JPH08279246A - Audio signal processor - Google Patents

Abstract

PURPOSE: To enable the simple processing of audio signals of plural channels without using plural interruption signals even with an audio signal processor having a single CPU, etc. CONSTITUTION: Relative offsets are generated in a relative offset generator 10 by the reproducing offset data of the reproducing encoding strings extracted by an offset extractor 9 and the reference offset data generated by a reference block pulse generator 11. The audio signals in a memory circuit 6 are controlled in reading out and are read out by these relative offset and are encoded by an encoding device 7. The audio signals are added with the offset in an offset adder 13 via a memory circuit 8 and are thereafter outputted.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio signal processing device for processing audio data of a plurality of channels in synchronization with video frame periods.

[0002]

2. Description of the Related Art Conventionally, as a method for highly efficiently encoding a video signal such as moving image data, there is an MPEG (Moving Picture Experts Group) method utilizing the correlation of video signals. Further, in this MPEG system, an audio encoding system has been proposed in which an audio signal is highly efficiently encoded together with a video signal. The coding algorithm used in this audio coding method is composed of three layers, that is, layers I, II, and III, and any layer can be used.

In an audio signal coding apparatus for coding and recording an audio signal in accordance with a frame period of a video signal of NTSC system used for color television broadcasting, for example, an audio signal sampled at a sampling frequency of 48 kHz is NTSC. According to the standard, 8008 samples are recorded for 5 frames of the video signal of the system, so that 1601 or 1602 samples of audio signals are recorded for 1 frame of the video signal. In the layer I of the MPEG system, one block is composed of 384 samples, so
An audio signal for 1001 blocks is recorded for a video signal of 40 frames. That is, at the beginning of the block of the audio signal to be recorded with respect to the video signal of one frame, the least common multiple of 8008 samples and 384 samples is 48048, and 8008 samples is 5
Since it is a frame, it has 240 phases.

When the audio signal is encoded, it is encoded according to the timing of the recording block pulse which becomes active every 384 samples.

The recording block pulse serves as a unit for encoding the audio signal, and the encoding of the audio signal is started by sending the recording block pulse to the audio signal encoding device.

Further, when reproducing audio data encoded in a video frame period, by reproducing the offset at the time of recording with respect to the reproduction frame pulse, that is, the reproduction offset data, the video frame and the audio signal at the time of recording are reproduced. It is possible to reproduce the phase relationship with.

[0007]

By the way, when two encoded audio data are switched and recorded, for example, as shown in the timing chart of FIG. 13, the two encoded audio data are reproduced according to the reproduction frame pulse SFP. The reproduced encoded sequence SP that has been
When performing recording by switching with, the relationship between the video frame number and the reproduction offset data SOF cannot be maintained.
That is, the recording block pulse RBP becomes discontinuous at the edit point P.

Therefore, in the audio signal processing apparatus capable of processing the audio signals of a plurality of channels for each channel, when the audio signals are coded for each channel, for example, as shown in FIG. There is a possibility that the recording block pulses RBP of the coded audio data for four channels from ch1 to ch4 have different phases.

As a result, a single central processing unit, that is, a CPU and a digital signal processor, without a parallel processing unit, is provided.
In an audio signal processing device using a processor (DSP) or the like, when performing a process of simultaneously switching and recording coded audio data for four channels, FIG.
As shown in FIG. 5, four recording block pulses RBP corresponding to the four encoded audio data to be handled, respectively.
Must be input to the encoder 60. That is, the encoder 60 outputs four interrupt signals corresponding to four channels by a single CPU or DSP.
Must be entered.

As described above, in the audio signal processing apparatus using a single CPU or DSP, when switching and recording encoded audio data of a plurality of channels,
An interrupt signal is required for each audio signal. As the number of interrupt signals increases, the processing in the system becomes more complicated, and the interrupt processing in the system takes time, resulting in inefficiency.

Therefore, in view of the above situation, the present invention makes it possible to perform switching signal recording by simple processing without using interrupt signals for a plurality of channels when switching recording an audio signal of a plurality of channels. A device is provided.

[0012]

An audio signal processing apparatus according to the present invention is supplied with encoded audio data including offset data indicating a relative position with respect to a reference video signal and a new audio signal, and uses a reproduction frame pulse. Offset extraction means for extracting offset data of the encoded audio data as reproduction offset data, reference block pulse generation means for generating reference offset data and reference block pulse according to the reproduction frame pulse, and reference block pulse generation The relative offset generation means for generating a relative offset from the reference offset data from the means and the reproduction offset data from the offset extraction means, and the storage means for storing the audio signal. of To solve the above problems by controlling the readout of the audio signal from the storage means by the pair offset.

[0013]

In the present invention, the offset data of the encoded audio data is extracted as the reproduction offset data by using the reproduction frame pulse, and the reference offset data and the reference block pulse are generated according to the reproduction frame pulse, A relative offset is generated from the reference offset data and the reproduction offset data, and reading of a new recorded audio signal is controlled by this relative offset, so that a plurality of channels of audio can be reproduced by one reference block pulse. Signal processing can be performed.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows a schematic configuration of an audio signal processing device according to the present invention. Further, FIG. 2 shows a timing chart of each signal in the audio signal processing device shown in FIG.

The audio signal processing apparatus of FIG. 1 is supplied with encoded audio data including offset data indicating a relative position with respect to a reference video signal and a new audio signal, and the encoded audio data is reproduced using a reproduction frame pulse. Offset extractor 9 which is the offset extracting means for extracting the offset data of 1) as reproduction offset data, and the reference block pulse generator which is the reference block pulse generating means for generating the reference offset data and the reference block pulse according to the reproduction frame pulse. 11, a relative offset generator 10 which is a relative offset generating means for generating a relative offset from the reference offset data from the reference block pulse generator 11 and the reproduction offset data from the offset extractor 9, Comprises a memory circuit 6 is a storage means for storing the I o signal, the relative offset from the relative offset generator 10 controls the reading of the audio signal from the memory circuit 6.

That is, in this audio signal processing apparatus, as will be described later, a reference block pulse, which is a reference signal serving as an encoding processing reference, is generated for each audio block length of an audio signal to be encoded, The difference between the offset with respect to the reference block pulse and the offset of the reproduction coded sequence that reproduces the encoded audio data is obtained as a relative offset, and as shown in FIG. 3, an audio signal is generated using the relative offset. It is read and encoded.

Thus, when the already-recorded encoded audio data of a plurality of channels is edited in synchronization with the video frame period, the relative position information of the encoded audio data can be recorded without interruption. Generate a reference block pulse that is a reference signal,
Regardless of the number of channels of the recorded encoded audio data, as shown in FIG. 4, only one reference block pulse enables processing of audio signals of a plurality of channels.

Here, the audio signal coding device and the audio signal decoding device which are the basis of the audio signal processing device shown in FIG. 1 will be described.

FIG. 5 shows a schematic structure of an audio signal coding apparatus for coding and recording an audio signal in accordance with an NTSC system video signal used for color television broadcasting, for example. FIG. 6 shows the audio signal coding apparatus. The timing chart of each signal in the signal encoding device is shown.

The audio signal AI is input to the audio signal encoding apparatus from the signal input terminal 22. The audio signal AI is a signal sampled at a sampling frequency of 48 kHz, and the sampling period T is 20.8 μs. This sampled audio signal is specified to record 8008 samples for 5 frames of an NTSC video signal, so 1601 or 1602 samples of audio signal are recorded for 1 frame of video signal. To be done. In the MPEG layer I, 1
Since the block is composed of 384 samples, an audio signal for 1001 blocks is recorded with respect to a video signal of 240 frames, and the beginning of the block of this audio signal has 240 phases.

The input audio signal AI is sent to the encoder 25. A recording block pulse RBP is input from the signal input terminal 23, and the recording block pulse RBP is input to the encoder 25 and the phase comparator 28. This recording block pulse RBP is 384
It is a pulse signal that becomes active for each sample.

In the encoder 25, the audio signal AI is encoded every 384 samples in accordance with the timing of the input recording block pulse RBP. The encoded audio data is written in the memory circuit 26.

Here, the input frame pulse IFP corresponding to the frame number FN of the video signal is input from the signal input terminal 24 to the phase comparator 28, and the recording frame pulse RFP is input from the signal input terminal 21 to the memory circuit 26. Has been done. The input frame pulse IFP and the recording frame pulse RFP have a frame frequency 2 of the video signal.
It is a pulse signal that becomes active every cycle of 9.96 MHz divided by the sampling frequency of the audio signal of 48 kHz, that is, every 1601 or 1602 samples.

In the phase comparator 28, from the input recording block pulse RBP and the input frame pulse IFP,
The head block information of the frame pulse and the offset between the recording block pulse RBP and the input frame pulse IFP are detected. The head block information of this frame pulse is sent to the memory circuit 26. The offset value is shown by the offset vector IOV in FIG. 6, and the offset adder 2 is used as the offset data OF.
7 is output.

The memory circuit 26 reads the accumulated encoded audio data based on the recording frame pulse RFP by using the head block information of the input frame pulse, and sends it to the offset adder 27. .

The offset adder 27 adds the offset data OF to the encoded audio data from the memory circuit 26 as recording offset data. The value of this offset data OF is the frame number FN
It is determined for each of the fields, and as shown in FIG.
It is a value according to the phase of the street. The output from the offset adder 27 is output as a recording coded sequence, that is, encoded audio data AD from the signal output terminal 29 and recorded on a recording medium (not shown).

As the encoded audio data AD, as shown in FIG. 6, the encoded audio data AD ₁
And encoded audio data AD ₂ . The encoded audio data AD ₁ is the input frame pulse IF
Since P and the recording block pulse RBP are coded while maintaining an asynchronous relationship, the data is difficult to record and reproduce audio data in video frame units. The encoded audio data AD ₂ is an encoded sequence rearranged so as to be completely synchronized with a video frame, and is easy to handle for recording / reproducing audio data in video frame units.

A schematic configuration of an audio signal decoding apparatus for reproducing audio data encoded and recorded in this way is shown in FIG. 8, and a timing chart of each signal in this audio signal decoding apparatus is shown in FIG. Figure 9
It is shown in FIG.

In the audio signal decoding apparatus shown in FIG. 8, the audio data coded and recorded by the audio signal coding apparatus described above is read out, and the reproduction coded sequence SP is reproduced at the timing of the reproduction frame pulse SFP.
Is input from the signal input terminal 31. This reproduction coded sequence SP is written in the memory circuit 35 and at the same time,
It is sent to the offset extractor 37.

The reproduction frame pulse SFP is input from the signal input terminal 33 to the offset extractor 37, and
The reproduction block pulse SBP is input from the signal input terminal 32. As a result, the offset extractor 37 separates the offset at the time of recording the reproduction coded sequence SP, and this offset is output to the phase calculator 38 as reproduction offset data SOF. This reproduction offset data SO
F also has 240 phases as shown in FIG.

The output frame pulse OFP is input from the signal input terminal 34 to the phase calculator 38. The phase calculator 38 performs position correction such that the output phase of the audio signal output from the signal output terminal 39 is deviated by the offset by the reproduction offset data SOF, that is, the offset vector OOV in FIG. 9, according to the timing of the output frame pulse OFP. Information is calculated. This position correction information is sent to the memory circuit 35 and the decoder 36.

In the memory circuit 35, the reproduced coded sequence SP is read by using the sent position correction information. The read reproduction coded sequence SP thus read is sent to the decoder 36.

In the decoder 36, the reproduced coded sequence SP from the memory circuit 35 is decoded. The decoded audio signal is output from the signal output terminal 39 as the output audio signal AO.

As described above, in the audio signal decoding apparatus for reproducing audio data in video frame units, by reproducing the offset at the time of recording with respect to the reproduction frame pulse SFP, the video frame and the audio signal at the time of recording are reproduced. It is possible to reproduce the phase relationship.

In the case of editing the coded audio data coded and recorded in the audio signal coding device, the coded audio data already coded and recorded in synchronization with the video frame period are There has been proposed an audio signal processing device for additionally recording a new audio signal without damaging the relationship between the video frame period and the block of encoded audio data.

FIG. 10 shows a schematic structure of this audio signal processing apparatus, and FIG. 11 shows a timing chart of each signal in this audio signal processing apparatus.

The encoded audio data is read from a recording medium (not shown) according to the timing of the reproduction frame pulse SFP input from the signal input terminal 45, and is reproduced as a reproduction encoded string SP from the signal input terminal 43 to the offset extractor 49. Is entered. The reproduction block pulse SBP from the signal input terminal 44 is also input to the offset extractor 49. As a result, this offset extractor 49 separates the offset at the time of recording the reproduction coded sequence SP, and this offset is sent to the recording block pulse generator 50 as reproduction offset data SOF.

The recording block pulse generator 50 uses the transmitted reproduction offset data SOF to record the audio signal AI in the unrecorded portion of the reproduction coded sequence SP shown in FIG.
When additional recording is performed, recording offset data ROF is generated such that the recording block pulse RBP is not interrupted and is continuous with the reproduction offset data SOF. Specifically, -131 is generated as the recording offset data ROF at the frame number FN2 after the edit point P.
The recording offset data SOF is sent to the phase calculator 51. The recording block pulse RBP generated by the recording block pulse generator 50 is sent to the encoder 46.

By performing such an operation, the recording block pulse RBP becomes continuous, and a block loss occurs even when decoding the encoded audio data AD in which the audio signal is newly additionally recorded and encoded. Never. That is, it becomes possible to decode all the blocks of encoded audio data as audio signals.

An audio signal AI for new additional recording is input to the encoder 46 from the signal input terminal 42, and the audio signal AI is encoded according to the timing of the recording block pulse RBP sent. . This encoded audio data is written in the memory circuit 47.

A recording frame pulse RFP is input to the memory circuit 47 from the signal input terminal 41, and the accumulated encoded audio data is encoded based on the recording frame pulse RFP and the encoding processing time t shown in FIG. It is read out after being delayed by only, and sent to the offset adder 48.

The reproduction frame pulse SFP is input to the phase calculator 51 from the signal input terminal 45. In the phase calculator 51, the reproduction frame pulse SF
According to the timing of P, the reproduction offset data SOF
The position correction information is calculated so that the output phase of the audio signal output from the signal output terminal 52 is shifted by the offset due to and is sent to the offset adder 48.

In the offset adder 48, the memory circuit 4
The position correction information from the phase calculator 51 is added to the encoded audio data from 7. The output from the offset adder 48 is the encoded audio data AD.
Is output from the signal output terminal 52 and recorded on a recording medium (not shown).

In the above audio signal processing device, the continuity of the relationship between the frame number FN and the offset is maintained from the recording of the first audio signal to the additional recording of the next audio signal. Therefore, as shown in FIG. 12, even in the case of encoding and additionally recording four channels of audio signals of channels ch1 to ch4, the encoder 46 can generate one recording block pulse RBP regardless of the number of channels. An interrupt (In
The terrupt) signal makes it possible to process all channels, and the encoder 46 provides encoded stream outputs of encoded audio data of channels ch1 to ch4.

For such an audio signal processing device, it is possible to switch and record encoded audio data.
Regarding the operation of the audio signal processing device according to the present invention,
The details will be described below.

The encoded audio data is read from a recording medium (not shown) and input to the offset extractor 9 from the signal input terminal 3 as a reproduction encoded sequence SP. The reproduction block pulse SBP from the signal input terminal 4 is also input to the offset extractor 9. As a result, in the offset extractor 9, the reproduction coded sequence SP
The offset at the time of recording is separated, and this offset is sent to the relative offset generator 10 as reproduction offset data SOF.

Here, the reference block pulse generator 11
Is free running in this audio signal processing device. A reproduction frame pulse SFP from the signal input terminal 5 is input to the reference block pulse generator 11, and a reference block pulse that does not become discontinuous at the timing shown in FIG. 2 in synchronization with the reproduction frame pulse SFP. KBP and reference offset data KOF are generated. The reference offset data KOF corresponds to the offset data OF shown in FIG. 7, and is input to the relative offset generator 10.

The relative offset generator 10 compares the reference offset data KOF from the reference block pulse generator 11 with the reproduction offset data SOF from the offset generator 9, and compares the relative offset by this comparison. Memory circuit 6 as offset HOF
Output to.

An audio signal is input to the memory circuit 6 from the signal input terminal 2 and stored therein. Therefore, in the memory circuit 6, the audio signal is read based on the relative offset HOF from the relative offset generator 10.

Specifically, as shown in FIG. 3, the input audio signal is sequentially written in the memory 70 from the head address ADD ₀ of the write address WA which is the write address of the audio signal. When reading the written audio signal, for example, a read address R that is a read address is used.
In A ₁ , the audio signal of the read address RA ₂ is actually read based on the relative offset.

The audio signal read from the memory circuit 6 is sent to the encoder 7. The encoder 7 receives the reference block pulse KBP generated by the reference block pulse generator 11, and the audio signal is encoded according to the timing of the reference block pulse KBP. The encoded audio data is written in the memory circuit 8.

A recording frame pulse RFP is input to the memory circuit 8 from the signal input terminal 1, and the accumulated encoded audio data is read out based on the recording frame pulse RFP and is supplied to the offset adder 13. Sent.

Here, the reference block pulse KB from the reference block pulse generator 11 is supplied to the phase calculator 12.
The reproduction frame pulse SFP is input from the P and signal input terminals 5. The phase calculator 12 calculates position correction information according to the timing of the reference block pulse KBP and the reproduction frame pulse SFP so that the output phase of the audio signal output from the signal output terminal 14 is shifted by the offset by the reproduction offset data SOF. The position correction information is sent to the offset adder 13.

In the offset adder 13, the memory circuit 8
To the encoded audio data from the phase calculator 1
The position correction information from 2 is added. The output from the offset adder 13 is output from the signal output terminal 14 as encoded audio data AD and recorded on a recording medium (not shown).

As a result, when the encoded audio data for a plurality of channels is switched and recorded, the continuity of the relationship between the frame number FN and the offset can be maintained.

Therefore, even when the audio signals for four channels from the channels ch1 to ch4 are encoded and switched and recorded, the encoder 7 adds one reference block pulse KBP regardless of the number of channels. The generated interrupt signal enables processing of all channels.

[0057]

As is apparent from the above description, the audio signal processing apparatus according to the present invention is supplied with encoded audio data including offset data indicating a relative position with respect to a reference video signal and a new audio signal. ,
Offset extraction means for extracting offset data of the encoded audio data as reproduction offset data using a reproduction frame pulse; reference block pulse generation means for generating reference offset data and reference block pulse according to the reproduction frame pulse; And a relative offset generation means for generating a relative offset from the reference offset data from the reference block pulse generation means and the reproduction offset data from the offset extraction means, and a storage means for storing the audio signal. By controlling the reading of the audio signal from the storage means by the relative offset from the relative offset generating means, the switching recording of the audio signals of a plurality of channels can be performed by one reference block pulse. Also in the audio signal processing apparatus including a CPU and the like, without using a plurality of interrupt signals, it is possible to easily handle the audio signals of a plurality of channels. Therefore, the overhead of the processing time due to the interrupt can be significantly reduced.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of an audio signal processing device according to the present invention.

FIG. 2 is a timing chart of each signal in the audio signal processing device of FIG.

FIG. 3 is a diagram showing a write address and a read address of an audio signal.

FIG. 4 is a diagram showing processing of audio signals of a plurality of channels.

FIG. 5 is a schematic configuration diagram of an audio signal encoding device.

6 is a timing chart of each signal in the audio signal encoding device in FIG.

FIG. 7 is a diagram showing a relationship between a frame and offset data.

FIG. 8 is a schematic configuration diagram of an audio signal decoding device.

9 is a timing chart of each signal in the audio signal decoding device in FIG.

FIG. 10 is a schematic configuration diagram of an audio signal processing device that additionally records an audio signal.

11 is a timing chart of each signal in the audio signal processing device of FIG.

12 is a diagram showing processing of audio signals of a plurality of channels in the audio signal processing device of FIG.

FIG. 13 is a timing chart of each signal when additionally recording encoded audio data.

FIG. 14 is a diagram showing recording block pulses of a plurality of encoded audio data.

FIG. 15 is a diagram showing processing of audio signals of a plurality of channels at the time of additional recording.

[Explanation of symbols]

 6, 8 Memory circuit 7 Encoder 9 Offset extractor 10 Relative offset generator 11 Reference block pulse generator 12 Phase calculator 13 Offset adder

Claims (2)

[Claims] 1. An audio signal processing device for processing a plurality of channels of audio signals in synchronization with a video frame period, wherein encoded audio data including offset data indicating a relative position to a reference video signal and a new audio signal are provided. Offset extraction means that is supplied and uses the reproduction frame pulse to extract the offset data of the encoded audio data as reproduction offset data, and the reference block pulse generation that generates the reference offset data and the reference block pulse according to the reproduction frame pulse. Means, relative offset generating means for generating a relative offset from the reference offset data from the reference block pulse generating means and the reproduction offset data from the offset extracting means, and the audio signal stored therein. That has a storage means, an audio signal processing apparatus characterized by controlling the reading of the audio signal from the storage means by a relative offset from the relative offset generating means. 2. An audio signal read from the storage means is encoded, and a phase calculated from the reference block pulse from the reference block pulse generating means and the reproduction frame pulse with respect to the encoded audio data. The audio signal processing device according to claim 1, wherein