JP2000197194A - Audio signal generating device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate an audio signal attended with an initial reflecting sound with a long delay time economically without imposing a load on a CPU of a computer system side. SOLUTION: A sound source LSI generates a sample of an instructed musical sound in a channel processing CH1 and the sample is transferred to a delay memory 30D in a RAM (system memory) of a computer system in a way of a burst signal. Furthermore, the sound source LSI receives the musical sound sample through burst transfer from the delay memory 30D in a plurality of channel processing sets CH2 or the like and generates samples of an initial reflecting sound and a rear reverberation sound from the musical sound sample. Then the sound source LSI accumulates the samples (mixer processing 203) and outputs the result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an audio signal generating apparatus having a function of generating an audio signal accompanied by an early reflection sound and a rear reverberation sound generated in an acoustic space.

[0002]

2. Description of the Related Art In recent years, many personal computers are equipped with a sound source LSI to perform various types of sound processing. Recently, in addition to processing to generate audio signals (including both instrument sounds and human voices) as sound processing, in order to simulate various acoustic spaces such as halls and churches, There is a need for an effect process for providing a sound field effect. Here, in an actual acoustic space, there are an early reflection sound and a rear reverberation sound as important elements that characterize the sound field. Of these, the initial reflected sound is the sound that has just been emitted from the sound source is reflected by the wall surrounding the acoustic space and reaches the ears of the listener, and is an element that strongly reflects the characteristics such as the shape and spread of the acoustic space. is there. Conventionally, as a means for generating the initial reflected sound, a method using an FIR filter as illustrated in FIG. 5 or a method using a comb filter as illustrated in FIG. 6 can be considered.

[0003]

By the way, in an actual acoustic space such as a concert hall, a large number of initial reflected sounds are distributed to the listener's ear at different timings through various propagation paths having different path lengths. To reach. If such an initial reflected sound is to be generated by an FIR filter as illustrated in FIG. 5, the following inconvenience occurs. That is, if this method is adopted, the delay time from when the sound source emits a sound to when each initial reflected sound reaches the listener's ear is relatively long, and furthermore, each initial reflected sound is transmitted to the listener's ear. Since the time interval between arrivals is not uniform, the audio signal generated from the sound source is stored in the memory for a considerably long time, and the audio signal is generated at each point in time corresponding to the arrival timing of each initial reflected sound to the listener. It is necessary to perform FIR filter processing as illustrated in FIG. 5 by a method of reading out a signal. Therefore, the memory capacity required for the delay processing D, D,... Of the FIR filter is enormous, but providing such a memory in the sound source LSI is not realistic from an economic viewpoint. It is also conceivable that the CPU of a personal computer that executes sound processing by using this sound source LSI instead of the sound source LSI performs FIR filter processing for generating the initial reflected sound, but this imposes an excessive burden on the CPU side. On the other hand, the method using the comb filter as illustrated in FIG. 5 does not require a large amount of memory capacity, but it is difficult to obtain a realistically reflected initial reflection sound obtained in an actual acoustic space. There is a problem.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and is capable of economically producing an initial reflected sound having a long delay time without incurring a load on a CPU in a computer system such as a personal computer. It is an object of the present invention to provide an audio signal generation device capable of generating an accompanying audio signal.

[0005]

SUMMARY OF THE INVENTION The present invention provides a plurality of audio signal processing means for generating or processing an audio signal, a system memory of a computer system, and a part of the plurality of audio signal processing means. The audio signal generated by the above is transferred to the system memory, and the audio signal transferred to the system memory is transferred to audio signal processing means for processing the audio signal at one or more timings thereafter. An audio signal generation apparatus comprising: a transfer unit; and an output unit that combines and outputs output signals of a plurality of audio signal processing units that perform generation of the audio signal and signal processing on the audio signal. Things.

Further, the present invention provides a plurality of audio signal processing means for generating an audio signal using a wave table stored in a system memory of a computer system, and an audio signal among the plurality of audio signal processing means. Audio signal transfer means for transferring the audio signal generated by the processing means to the system memory and transferring the audio signal transferred to the system memory to another audio signal processing means among the plurality of audio signal processing means And an output unit for combining and outputting the output signals of the plurality of audio signal processing units.

[0007]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments will be described to make the present invention easier to understand. Such an embodiment shows one aspect of the present invention, and does not limit the present invention, and can be arbitrarily changed within the scope of the present invention.

A. Configuration of Embodiment FIG. 1 is a block diagram showing a configuration of a computer system A including an audio signal generation device according to an embodiment of the present invention. In FIG. 1, a CPU 20 is a means for controlling the entire computer system A, and is connected to other components via a bus 60.

The RAM 30 is a readable and writable memory corresponding to the system memory of the computer system A. The RAM 30 has a large number of storage areas. One storage area is a work area for the CPU 20, another storage area is a well-known wave table for storing various waveform data WD, and another storage area is a storage area. Each is used as a delay memory in effect processing.
In the present embodiment, by using the delay memory set in the RAM 30, the above-described sample data of the initial reflected sound can be generated.

The ROM 40 is a read-only memory in which a boot program and the like are stored. Hard disk 50
Is provided in the computer system A as a secondary storage device. In this hard disk 50,
Program such as application program, device driver, microprogram MP and control data CD
And various data such as waveform data WD. The device driver, microprogram MP, waveform data WD, and the like stored in the hard disk 50 are transferred to the RAM 30 via the bus 60 when the OS starts.

Here, the waveform data WD in the hard disk 50 is data obtained by sampling musical tones actually produced by musical instruments corresponding to various timbres such as a guitar and a piano. This waveform data W
D is selected from data stored in the hard disk 50 corresponding to frequently used tones.
Stored in the AM 30. In addition to the time when the OS starts, RA
Waveform data WD is exchanged between M30 and hard disk 50 as needed. In the present embodiment, the waveform data WD is stored in the RAM 30 in a continuous linear area where physical addresses are continuous. this is,
If the physical address is discontinuous, it is necessary to calculate the physical address from the logical address.
This is for reducing the load of S.

The control data CD in the hard disk 50 is referred to by a higher-level application, and is transferred to the RAM 30 as needed. The control data CD is data for driving a tone generator LSI 10 to be described later in response to a performance instruction such as a MIDI format given from an application program, and stores tone color information (waveform data WD corresponding to the tone color). Address, pitch information, volume information, and effect information for instructing effect processing. This effect information includes effect information for instructing the addition of both the early reflection sound and the rear reverberation sound generated in an acoustic space such as a hall.

The microprogram MP is a microprogram executed by the sound source LSI 10 in the sound source board 100 described later. This microprogram MP
As described above, the hard disk 50
Is transferred from the RAM 30 to the RAM 30.
Is transferred to the sound source LSI 10.

The bus 60 is connected to the computer system A
Is a common information transmission path between the respective components. The bus 60 may be of any type as long as it can transfer a large amount of data at high speed. In this embodiment, however, a PCI bus (Peripheral Component Int.) Having a burst transfer mode is used.
erconnect Bus) is used as the bus 60.

The sound source board 100 becomes a component of the computer system A shown in FIG. 1 by being mounted on an expansion slot (PCI slot) not shown. The sound source board 100 includes a sound source LSI 10 that generates output audio data SD, and a DAC 16 that converts the output audio data SD into an output audio signal S that is an analog signal and outputs the signal. Note that the sound source LSI 10 may be mounted on a motherboard of the computer system A.

The sound source LSI 10 includes a control unit 11 and a PCI
It comprises a bus interface 12, an input buffer 13, an operation unit 14, and an output buffer 15.

First, the control unit 11 forms the center of control of the sound source LSI 10, and the RAM is used when the above-described OS is started.
The microprogram MP transferred from the memory 30 is stored in a memory (not shown) provided therein to control each component in the sound source LSI 10 and transmit information between the components according to the microprogram MP. Instruction i
Issue

Here, the timing control of the tone generator LSI 10 performed by the control unit 11 will be described. In the sound source LSI 10, under the control of the control unit 11, a process of generating the output audio data SD is performed in fixed time units called subframes. In the present embodiment, the time (approximately 5.3 ms) for outputting 256 samples of the output audio data SD at the output sampling rate (DAC rate = 48 kHz) is defined as one frame, which is further divided into four (for 64 samples). Is about 1.3 ms) as one subframe. The 64 samples collectively generated in a certain sub-frame are temporarily stored in the output buffer 15, and are supplied to the DAC 16 in the next sub-frame at a speed corresponding to the output sampling rate.

As shown in FIG. 2A, the sound source LS
In I10, channel processing CHN (N = 1 to 64) consisting of 64 channels is executed in each subframe under the time division control by the control unit 11. These channel processes CH1 to CH64 are collectively performed asynchronously with the output sampling rate. In each channel processing, 64 samples of audio data to be output in the next subframe can be created.

In the sound source LSI 10, in each subframe, as shown in FIG. 2B, 64-sample output audio data SD obtained by accumulating the audio data created in each channel processing of the previous subframe is output. , Output sampling frequency (DAC rate = 48
kHz) to the DAC 16 in sequence.

The processing that can be executed in each of the channel processes CH1 to CH64 depends on the contents of the microprogram MP created in advance for the channel.
The processing actually executed in each channel processing is specified by the control data CD stored in the RAM 30 for each channel by the CPU 20 according to the application program.

In this embodiment, each of the channel processes CH1 to CH1
The microprogram MP is created by the CH 64 so that the following processing can be performed, for example.

A. Audio signal generation processing This is processing for generating sample data of musical tones having a specified tone and pitch according to the control data CD.

B. Effect processing Initial reflection sound generation processing When a user specifies a desired one of various sound spaces such as a concert hall, a plurality of initial reflections corresponding to the sound space are obtained from the sample data generated by the audio signal generation processing. This is a process of generating sound sample data. As described above, in order to generate the initial reflected sound, it is necessary to perform a delay process with a large delay amount on the original musical sound waveform. Therefore, in this embodiment,
When sample data generated by an audio signal generation process of a certain channel process is generated, the sample data is output as a musical tone, and is burst-transferred to the RAM 30 to perform another channel process corresponding to the number of initial reflected sounds to be generated. , The initial reflected sound generation processing is executed by receiving the sample data from the RAM 30 by burst transfer at an appropriate timing according to the required delay amount. In a plurality of channel processes applied to the generation of the initial reflected sound, the sample data burst-transferred from the RAM 30 is subjected to coefficient multiplication or simple filter processing, and is output as sample data corresponding to the initial reflected sound.

Rear reverberation sound generation processing This processing is executed together with the above-mentioned initial reflection sound generation processing. In this processing, the sample data of the rear reverberation corresponding to the acoustic space specified by the user is generated from the sample data generated by the audio signal generation processing. The process of generating the sample data of the rear reverberation sound is performed by R
This is performed by receiving the sample data burst-transferred to the AM 30 by burst transfer at an appropriate timing.

C. Mixer processing This processing is performed in the audio signal generation processing and effect processing described above. In the mixer process, an accumulation process is performed in which the sample data generated in the audio signal generation process and the effect process is added to the sample data already stored in the output buffer in the subframe, and returned to the output buffer.
This mixer processing can be permitted or prohibited by the control data CD. The above-described processes will be described in detail when describing the operation of the present embodiment.

Next, the PCI bus interface 12
It has a bus master function. Therefore, the sound source LSI1
0 indicates that the control data CD, the waveform data WD, or the data necessary for the effect processing can be read directly from the RAM 30 without imposing too much load on the CPU 20.

Next, the input buffer 13 is means for receiving and storing control data CD, waveform data WD, or data necessary for effect processing, which are burst-transferred from the RAM 30 via the PCI bus interface 12.

Next, the arithmetic section 14 is means for executing arithmetic operations for audio signal generation processing and effect processing according to the instruction i supplied from the control section 11. Further, the arithmetic processing unit 14 includes effect processing buffers 141 and 142 therein. These effect processing buffers 141 and 142 are provided to implement a relatively long delay by using the RAM 30 of the computer system A. That is, these buffers are provided as buffers for transferring data to be delayed to the RAM 30 by burst transfer. In the present embodiment, two buffers are prepared for effect processing in order to execute two types of effect processing including initial reflection sound generation and rear reverberation sound generation. However, three or more types of effect processing are performed. If so, it is sufficient to prepare a number of buffers corresponding to the type.

Next, the output buffer 15 is composed of a first buffer 151 and a second buffer 152.
Each role is switched in the subframe cycle. That is, in a certain subframe, the first buffer 1
Reference numeral 51 denotes a sample created in each audio signal generation process and a sample created in the effect process (64 each).
Are written (accumulated), and 64 samples (64 samples accumulated in the previous subframe) stored in the second buffer 152 are output to the DAC 16 as output audio data SD according to the output sampling rate. Is read. In the next subframe, 64 samples stored in the first buffer 151 are read out to the DAC 16 as output audio data SD, and an accumulation process is performed in the second buffer 152.

B. Next, the operation of the present embodiment will be described.

(1) General Operation In the computer system shown in FIG. 1, the CPU 20 may execute an application program including sound processing. In this case, the execution of the application program causes the specific tone generation process to be executed by the CPU.
20 may execute. In this case, the CPU 20
Select a specific channel to use for shaping the music,
The control data CD for instructing generation of the musical tone is written in an area of the RAM 30 corresponding to the channel. As described above, the control data CD includes timbre information (an address at which the waveform data WD corresponding to the timbre is stored), pitch information, volume information, and effect information for instructing effect processing. ing.

On the other hand, the control unit 11 of the sound source LSI 10 executes channel processing CH1 to CH64 for each subframe according to the microprogram MP. Each channel processing CH executed according to this microprogram MP
1 to CH64 each include an audio signal generation process.

In the audio signal generation processing in each channel processing, control data C corresponding to the channel is generated.
When D is stored in the RAM 30 as described above, the control unit 11 loads the control data CD from the RAM 30 into the input buffer 13.

Next, the control section 11 takes into account the pitch information, sampling rate, and output sampling rate (DAC rate) contained in the control data CD, and generates a waveform necessary to create 64 samples for one subframe. The data WD is taken into the input buffer 13 from the RAM 30.
This capture is performed by burst transfer by designating the start address and the number of samples. That is, RA
A predetermined continuous area in the waveform data WD of M30 is cut out and transferred collectively to the sound source LSI 10.
After that, based on the pitch information and the volume information in the control data CD, the waveform data WD is subjected to a conversion process to create 64 samples and stored in the output buffer 15. When effect processing is performed, the sample created by the audio signal generation processing is stored in, for example, the effect buffer 1.
At the same time, the output buffer 15 accumulates (mixer-processes) the sample created by the audio signal generation process and the sample created by the effect process.

The 64 samples thus stored in the output buffer are read out as output audio data SD in the next subframe, and output from the DAC 16 as an output audio signal SD.

(2) Operations Specific to this Embodiment In this embodiment, the initial reflected sound generation processing 201 shown in FIG.
And an effect process including a rear reverberation generation process 202.

Here, the initial reflected sound generated in the initial reflected sound generation processing 201 will be described with reference to FIG. In FIG. 4, W1 to W4 are walls of an acoustic space to be simulated which is designated by a user, M is a position of a listener, S is a position of a sound source, and L0 is a direct sound emitted from the sound source and heard by the listener. Is shown.

FIG. 4 shows that the sound output from the position S
2 shows a state where the light is reflected by the light 2 and becomes a primary reflected sound L1 and reaches the listener. 1 that is heard by this listener
The next reflected sound L1 moves the sound source to a virtual sound source position S ′ obtained by mirror-inverting the position S with respect to the wall W2, and is the same as that heard by the listener when the same sound is emitted from this virtual sound source position. Become. Therefore, for the sound emitted by the sound source,
By providing a propagation delay and an attenuation corresponding to the transmission path from the position S ′ to the position M, a sound corresponding to the primary reflected sound L1 from the wall W2 can be obtained.

In FIG. 4, the sound output from the position S is reflected by the wall W2, and the reflected sound is further reflected on the wall W4.
, And a secondary reflected sound L2 reaches the listener. The secondary reflected sound L2 heard by the listener obtains a virtual sound source position S ″ obtained by mirror-inverting the position S with respect to the wall W4 as a reference, and this virtual sound source position S ″
Is further obtained by obtaining a virtual sound source position S "" mirror-inverted with respect to the wall W2 ', moving the sound source to this virtual sound source position S "", and listening to the listener when the same sound is emitted therefrom. Will be the same as The wall W2 'is a mirror image of the wall W2 with respect to the wall W4. In this case, a sound corresponding to the secondary reflected sound L2 can be obtained by adding a propagation delay and an attenuation corresponding to the transmission path from the position S ′ ″ to the position M to the sound emitted from the sound source. .

In FIG. 4, 25 virtual sound source positions obtained by the above-described method are indicated by "x" marks. Each propagation path from each of these virtual sound source positions to the listener's position M can be modeled by a combination of delay processing and multiplication of an attenuation coefficient or simple low-pass filtering.

In the present embodiment, sample data corresponding to the primary reflected sound and the secondary reflected sound is generated by the processing combining the delay processing and the coefficient multiplying processing.
This processing is the initial reflected sound generation processing 201 shown in FIG.

Here, with reference to FIG. 3 again, the effect processing performed in the present embodiment will be described. According to the instruction from the CPU 20, the sound source LSI 10 in the present embodiment, for example, as shown in FIG.
A sample of a specific musical sound is generated by the audio signal generation process 1 and a plurality of samples of the initial reflected sound described above are generated by the effect processing of the other channel processes CH2 to CHk (initial reflected sound generation process 201). The reverberation sample is generated by the effect processing of the channel processing of, and the accumulation of these samples (mixer processing 20
3) can be output from the output buffer 15.

In this case, in the channel processing CH1, a tone sample (64 samples) is generated in accordance with the control data CD by the above-described audio signal generation processing, and this sample is stored in the output buffer 15, and the sample is used for the effect. RAM 3 via buffer 141
The burst transfer is performed to the delay memory 30D in the "0".

Next, for example, in the channel processing CH2, the sample stored in the delay memory D is stored in the input buffer 1 at a time when a predetermined time has elapsed from the time when the generation of the musical tone sample was started by the channel processing CH1.
3 to generate a sample of the early reflected sound. Here, the delay time from the start of musical tone sample generation by the channel processing CH1 to the reception of the sample from the delay memory 30D in the subsequent channel processing CH2 is the length of the propagation path of the initial reflected sound to be simulated. It is time corresponding to. Also,
The filtering performed on the sample received from the delay memory 30D simulates the attenuation characteristic of the propagation path of the initial reflected sound.

The same applies to other channel processing for generating the sample of the initial reflected sound, and a predetermined time corresponding to the propagation path length of the initial reflected sound to be simulated has elapsed after the start of the generation of the sample of the musical sound. At this point, a sample of a musical tone is received from the delay memory 30D by burst transfer, and a sample of an early reflected sound is generated from this sample.
The accumulation with the sample in the output buffer 15 is performed.

The rear reverberation generation process 202 is executed by another channel process. The rear reverberation generation processing 202 includes filter processing such as an all-pass filter and a comb filter (not shown). Here, the all-pass filter has a function of giving a phase delay according to the frequency to a constituent spectrum of a waveform to be processed and generating a waveform having low similarity with the original waveform. . The comb filter is composed of a loop including a low-pass filter and a delay. The sample to be processed is taken in the loop and circulated, thereby outputting a waveform sample in which a high band is attenuated with the passage of time. . According to this comb filter, it is possible to simulate a phenomenon in which a high-frequency component of a reflected sound is attenuated each time reflection is repeated in an acoustic space. In the rear reverberation generation process 202, a sample of the rear reverberation is generated by reading out a tone sample stored in the delay memory 30D and performing a filtering process in combination with an all-pass filter or a comb filter. Fifteen
Performs accumulation with the sample in. Rear reverberation generation processing 20
In 2, the delay processing is performed several times on the sample read from the delay memory 30D or on the sample in the middle of the filter processing. In the delay processing, a sample to be processed in a certain channel process is burst-transferred to the RAM 30, and after a necessary time elapses, the sample is
And received by burst transfer.

The effect processing specific to the present embodiment has been described above. There are the following two methods for controlling a plurality of channel processings constituting this processing.

A. CPU 20 (application program) -driven control system In this control system, the CPU 20
According to the application program, selects a channel for generating a sample of a musical tone, a plurality of channels for generating a sample of an early reflection sound, and a channel for generating a sample of a reverberant sound, and controls each channel processing. The data CD is provided to the sound source LSI 10 via the RAM 30. More specifically, the CPU 20 determines a channel for generating a sample at a timing at which a sample of a musical tone is to be generated, and specifies information for specifying a musical tone to be generated, and a delay memory 30 for the generated musical tone sample.
The control data CD including the information for instructing the burst transfer to D is written into the area in the RAM 30 corresponding to the channel.

The tone generator LSI 10 transmits the control data CD to the RA in the channel processing of the subsequent subframe.
When read from the M30, a tone sample is generated in accordance with the control data CD, the sample is stored in the output buffer 15, and the delay memory 30D in the RAM 30 is subjected to burst transfer using the effect buffer 141.
Send to

The CPU 20 generates the first initial reflected sound after the tone generation (the initial reflected sound reaching the listener via the shortest propagation path).
Is determined, a channel for generating a sample of the initial reflected sound is determined, and the delay memory 30D
And information indicating that a tone sample stored in the RAM 30 should be received by burst transfer, and control data CD including a filter coefficient and the like for a filter process applied to the received sample are stored in the RAM 30 corresponding to the channel. Write to the area.

The sound source LSI 10 transmits the control data CD to the RA in the channel processing of the subsequent subframe.
When read from M30, according to the control data CD, a musical tone sample is received from the delay memory 30D by burst transfer, and is subjected to a filtering process to generate an initial reflected sound sample. Performs accumulation with the sample in.

The generation of other samples of the early reflection sound and the rear reverberation sound are also performed under the initiative of the CPU 20 in the same manner as described above.

In the case of this control method, when a sample of an early reflection sound or a rear reverberation sound is to be generated, the channel processing which is available at that time is applied to the generation of the sample, so that the flexibility of the system can be increased. There is an advantage that you can.

B. In the case of this control method, the CPU 20 generates a channel for generating a sample of a musical tone and a plurality of channels for generating a sample of an initial reflected sound at the timing when a musical tone sample is to be generated. And the channel for generating the sample of the rear reverberation are collectively determined. Then, information for designating a musical tone to be generated and a delay memory 3 for a sample of the generated musical tone.
Control data CD including information for instructing burst transfer to 0D.
The control data CD for generating samples of each initial reflected sound and the control data CD for generating samples of the rear reverberation sound are similarly written in the RAM 30 corresponding to each sample generation channel. Write to the area.

Here, each control data CD for generating each initial reflected sound sample includes information indicating a waiting time until reading of a musical tone sample from the delay memory 30D, and a filter for the read sample. The filter coefficients used when performing the processing are included. Similarly, the control data CD for generating the reverberation sample also includes information (when and where to specify the sample transfer procedure to be performed with the RAM 30 before the final reverberation sample is passed). And what time it is read), and information specifying the operation to be performed on the read sample.

The sound source LSI 10 converts these control data CD into R in the subsequent channel processing of the subframe.
Read from AM30. Then, in the channel processing for generating a musical tone sample in the subframe immediately after that (channel processing for reading out control data CD for generating musical tone samples), musical tone samples are generated in accordance with the control data CD, and this sample is generated. The delay memory 30D in the RAM 30 is stored in the output buffer 15 and is subjected to burst transfer using the effect buffer 141.
Send to

The channel processing for generating the sample of the initial reflected sound (the control data CD for generating the sample of the musical sound)
In the channel processing of reading out the data, the CPU waits until the time specified by the control data CD elapses, and after the elapse of the specified time, receives the sample of the musical tone from the delay memory 30D by burst transfer, Is generated, and the sample of the initial reflected sound and the sample in the output buffer 15 are accumulated.

The same applies to the channel processing for generating the sample of the rear reverberation sound. In accordance with the control data CD supplied via the RAM 30, a sample of the rear reverberation sound is generated, and this is accumulated with the sample in the output buffer 15. I do.

In the case of this control method, necessary control data C
After D is transferred from the CPU 20 to the sound source LSI 10, samples of musical tones, early reflections, and rear reverberation are generated under the initiative of the sound source LSI 10, so that there is an advantage that the load on the CPU 20 can be reduced. The channel for generating the sample of the early reflection sound or the rear reverberation sound may be fixed, instead of being determined each time.

The embodiments of the present invention have been described above. However, the scope of the present invention is not limited to the embodiments, and various modifications may be made to the above embodiments. For example, in the above-described embodiment, one channel processing is used to generate a musical tone sample and one initial reflection sound sample. However, for each of the musical sound sample generation and each initial reflection sound sample generation, two channel processings are used. It is also possible to generate a sample of the musical sound and the initial reflected sound, each consisting of two channels on the left and right sides, by applying the same channel processing. Each of the early reflections generated in the acoustic space arrives at the listener from various Japanese music as illustrated in FIG. 4. In this way, if two channel processes are used to generate the samples of each initial reflection, Since the delay time difference and the sound volume balance of the sounds emitted from the left and right speakers can be adjusted, the sound field can be reproduced with a great sense of reality.

Further, in the above-described embodiment, the audio signal generation device of the form inherent in the computer system has been described as an example. However, the scope of the present invention is not limited to this, and a plurality of audio signal processing means for generating an audio signal using a wave table stored in a system memory of a computer system and a plurality of audio signal processing means Transferring the audio signal generated by one of the audio signal processing means to the system memory, and transferring the audio signal transferred to the system memory to the other audio signal processing means among the plurality of audio signal processing means. IC card or sound card (sound source LSI 10 or sound source board 1 in the above-described embodiment) comprising an audio signal transfer means for outputting the sound signal and an output means for synthesizing and outputting the output signals of the plurality of audio signal processing means.
(Equivalent to 00) may be manufactured and distributed to general users, and the user may mount this IC or card in a computer system and use it as an audio signal generation device.

[0063]

As described above, the audio signal generating apparatus according to the present invention comprises a plurality of audio signal processing means for generating an audio signal or performing signal processing on an audio signal; a system memory of a computer system; The audio signal generated by a part of the audio signal processing means is transferred to the system memory, and the audio signal transferred to the system memory is processed at one or more timings thereafter. A computer configured to include an audio signal transfer unit for transferring the signal to the signal processing unit and an output unit for combining and outputting output signals of a plurality of audio signal processing units for generating the audio signal and performing signal processing on the audio signal; System side Without causing the load on the CPU, and, economically, there is an effect that it is possible to generate an audio signal with a long initial reflected sound delay time.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a computer system including an audio signal generation device according to an embodiment of the present invention.

FIG. 2 is a time chart showing a form of timing control of a sound source LSI in the embodiment.

FIG. 3 is a diagram showing an effect process executed in the embodiment.

FIG. 4 is a diagram illustrating an initial reflected sound generated in the effect processing.

FIG. 5 is a block diagram illustrating an example of a configuration of a means for generating an early reflection sound using an FIR filter;

FIG. 6 is a block diagram illustrating a configuration of a means for generating an initial reflected sound by a comb filter.

[Explanation of symbols]

A: Computer system, 20: CPU, 30:
... RAM (system memory), 10 ... tone generator LSI (plural audio signal processing means), 11 ... control unit, 12
... PCI bus interface, 13 ... input buffer, 141, 142 ... effect processing buffer (above, audio signal transfer means), 15 ... output buffer (output means).

Claims (5)

[Claims] A plurality of audio signal processing means for generating an audio signal or performing signal processing on the audio signal; a system memory of a computer system; and an audio signal generated by a part of the plurality of audio signal processing means. Audio signal transfer means for transferring the audio signal transferred to the system memory to audio signal processing means for performing signal processing on the audio signal at one or more timings thereafter, An audio signal generating apparatus comprising: output means for synthesizing and outputting output signals of a plurality of audio signal processing means for performing generation and signal processing on an audio signal. 2. According to an application program,
A system control processor that outputs control data necessary for generating an audio signal and control data necessary for effect processing on the audio signal; a system memory; and a sound source processor that executes sound processing under the control of the system control processor. A plurality of audio signal processing means for generating an audio signal or performing an effect process on the audio signal in accordance with the control data output from the system control processor; and the plurality of audio signal processing means in accordance with the control data output from the system control processor. The audio signal generated by a part of the audio signal processing means is transferred to the system memory, and the audio signal transferred to the system memory is transferred to the audio signal at one or more subsequent timings. Audio signal transfer means for transferring the audio signal to audio signal processing means for effecting the audio signal, and output means for synthesizing and outputting output signals of a plurality of audio signal processing means for generating the audio signal and effect processing on the audio signal. An audio signal generation device, comprising: 3. The audio signal generation device according to claim 2, wherein the system control processor supplies control data necessary for the effect processing to the sound source processor at a timing at which the effect processing is to be performed. 4. The system control processor supplies the control data including information for specifying a timing at which the effect processing is to be performed to the sound source processor, and the sound source processor is specified by information included in the control data. 3. The audio signal generating apparatus according to claim 2, wherein the audio signal transferred to the system memory is received by the audio signal transfer unit at a given timing, and effect processing is performed by the audio signal processing unit. 5. A plurality of audio signal processing means for generating an audio signal using a wave table stored in a system memory of a computer system; and a plurality of audio signal processing means for generating the audio signal. Audio signal transfer means for transferring the audio signal transferred to the system memory, and transferring the audio signal transferred to the system memory to another audio signal processing means among the plurality of audio signal processing means; Output means for synthesizing the output signal of the audio signal processing means and outputting the combined signal.