time division channel output from the eight-stage delay unit 16 are output. Are added at each time slot timing, and the eight system addition outputs are stored in each stage of the eight-stage delay unit 16. As a result, the eight-stage delay section 16 has eight systems of level-controlled external input data, eight systems of waveform data of the first time division channel, eight systems of waveform data of the second time division channel, and a third system. Waveform data obtained by channel-accumulating the above four data, which is the result of adding the waveform data of eight systems of the time division channel for each system, is stored.

Such an addition operation is performed between the time points t3 to t31. At the time point t1, the eight-stage delay unit 16 supplies the eight-level externally-inputted data and the first time-division data. Waveform data obtained by channel-accumulating the 32 data, which is the result of adding eight channels of waveform data to eight channels of the 31st time-division channel for each channel, is stored. Then, the selection unit 18 selects the delay output from the eight-stage delay unit 16 at time t31. Then, from time t31 to time t0 (start timing of the next DAC cycle), the waveform data of eight systems of the 32nd time division channel output from the multiplier 13 and the one-time division data output from the eight-stage delay unit 16 The eight channels of waveform data for the channel are added at each time slot timing, and the added output of the eight channels is stored in each stage of the eight-stage delay unit 16. As a result, the eight-stage delay unit 1
Reference numeral 6 denotes eight levels of external input data whose level is controlled, eight types of waveform data of the first time division channel, or thirty-second data.
Waveform data obtained by channel-accumulating the data of the above-mentioned 33 systems, in which the waveform data of the eight systems of the time-division channel are added for each system, are stored. That is, 3
By repeatedly performing the addition operation twice, it is possible to accumulate the data of 33 systems of eight channels without providing a new adder.

The tone data ADD obtained by channel-accumulating the 33 data of the eight systems is supplied to the eight-stage delay unit 1 between the start timing t0 and the time t1 of the next DAC cycle.
The data is sequentially transferred from the 6th to the 8th stage register 17. This transfer mode is shown in the column of "8-stage register loading" in FIG. The transferred musical tone data ADD is made into eight sets of musical tone data, each of which is stored in the eight-stage register 17.
The eight sets of tone data ADD are read out from the eight-stage register 17 to the DSP 20 at a predetermined timing, and reverb, chorus, and variation effects are applied to two sets each. The total six sets of tone data and the remaining dry Volume control is performed for the two systems, and arbitrary mixing is performed with the 8-input 2-output multi-effect unit 20-.
2 is performed. As a result, the tone data of the L channel and the R channel, which are stereo, are output from the DSP 20 every one DAC cycle.

The processing in the DSP 20 is performed by the DSP 2
The timing at which the multiplier 20-1 is not used in this process is shown in the column of the multiplier non-use timing in FIG. Eight times after the time t1 in the multiplier non-use timing are used, and the external input data from the external input unit 21 and the level control data generation unit 2 in the tone generator register 2 are used.
-2 is multiplied by the level control data for each of the eight systems output from -2. Then, the result of this multiplication is stored in an 8-stage FIFO.
By storing the data in 19, the channel-accumulated eight-system external input data whose level is controlled in the next DAC cycle.

More specifically, in each step of the DSP 20, it is determined whether or not two inputs of the multiplier have selected valid values. When the information indicating the non-use of the multiplier is detected during the execution of the micro-program, the information indicating the use / non-use of the multiplier is embedded as the timing not used or in the micro program executed by the DSP 20. As the timing when the multiplier is not used, the external input data from the external input unit 21 and the level control data generation unit 2-2 in the tone generator register 2 are used.
Is multiplied with the level control data classified into eight channels output from the DSP 20 using a multiplier in the DSP 20. Accordingly, it is not necessary to add new hardware such as a multiplier, and the external input data from the external input unit 21 and the level control data in the tone generator register 2 are not affected without affecting the execution of the microprogram in the DSP 20. It can be multiplied by the level control data output from the generator 2-2 and classified into eight systems.

In the above description, the time division channel is 32
Although the number of channels has been described, the number of channels is not limited to 16 channels, 64 channels, or other channels. Each time-division channel has 8 channels.
Although the system is constituted by the system (8 time slots), the invention is not limited to this, and the system can be constituted by an arbitrary number of systems.
When the number of systems is n, the eight-stage delay units 16, 8
Each of the stage register and the 8-stage FIFO has n stages. Further, the multiplication unit 20-1 performs n multiplications and performs an 8-input / 2-output multi-effect unit 20-n.
In No. 2, effect addition and mixing are performed on the data input n.

[0027]

As described above, according to the present invention, when the waveform data of the first time division channel, which is the first time division channel of the sampling cycle (DAC cycle), is input to the addition input A of the addition means, the addition is performed. By selecting and inputting the external input data as the input B, the addition of the external input data and the first time-division channel is performed for each system. Next, when the waveform data of the second time division channel is inputted to the addition input A of the adding means, the output of the eight-stage delay section is inputted as the addition input B, so that the external input data and the first time division channel are inputted. The waveform data of the second time-division channel is added to the tone data obtained by adding the above for each system. In addition, the third
When the waveform data of the channels after the time division channel is input to the first addition input of the addition means, the same processing as the processing of the second time division channel is performed. Waveform data obtained by adding input data and all time-division channels for each system can be obtained. Therefore, the present invention does not require a new adding means when channel-accumulating external input data, and
Thus, it is possible to provide a channel accumulator for a sound source that performs a time-division channel operation without reducing the number of sounding channels.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration example of an embodiment of a sound source having a channel accumulator according to the present invention and performing a time-division channel operation.

FIG. 2 is a diagram showing a timing chart for explaining the operation of the channel accumulator of the present invention.

[Explanation of symbols]

1 1-chip microcomputer, 2 sound source register, 3 MID
I input section, 4 panel display, 5 panel operator, 11
32ch time division OSC, 12 32ch time division EG,
13 multiplier, 14 combining unit, 15 adder, 168
Stage delay unit, 178 stage register, 18 selection unit, 1
9 8-stage FIFO, 20 DSP, 21 external input unit, 2
2L, 22R DAC, 23L, 23R sound system

Claims (3)

[Claims] 1. A first addition input to which waveform data of a plurality of time-division channels arranged so as to substantially divide one sampling period are sequentially inputted, and a second addition input to which an output of a selection means is inputted. And delay means for delaying the addition output from the addition means by a time corresponding to the one-time-division channel; and delay output from the delay means and external input data. Selecting means for selecting any one of them and inputting the selected signal to the second addition input of the adding means, wherein the waveform data of the first time division channel, which is the first time division channel in one sampling cycle, At the timing of input to the addition input, the selection means selects the external input data,
The adding means performs addition of the external input data and the waveform data of the first time division channel, and at the timing when the waveform data of the second time division channel is input to the first addition input of the addition means, The selection means selects the delay output output from the delay means until one sampling period is completed and inputs the delay output to the second addition input of the addition means, so that the external input data and the waveform data of all time division channels are A channel accumulator for a sound source that performs a time-division channel operation, wherein the channel accumulation is performed. 2. One sampling period is arranged so as to be substantially equally divided, and at the same time, n (n = n
A first addition input in which waveform data of a plurality of time-division channels composed of waveform data of 1, 2, 3, 4,... Addition means having two addition inputs; n-stage delay means for delaying the addition output from the addition means by a time corresponding to the one time division channel; and time division output from the n-stage delay means. Selecting means for selecting either one of the selected delayed output composed of n systems and the time-divided external input data composed of n systems and inputting the selected data to the second addition input of the adding means. The selecting means selects the external input data at the timing when the waveform data of the first time-sharing channel, which is the first time-sharing channel in the above, is input to the first addition input of the adding means. By Rukoto,
A timing at which the external input data and the waveform data of the first time-division channel are added for each system in the adding means, and a timing at which the waveform data of the second time-division channel is input to the first addition input of the adding means; Wherein said selecting means selects a delay output output from said n-stage delay means until one sampling period ends and inputs it to a second addition input of said adding means, whereby said external input data and all time division A channel accumulator for a sound source performing a time-division channel operation, wherein channel accumulation for each channel with channel waveform data is performed. 3. A multiplier which takes in waveform data for one time-division channel stored in the delay means or the n-stage delay means at a timing when one sampling period is completed, and performs mixing and effect processing on the waveform data. The multiplier further comprises: at the timing when the multiplier is empty, the multiplier multiplies the external input data by a level control signal, and the level-controlled external input output from the multiplier. 3. A channel accumulator for a sound source performing a time-division channel operation according to claim 1, wherein data is input to said selecting means via a first-in first-out means.