JPS6352380A - Recording device for compressed voice signal - Google Patents

Abstract

PURPOSE:To make it possible to record a compressed voice signal of the same voice signal with the last part of a sector one before by providing two main memories having capacity to store (N-1)/N of voice data of the quantity of one sector and two auxiliary memories having capacity to stone 1/N of the rest. CONSTITUTION:The quantity (length) of voice signals compressed and stored in a main memory formed of the first, second main memories 8, 9 and auxiliary memories 12, 13 in overlap periods T3, T4 is determined based on compression ratio and access time in the part of periods T3, T4, and voice signals of the quantity for are sector are compressed and recorded. Accordingly, capacity of the first and second main memories that store voice data of voice signals of the length compressed and recorded in the period (T4-T3) and capacity of the first and second auxiliary memories 12, 13 that store voice data of voice signals of the length compressed and recorded in the period T3 become constant independently of compression ratio, and the sum of capacity of the first main memory 8 and the first auxiliary memory 12 and the sum of capacity of the second main memory 9 and auxiliary memory 13 become capacity of voice data of the quantity for one sector newly recorded in the period T4.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application] The present invention relates to a compressed audio signal recording device for time-compressing and recording audio signals on a recording disk such as a magnetic disk for an electronic still camera.

[Conventional technology]

Conventionally, the recording medium of electronic still cameras for still image capturing has been electronic still cameras, as described in the July 2, 1984 issue of Nikkei Electronics J, a magazine published by Nikkei McGraw-Hill, pages 80-85. It is possible to use a flexible magnetic disk called a camera magnetic disk, and its specifications are as follows.

Regarding recording formats, etc., a unified standard has been proposed by the so-called "Electronic Still Camera Council."

A magnetic disk for an electronic still camera has a plurality of concentric tracks formed thereon, and a pulse generator yoke (hereinafter referred to as a PG yoke) made of a magnetic piece for detecting a reference position is provided on the center hub, so that still images can be captured. The video signal is recorded according to a so-called video floppy recording format.

By the way, the magnetic disk for electronic still cameras includes:
Not only still image video signals, but also digital/information signals,
It is also being considered to record audio signals for commentary on still images.

The audio signal is, for example,
No. 6, 8AK'J of Utility Application No. 178417/1986
As described in Book I and the drawings, it has been proposed to perform time compression and FM modulation, and to divide and record on each track of a disc in the format described below.

First, the track format of each track will be explained.

As shown in FIG. 4(a), each track (1
) are four equally spaced sectors (So), (81), (S2), (83) based on the data start point (DSP) obtained by detecting the leakage magnetic flux of the PG yoke.
and between point CD5P) and the first sector (SO), between each sector (SO) to (Ss), and between the fourth sector (Ss) and point (DSP), respectively. A space (Ss) is provided for spacing.

Note that the periods of the leading and trailing spaces (Ss) are set to at least 2H (H is the horizontal scanning period), and each sector (Ss) is set to at least 2H (H is the horizontal scanning period).
The period of the space (Ss) between o) and (St) is set to any period other than 0, for example, about 30μ.

Next, the sector format of each sector (So) to (Ss) will be explained.

The sector format of each sector (So) to (Ss) is as shown in FIG. 4(b), which is an enlarged view of the sector (St).
2 are provided, and are arranged from the trailing edge of the period T+ to the leading edge of the period T2, and at this time, a pedestal period of a standard level for separation is provided between each period Tα to Tγ.

In the start and end flag periods Tα and Try:, start and end flags indicating the start and end of a sector are recorded, and at this time, the levels of both flags are set to hi-re/1 based on whether an audio signal is recorded or not. ．． / (Positive) Mata becomes Lorepenosi (Negative).

Further, in the control code period iβ, information regarding the recorded audio signal, such as the number of the track where the corresponding video signal is recorded, the compression ratio, etc., is recorded.

Furthermore, in the audio signal period TT, a compressed audio signal obtained by time-compressing the audio signal is FΔf modulated and recorded, and at this time, in an overlap period T3 at the beginning of the period Tr, the compressed audio signal recorded in the previous sector is The same signal as the last part of the audio signal is recorded, and a new compressed audio signal is recorded during the remaining period T4.

Note that each period T+-T9 in FIG. 4(b) differs depending on the video signal format of the disc, for example, 5
In the case of a format of 25 lines and 60 fields, the settings are as shown in the table below, and the recording time for one sector is approximately 4
It becomes m5ec.

table On the other hand, the compression ratio for audio signals is 320 times and 640 times.

It is considered to compress time by setting the compression ratio to 1280 times, 320 times, 640 times, 1280 times.
If each track is doubled, one track takes approximately 5Sec, 10sec, 2Qs, excluding the overlap period T3.
It becomes possible to compress and record each of the ec audio signals, and at this time, as the compression ratio increases, the frequency characteristics during reproduction deteriorate.

Furthermore, it has been considered that a compressed audio signal formed by time compression is subjected to FM modulation using a carrier wave having a center frequency θλh.

Note that at each compression ratio mentioned above, one track takes about 5 seconds.

When compressing and recording audio signals of 10 SeC and 20 SeC, the amount of audio signals that can be recorded in one sector period T4 is approximately 1.3 seconds, 2.5 seconds, and 5 seconds.
Each signal has a length of c.

The amount of audio signals compressed and recorded in the period T4 of each sector (SO) to (S+) changes depending on the total amount of audio signals to be recorded and the compression ratio. For example, at a compression ratio of 640 times, 7 When recording an audio signal, on any one track, one sector amount of 2.5 SeC is recorded in period T4 of sector (So) and (81), and the remaining 2 SeC is recorded in sector (S2). When recording a 12-second audio signal on a disc with the same compression ratio, one sector is recorded in period T4 for each sector (SO) to (S3) of the first track. 15 seconds are recorded, and the remaining 2 seconds are recorded in the period T4 of the sector (So) of the next track.

By the way, the overlap period T3 of each sector (SO) to (S3) is provided to ensure continuous reproduction of the audio signal during playback, and one overlap period T3 is provided when the compression ratio is 320 times, 640 times, or 1280 times. Approximately 1.4 m5eC, 2.9 m5ec compressed and recorded in the last part of period T4 of the previous sector
, 5.7 m5ec, and the respective audio signals are compressed and recorded.

That is, during playback, the start flag of the flag period Tα of each sector (SO) to (S3) is detected, and the compressed audio signal recorded in each sector (SO) to (S3) is extracted. Each extracted sector (So) ~ (
It is necessary to time-expand and synthesize the compressed audio signal in S3) to reproduce and form a continuous audio signal.

If the overlap period T3 is not provided in each sector (SO) to (S3), if the extraction timing of the compressed audio signal shifts even slightly due to jitter, changes in circuit time constants based on temperature, etc., the timing shown in Figure 84 ( A part of the pedestal period before and after the period TTO in b) is extracted as being considered as part of the period of the compressed audio signal, and at this time, the pedestal period part becomes a so-called silent part, so the reproduced sound This causes the inconvenience that the signal may be interrupted.

Therefore, an overlap period T3 is provided in each sector (SO) to (S3), and the overlap period T3 is
, the compressed audio signal of the last part of period T4 of the previous sector is recorded in an overlapped manner, and during playback, each sector (
SO) to (S3), for example, the compressed audio signal from the center of the overlap period T3 to T3/2 before the end of the period T4 is extracted, and even if the above-mentioned extraction timing shift occurs, the reproduction formation The idea is to prevent the audio signal from being interrupted.

[Problem/point that the invention attempts to solve]

The specific configuration of a device that reproduces and forms an audio signal by reproducing a compressed audio signal recorded in the above format is described in the specifications and drawings of the above-mentioned applications. , the specific structure of an apparatus for compressing and recording in the above format has not been invented.

By the way, when an audio signal is time-compressed and recorded, the time is generally compressed by using a memory such as a delay line or a digital memory, and by making the writing speed and reading speed of the memory different.

On the other hand, as described above, in the period T4 from sectors (So) to (S3), the amount of one sector is determined by the compression ratio and the like.

For example, only about 10 seconds of audio signals can be recorded.

Therefore, the entire view of the audio data to be recorded, which is still a digital signal of the audio signal, is written in the memory in advance and held therein, and after writing, it is read out at a reading speed higher than the writing speed, for example, in increments of I trough. It is conceivable to sequentially record compressed audio signals in the period T4 of sectors (So) to (S3) of each track (1).

However, since the total fr (length) of the audio signal to be recorded is not constant, it is difficult to determine the memory capacity, and even if it were determined based on an assumption, the capacity would be extremely large. It is difficult to form a device using a large-capacity memory not only in terms of circuit configuration but also in terms of cost.

Furthermore, it is not possible to simply change the write speed and read speed of the memory in sectors (So) to (S3) of each track (1).
), it is not possible to record the same compressed audio signal as the last part of the previous sector.

[Means for solving problems]

This invention has been made with the above-mentioned points in mind, and it divides each concentric track formed on a recording disk into a plurality of sectors, and also compresses the audio signal into each sector by time-compressing the audio signal. In a compressed audio signal recording device that sequentially performs divided recording, (N-1)/N (N>1) of l sector amount of audio data formed by dividing the audio signal into a predetermined amount of new recording in one sector. two main memories with a capacity to store , and two submemories with a capacity to store the remaining 9 I/Ns, the first (N-1)/N of audio data of 1 sector amount, and the remaining ] /N to one of the main and secondary memories, and then the first (N-1)/N of the next one sector amount of audio data.
N, a write control means for alternately writing the remaining 1/N of the remaining 1/N into the other main and submemories; and the one submemory during writing to the one main memory;
j sequential reading of the other main and submemories.

the other secondary memory while the other main memory is being written;
a read control means for alternately performing sequential reading of the one main and submemories at a higher speed than writing;

Based on the audio data read from each memory, 1
At the time of recording each sector, a compressed audio signal for recording of l sector consisting of the remaining 17N of audio data of 1 sector amount recorded in the previous sector and 1 sector amount of audio data recorded in the sector is generated. The compressed audio signal recording apparatus is characterized by comprising a recording output means for sequentially outputting the compressed audio signal to the recording means of the disk.

[For production]

Then, the main memory of 2@ is transferred to the 1st. As the second main memory, 2
The first . When the second secondary memory is used, the write control means causes each sector of audio data to be transferred to one memory pair with a capacity of one sector formed by the first main and secondary memories, and the second main and secondary memories. The first (N-+7/N) of the 1-sector volume of audio data is written to the other memory pair of 1-sector volume t to form the other memory pair, and at this time, the first (N-+7/N) of the 1-sector volume of audio data is written to each of the two memory memories. The remaining [/N] of audio data of one sector amount is written in each of the two submemories.

Further, the read control means sequentially reads the first secondary memory and the second main and secondary memories at a higher speed than the writing speed while writing to the first main memory, and while writing to the second main memory, the second secondary memory Ah, the first lord.

Secondary memory is faster than compression:’l! It is then read out.

Therefore, each track of the recording disk is shown in FIG. 4(a).
, (b) four sectors (SO) to (S
3), the audio signal is time-compressed, and each sector (SO
) to (S3), sector (So) to
(S3) Of the audio data of one sector amount which is equal to the audio signal that can be compressed and recorded in each period T4, period (T4-
The first (N-re/
N audio data is the first. 17 which are written in each of the second main memories and compressed and recorded in the longer portion of the remaining period T3.
'N audio data is the first. At this time, while writing the first (N-1)/H audio data of one sector amount of audio data to the first main memory, 1/N of the first sub memory is written to the second sub memory. The audio data of 1 sector is read out as the audio signal recorded in the overlap period T3, and the audio data of 1 sector of the second main and secondary memories is read out as the audio signal of 1 sector newly recorded in the period T4. Similarly, while writing the first (N-1)/N of the next one sector amount of audio data to the second main memory, the 17
N audio data is read out as an audio signal recorded in the overlap period T3, and the first main audio data is read out as an audio signal recorded in the overlap period T3.

One sector of audio data in the sub-memory is read out as l sector's worth of audio signals newly recorded in period T4.

And since the read speed is faster than the write speed,
The audio data read from each memory becomes compressed audio signal data obtained by time-compressing the written audio signal, and is transferred from the recording output means to the recording means of the recording disk such as a magnetic head during an overlap period T3 and a period T4. A compressed audio signal for recording of one sector is output, and the compressed audio signal is recorded on the recording disk in the format of the fourth f3 (a) and (b).

〔Example〕

Next, the present invention will be explained in detail with reference to FIGS. 1 to 3 showing one embodiment thereof.

(Structure) Figure 1 shows four sectors (So) to (S3) of each concentric track (1) formed on a magnetic disk for an electronic still camera in the format shown in Figures 4(a) and (b). )
2 shows the configuration for recording compressed audio signals. In the same figure, (2) is the recording audio input terminal into which the analog audio signal for recording is input, and (3) is connected to the input terminal (2). (4) is a noise reduction circuit connected to the filter (3),
Logarithmically compresses the amplitude of the input audio signal.

(5) is an analog/digital conversion circuit (hereinafter referred to as A/D conversion) connected to the noise reduction circuit (4), and the A/D conversion circuit (6) is connected to the noise reduction circuit (4). Based on the conversion clock signal,
The input analog audio signal is converted into digital audio data at a sampling frequency according to the compression ratio, and 1-byte (8-bit) audio data is sequentially output.

(7) is an analog switch for writing into the main memory into which the audio data of the conversion circuit (6) is input;
Second main memory write control signal Wm 1. Based on Wm 2, the input audio data is output to the 2 output terminal (7a).
, (7b) for switching output.

(8) and (9) are the first .
The second main memory is the period T4 in FIG. 4(b).
(N-1) of 1 sector amount of audio data compressed and recorded in
/N (N>1), that is, the capacity to record audio signal data whose length is shorter than the entire audio signal recorded in period T4 by the audio signal recorded in overlap period T3 in FIG. Based on the control signals Wm l and Wm 2 input to the write control terminal (w), the output terminals (7a) and (7b) of the switch (7) are switched and outputted to the data input terminal (1). While the audio data is written respectively, the first . Second main memory read control signal Rm+, Rm
Based on No. 2, the written audio data is read out from the data output terminal (0) at a read speed higher than the write speed.

00 is an analog switch for reading main memory, through which the audio data read from both main memories (8) and (9) is input to two input terminals (IOn) and (10b), respectively, and control signals Rm+ and Rm2 are input. Originally, both main memories (
8) Synthesize and output the audio data read from (9).

Ql) is an analog switch for writing into the sub-memory into which the audio data of the conversion circuit (6) is input;
Second sub-memory write control signal Ws +, Ws 2
Based on the above, input audio data is input to two input terminals (1
1a), (Itb) to switch output.

of α, α4 is the first . It is a second sub-memory, and has a capacity of 1/H of one sector of audio data compressed and recorded during the above-mentioned period T4.

In other words, the capacity is set to record the remaining length of audio signal data after excluding the audio signals written to both main memos 'J (8) and (9) from all the audio signals recorded in period T4. , based on the control signals Ws l and Ws 2 input to the write control terminal (W), the data input terminal (
1), the audio data switched and output is written respectively, and the audio data input to the read control terminal (r)K, which will be described later, is written. Second sub-memory read control signal Rs l, Rs
2, the written audio data is read out from the data output terminal (0) at a read speed higher than the write speed.

α(1) is an analog switch for reading the sub-memory, through which the audio data read from the Ryokawa memories α2 and Q3 is input to the two input terminals (+=1i1) and (14b), respectively, and the control signal RsI, Based on Rs 2, the audio data read from the Ryokawa memories (2) and (13) are synthesized and output.

α9 is a digital/analog conversion circuit (hereinafter, digital/analog conversion is referred to as D/A conversion) into which the audio data output from switch 00, 0■ is input;
Based on the D/A conversion clock signal of the conversion clock terminal αO, the input audio data is returned to an analog audio signal, and during the period Tγ shown in FIG. 4(b), a compressed audio signal for recording one sector is generated. Generate and output.

The α power is a flag and control generation circuit that outputs the start and end flags of the periods Tα and Tδ and the control code of the period Tβ shown in FIG. 4(b).

0 seconds is an adder circuit connected to the conversion circuit 09 and the generation circuit α power, and the 8E compressed audio signal outputted from the conversion circuit αQ and the start signal outputted from the generation circuit α power.

The end flag and control code are added to form a complete signal in the one sector format shown in FIG. 4(b) and output.

α9 is a pre-emphasis circuit to which the output signal of the adder circuit α& is input, and pre-emphasizes the input signal and outputs it. (A) is an FM modulation circuit connected to the α field of the pre-emphasis circuit, which FM modulates carrier waves with six center frequencies based on the input signal, forms a recording signal consisting of an FM high frequency signal, and outputs it. . CI+ is a recording amplifier connected to the modulation circuit (4), which amplifies the input recording signal and outputs it.

is an analog switch for recording output that has two input terminals (22a) and (22b);
The recording signal of the amplifier C1l is input to fL), and the other input terminal (22b) forms an idle terminal, and each sector (SO) to
It is switched to the input terminal (22n) only during the recording period (S3), and a recording signal is output to the recording magnetic head of the disk via the recording audio output terminal.

C! The gradient is determined by detecting the waning magnetic flux of the PG yoke provided on the disk and determining the data start point (
This is a data start point detection circuit that detects the data start point CDSP) and outputs a detection signal DC at the detection timing of the data start point CDSP). The problem is the detection signal Dtfm
It is a recording gate generation circuit that can be inputted.It receives 3 fsc of the reference clock terminal ω together with the detection signal Dt, counts the reference clock signal based on the input timing of the detection signal Dt, and counts the reference clock signal to the sector specified by the sector selection signal ( So) to (S3) A recording gate signal Or indicating each recording time is generated and output.

Reference numeral 3 denotes a memory control circuit to which the recording gate signal Or of the generation circuit - is inputted, and a start signal St indicating the period of the audio signal to be recorded at the start signal input terminal +3 is inputted together with the recording gate signal Gr, and both signals Gr , S t, each write control signal Wm + , Wm2゜W
s+, Ws2 and each read control signal Rm+, 1
m2. Rs+.

R82 is formed and output.

Note that the write control means is formed by the switch (7), Qη and the control circuit 2, and the read control means is formed by the switches QCj, Q■ and the control circuit Cε, and the D/A conversion circuits 0Q to 0Q form the read control means. A recording output means is formed by the signal path of the output terminal.

On the other hand, the video signal format of the disc is set to 525 fins and 60 fins, and during recording and playback, the disc ! The magnetic head rotates at a period of /60 sec, and the time during which the overlap period Ta and period T4 in FIG. 4(b) are accessed by the magnetic head is constant regardless of compression.

Also, based on the signal recording band of the disk, each memory (
8), <9), 02), 03 read speed and each memory (8), (9), α4. The D/A conversion frequency of the audio data read from Zeng is fixed,
The frequency of the compressed audio signal recorded on the disc remains constant regardless of the compression ratio.

Since the compression ratio of the audio signal is determined by the ratio of the A/D conversion frequency and the D/A conversion frequency when generating audio data, the A/D conversion frequency is determined by the compression ratio and the D/A conversion frequency. Determined based on.

Furthermore, period, period T3. The amount (length) of the audio signal compressed and recorded by the UE in each T4 is determined by the compression ratio and the period T3. It is determined based on the access time of the portion T4, and as mentioned above, when the compression ratio is 320 times, 640 times, and 1280 times, the overlap period T3 is approximately 1.4 m5.
The audio signals of eC, zgmsec, and 5.7m5ec are compressed and recorded, and in one period T4, approximately 1.3SeC,
Audio signals of 1 sector each of 2, 5, and 5 seconds are compressed and recorded.

Therefore, the first . 2nd main memo l)
(8) and (9), the first one storing the audio data of the audio signal having the capacity and the length to be compressed and recorded in the period T3;
2nd submemory 1, 10Z), (The capacity of 13 is constant depending on the compression ratio, 1st main and submemory 1) (8),
(a) Sum of capacities, 2nd main, sub memo 1, 1 (9),
The sum of the capacities of 03 is the capacity of one sector of audio data newly recorded in period T4.

(Movement) Now, the audio signal shown in Fig. 2 (a) is input to the input terminal (2), and based on the operation of the record button of the device, the high level start signal St shown in Fig. 2 (b) is input to tx. Assume that an input is made during a period of ~ty, and a command is given to record an audio signal of tx-ty.

At this time, the audio signal at the input terminal (21) is filtered in a band corresponding to the compression ratio by filling (3) to prevent aliasing noise during D/A conversion (based on the digital processing at the subsequent stage). Take control.

By the way, the signal recording band of the disk can be set on the groove 32 based on the center frequency (6-stop) of the carrier wave of the modulation circuit.

Therefore, the compression ratio is 320x, 640x, 1280x
When doubled, that is, approximately 5 seconds per track, I
Q.

When recording an audio signal of
The band is variable.

The audio signal band-limited by the fill (3) is then logarithmically compressed in amplitude by a reduction circuit (4) and then input to a conversion circuit (5).

At this time, the conversion circuit (5) samples the input audio signal based on the clock signal of the input terminal (6), A/D converts it into digital audio data, and switches the switch (
7) Output audio data to αη.

Note that when the compression ratio of the audio signal is 31 times, the frequency of the clock signal at the beginning of the input terminal is a predetermined frequency set based on the signal recording band of the disk.

For example, since it is fixed at 3 fsc (fsc is the color subcarrier frequency), the 70 clock signal at the input terminal (6) needs to be a Si-divided signal of the clock signal at the input terminal QFj. The clock signal at the input terminal (6) is formed by dividing the frequency of the 70 clock signal at the input terminal αG by 1.

Then, each control signal Wm r , Wm2 , 'iV
Switch based on s t, Ws 2 (7),
(By switching 1η, the audio data output from the conversion circuit (5) is divided into 1-sector units that are compressed and recorded during period T4 in FIG.
(8), one memory pair with a capacity of one sector of audio data formed by α2, and the second main and sub memories (9),
The l-sector amount of audio data formed by Q3 is written alternately to the other memory pair.

By the way, the audio signal of tx-ty in FIG. 2(a) is
If we time-divide σ corresponding to l sector amount of audio data by 2, that is, by each period Ta, then tx~ta, ta
~tb, tb ~tc, tc-ta period Ta
and the remaining td-ty part.

Note that the period Ta changes depending on the compression ratio, and when the compression ratio is 32
Approximately 1.3 SeC for 0x, 640x, and 1280x
, 2,5511 IC.

5 seconds each.

The control circuit (■) receives the start signal St of the input terminal gold.
Based on the rising edge of , the period Ta is counted repeatedly, and the periods Twm and Tws obtained by dividing the period Ta into (N-1)/N and I/N are counted, and tx-Vta, tb
In the odd-numbered period Ta of ~Lc, FIG. 2(C),
As shown in (d), tx -ta', tb -t
A high-level control signal Wn is output during a period Twm of c', and a high-level control signal Ws+ is output during a period Tws of ta' to ta and tc' to tc.
-tb, tc-md even numbered period Ta is,
As shown in (0) and (r) in the same figure, ta -tb'
, tc - td', and outputs a high level control signal Wm2 during the period Twm, and also outputs the high level control signal Wm2 during the period Twm of tb' to tb, t
High level control signal Ws2 during period TWS from d' to td.
Output.

Furthermore, the switch (7) outputs the input audio data from the output terminal (7a) during the odd-numbered period Twm when the control signal Wmt is at a high level, and outputs the input audio data from the output terminal (7a) during the even-numbered period Twm when the control signal Wm2 is at a high level. During the period TWm, the input audio data is also output from the output terminal (7b), and the switch Ql) outputs the input audio data from the output terminal C11a) during the odd-numbered period Tws when the control signal Wsl is at a high level. At the same time, the input audio data is output from the output terminal (Ilb) during even-numbered periods Tws when the control signal Ws2 is at a high level.

Also, 1st. 2nd main memo') (8) and (9) are
The first . 2nd secondary memory (a), α
J is controlled to be written by the high level of control signals WSI and Ws2, respectively.

Therefore, in the first period Ta of tx-ta, the first period Twm is caused by the high level of the control signal W+n+.
, i.e., the overlap period T of the period T4
The audio signal data that is compressed and recorded in a short part by 3 is
The control signal W is written into the first main memory (8) and the control signal W
Due to the high level of s+, the audio data of the remaining period Tws, that is, the data of the audio signal to be compressed and recorded in the length of the remaining overlap period T3, is transferred to the first sub memory l.
) 03 is whistled.

Furthermore, in the second period Ta of ta-tb, the audio data of the period Twm is written to the second main memory (9) by the high level of the $ control signal Wm2, and the high level of the control signal WS2 The audio data of the remaining period Tws is written to the second submemory l)'13.

Thereafter, by repeating the same operation, the third period Ta
One sector of audio data is stored in the first main and secondary memories (8).
, divided into Q2 and written in 1 of the fourth period Ta.
The sector amount of audio data is stored in the second main and sub memories (9),
It is divided into 0.3 parts and written.

Note that for the remaining audio data between td and tyO, which is less than the period Tm, the control signal Wm+ is applied between td and tyO.
becomes high level and is written to the first main memory (8).

On the other hand, when the first write to the second main memory (9) is reached, the magnetic head is moved to an access position of an unrecorded empty track in order to start recording on the disk.

The track access time of the magnetic head is approximately 6m5e.
c/) is set on the rack, and the magnetic head is in the initial position.
Even if the car moves a large number of tracks, for example 50 racks, and moves to the access position of an empty track, the time required for the move will be 304 khu.

As mentioned above, even when the period Ta is the shortest, that is, when the compression ratio is 320 times, it is approximately 1635 times, which is the time required for the movement of the magnetic head.

For example, 304m5ec is sufficiently short compared to the period Ta,
The magnetic head instantly moves to the access position of the empty track, and preparation for recording is completed.

Further, when the detection circuit - detects the data start point (DSP) shown in FIG. 4(a), it outputs a detection signal Dt to the generation circuit C25+.

Then, the generation circuit counts the clock signal of the input terminal +3 based on the input timing of the detection signal Dt, and calculates each period Twm from the period Twm in which the recording preparation is started.
That is, in each period TWm after ta, which is delayed by a period Ta from the start of writing of audio data ty, a recording gate signal Or that becomes high level is generated only during the recording period in which the sector specified by the sector selection signal is accessed. and output it.

That is, based on the input timing of the detection signal Dt,
The generation circuit has an empty track (1) shown in FIG. 3(a).
At the same time as detecting the data start point (DSP), the clock signal of the input terminal ■ is counted based on the data start point (DSP), and in each period Twm, after the movement time until the recording preparation is completed has elapsed. (
, recording periods Tso, Ts+ in which each sector (SO) to (S3) is accessed by the magnetic head.

TS2 and Tsa are each detected.

Then, during the period Tvm in which the sector (SO) is specified by the sector selection signal, the generation circuit generates a high-level recording gate signal Gr during the recording period Tso, as shown in FIG. 3(b). period T during which each of sectors (Sl) to (S3) is specified by the sector selection signal.
As shown in FIG.
．． A high-level recording gate signal Or is generated and output for each TS3.

Note that each recording period Tso to TS3 is shown in Fig. 4 (
This is a sufficiently shorter time than B 4 m5ec of the format b), that is, the period T W m.

During the period 'pwmK' during which audio data is written into the second main memory (9), the second main memory (alpha) and the first main memory are written.

Submemory l) (8), (Read out the audio data held in L21 and write it to the odd numbered sector (So) or (S2
), and during the period Twm during which the audio data is written to the first main memo (January 8), the compressed audio signal is stored in the first secondary memory (A), the second main and secondary memories (9), and Q3. The control circuit (
28) is based on the input recording gate signal Gr,
At the timing explained below, each control signal Rrr++,
1m2. Rs+ and R52 are formed and output.

That is, the control circuit controls the input recording gate signal O.
Based on r, an overlap period T3 is delayed by the period T++Ts in FIG. 4(b) from the leading edge of each recording period Tso-TS3, and a main memory shorter than period T4 from the trailing edge of the overlap period T3 by the period T3 Data recording period Trm (= T4 - Ta), nll memory data recording period T of the length of period T3 from the trailing edge of the period Trm
rs (==Ta) is detected and the recording period during the period Twm in which audio data is written to the second main memory (9), that is, the recording period Tso of each of the odd-numbered sectors (80) and (82).

TS2 outputs a high-level control signal R32 during period T3, and outputs high-level control signals 几no and Rs+ during periods Trrn and Trs.
) during the period Twm in which audio data is written,
That is, in the recording periods TSI and TS3 of even-numbered sectors (81) and (83), a high-level control signal Rs+ is output in the period T3, and the periods Trm and T
A high level control signal Rm2.rs is applied to Rm2.rs. Output R82.

Note that the recording period Ts of the first recorded sector (So)
Since there is no audio data to be recorded in the overlap period T3 in the period T3, the output of the control signal R82 in the period T3 is stopped.

Now, during the period Twm between ta and tb' in FIG. 3
When the recording gate signal Gr of the recording period Tso shown in FIG.
SO), as shown in FIG. 3(f') and (g), during the periods Trm and Trs during the recording period Tsa,
High-level control signals Rrru and Rs+ are output, respectively.

Next, during the period Twm from tb to tb' in FIG. 2(C) during which the second writing to the first main memory (8) is performed, the controller selects the sector selection signal that specifies the safeter (St) as shown in FIG. When the recording gate signal Or of the recording period Ts+ in (C) is input to the control circuit (28), the second recording is performed in the sector (St) at this time, so that the signals shown in FIG.
As shown in (i) and (j), during the period T3. during the recording period Tsl. High level control signal R to Trm and Trs
s+, 1m2. R82 is output respectively.

Thereafter, in the same manner, the second writing to the second main memory (9) is performed during the period T of tc - td' in FIG. 2(e).
zvrr+: Then, the recording gate signal G of the recording period TS2
Based on the period T3.r in the recording period TS2. Trm,
High-level control signals R32, Rmt, and S are applied to Trs.
The period Tw from td of the second section (C) during which 1 is output and the third write to the first main memory (8) is performed.
Based on the recording gate signal Gr of the recording period Ts3, the period T'3 . Trm, Trs
and a high level control signal Rs+.

1m2. Rs2 is output respectively.

Then, each recording period Tso shown in FIG. 2(g).

Based on the high level recording gate signal Or of Ts l , TS2 and Tsa, control signals Rm+ and Rs+
, 1m2. R82 is the same figure), (i),
(,i), (k) become high level in the order shown respectively.

The width of the high level of each signal Gr, Rmt, . In (k), each signal Gr, Rm+,...
The time axis is shown enlarged to clarify the level change of R32.

Then, the first . The second main memories (8) and (9) can each be controlled for reading, and the control signals Rs+ and R32
1st place due to the high level of. If the second sub-memory access and αa are each controlled for reading, and at this time the compression ratio of the audio signal is set to 1, then each memory (g), (9), o
3. Zeng reads audio data at a read speed that is ΔI times the write speed.

Further, the switch α0 outputs the audio data of the input terminal (108) read from the first main memory (8) to the conversion circuit 05 during the period Trm when the control signal Rm+ is at a high level, and the control signal Rm2 is output to the conversion circuit 05. The input terminal (
The audio data of 10b) is output to the conversion circuit α, and the switch a→ is operated during the period T during which the control signal Rs+ is at a high level.
3. Output the audio data of the input terminal (14a) read from the first sub-memory α2 to Trs to the conversion circuit (15), and set the period T during which the control signal R82 is at a high level.
3. The audio data at the input terminal (14b) read from the TrsK second sub-memory α4 is output to the conversion circuit 0.

Therefore, during the period Twm of ta-tb' in FIG. 2(e) when the audio data is first written into the second main memory (9), the periods Trm and Trs during the recording period Tso, that is, sectors (SO) While the period T4 part is accessed, the first one sector of audio data held in the first main and sub memo IJ (8) and CLa is time-compressed at a compression ratio of ~■ times and sent to the conversion circuit. You can input only α.

Also, during the period Twm of tb-tc' in FIG. 2(C) when the audio data is written to the first main memory (8) for the second time,
The first period T3 of the recording period Ts+, that is, the sector (
When the period T3 part of St) is accessed, the first
The same audio data as the last part of the previous sector (So) held in the sub-memory a is time-compressed at a compression ratio of M times and input to the conversion circuit 09, and the period following period T3 is Trm, Trs, i.e. sector (St)
When a portion of period T4 of is accessed, the second master.

One sector of audio data held in the submemory 1) (9), 03 is time-compressed at a compression ratio of M times and input to the conversion circuit αQ.

Thereafter, in the same manner, during the recording period TS2 in the period Twm of tc-td', the first period Ta, that is, the sector (
When the period T3 part of S2) is accessed, the second
The same audio data as the last part of the previous sector (St) held in the sub-memory μs is time-compressed and input to the conversion circuit Q5, and the period Tr following the period T3 is
m, Trs, that is, when the period T4 portion of the sector (S2) is accessed, the first main and secondary memories (8)
, 1 sector of audio data held in O'ZI is
It is time-compressed and input to the conversion circuit α9, and during the recording period Tss in the period Twm from td, when the part of the period T3 of the sector (S3) is accessed, the first sub-memory α
The successfully held audio data is time-compressed and input to the conversion circuit αQ, and the period Tr following period T3 is
When the portions m and TrS are accessed, one sector of audio data held in the second main and submemories (9) and a3 is time-compressed and input to the conversion circuit αQ.

Note that the first main memory (8) is located at td-ty in FIG. 2(C).
It is necessary to record the remaining audio data written in the sector (SO) of the next empty track.

Therefore, when the period Twm from td ends, the magnetic head moves to the access position of the next empty track, and the next period Twm after the period Twm from td ends.
In n, when the period T3 part of the sector (SO) is accessed, the audio data held in the second sub-memory αQ is time-compressed and input to the conversion circuit 05 based on the control signal R82. When the period Trm following the period T3 of SO) is accessed, the audio data held in the first main memory (8) is time-compressed and input to the conversion circuit 0Q based on the control signal Rm+.

Then, the audio data input to the conversion circuits 0 and 9 is converted into D and /A based on the clock signal of the input terminal 0Q, and at this time, each memory (8), (9), α
Audio data can be read out from Z, 03 at a reading speed M times the writing speed, and the conversion circuit α9 performs digital conversion at a speed M times the analog conversion of the conversion circuit (5). is the input terminal (2
] The audio signal is time-compressed at a compression ratio of 320 times, 640 times, or 1280 times, that is, a compressed audio signal is generated and output. In addition, input terminal α
The frequency of the Q clock signal is set to, for example, 3 fsc based on the signal recording band of the disk.

That is, from the conversion circuit Q5, each sector (SO) to (
S4) Recording period Tso during which each is accessed,...
・.

Tsa includes the audio data of the overlap period T3 consisting of the remaining I/N of the audio data of one sector of the previous sector, and the audio data of one sector amount newly recorded in the period T4 of the sector. A compressed audio signal for recording in one sector is output at the timing of period Tγ shown in FIG. 4(b).

The conversion circuit (ISO compressed audio signal is input to the pre-emphasis circuit QC4 via the adder circuit α mark, and at this time, each sector (SO
) to (S4), the start flag, the end flag, and the control code are output at the timing when the period Tα, Tδ portion and the period Tβ portion are accessed, so the pre-emphasis circuit 09 is provided with a start signal for the compressed audio signal. , an end flag and a control code are input in the format shown in FIG. 4(b).

Furthermore, the recording signals of each one sector (So) to (SS) that have been pre-emphasized by the pre-emphasis circuit 01 are modulated by the modification circuit (1) and converted into recording signals, and then sent to the amplifier Q11. You can input it to the switch ■ through it.

And generation circuit c! Based on the recording gate signal Or in 5), the switch ◎ is switched to the input terminal (22a) during each recording period Tso to Tsa, so that when each sector (SO) to (SS) is accessed by the magnetic head, , the recording signal in the format shown in FIG. 4(b) to be recorded in each sector (SO) to (SS) is supplied from the output terminal Ω to the magnetic head,
A compressed audio signal is recorded on SS) in the format shown in (1) of the same figure.

Therefore, the first. Second sub-memory α2. αa is shared to store the audio signal compressed and recorded in the overlap period T3 of each sector (SO) to (SS) and the audio signal compressed and recorded in the last part of the period T4, and The first and second main memories (8) and (9) have a capacity to store (N-1)/N of data, and the first and second main memories (8) and (9) have a capacity to store the remaining I/N. A second sub-memory (6) is provided,
Compressed audio signals can be recorded on a disk in the format shown in Figure 4(b), with a relatively small capacity, that is, 2
A compressed audio signal recording device can be easily and inexpensively formed using a memory having a capacity to store audio data of a sector amount.

If the D/A conversion frequency of the conversion circuit 0 based on the lock signal is set to 3 fsc, in order to store 2 sectors of audio data, the first main, sub memory (8), α and second main , the submemory (9), and α3 can be formed using a memory having a capacity of approximately 90 bytes.

By the way, in the case of Figure 1, the data start point)
Each sector (
Recording periods Tso to Tsa are set for SO) to (SS).

The tolerance of the detection timing of the data start point (DSP) is set to ±IH based on the standard, and when the device or disk is replaced, the tolerance of the detection timing of the data start point (CD5P) is set to ±IH.
A shift in H occurs, and a shift occurs in the access position of the magnetic head.

However, in the case of Fig. 1, the device and disk are not replaced during recording, and if the compression ratio is 640 times, for example, each sector (SO) to (SfK) is
S) is recorded, so the data start point)
The difference in the recording timing of each sector (SO) to (SS) based on the difference in the detection timing of the (DSP) is, for example, only ±3μ3μ east, and at this time, each sector (
A space of approximately 30 μ east between SO) and (SS) (
Since the space (Ss) is provided, the discrepancy in recording timing is absorbed by the space (Ss), and sectors (SO) to
Overlapping recording of (SS) etc. does not occur.

In the above embodiment, the tx-ty audio signal was recorded on the disk, but it is of course possible to record an audio signal shorter than 1 trough or an audio signal for several tracks on the disk.

Further, in the embodiment, the audio signal is converted into digital audio data and stored in the memories (8), (9), α2. αa
The audio signal may be written directly into the memory as audio data; in this case, the conversion circuit (5)
, αQ, etc. can be omitted.

Further, in the above embodiments, magnetic disks for electronic still cameras having the formats shown in FIGS. 4(a) and 4(b) were used as recording disks, but various magnetic, optical, etc. Of course, it can be used as a recording disk.

〔Effect of the invention〕

As described above, according to the compressed audio signal recording device of the present invention, there are two main memories each having a capacity to store (N-1)/N of audio data of one sector amount, and the remaining 1/N. Each sector is provided with two sub-memories with a capacity of
It is possible to record a compressed audio signal for recording of one sector consisting of /N and one sector amount of audio data recorded in the sector, and for example, to record a compressed audio signal of a magnetic disk for an electronic still camera. It is something that can be done.

[Brief explanation of the drawing]

1 to 3 show one embodiment of the compressed audio signal recording device of the present invention, FIG. 1 is a block diagram, and FIG.
) to (k) and FIGS. 3(a) to (j) are timing charts for explaining the operation, and FIGS. 4(a) and (b) are track formats of a magnetic disk for an electronic still camera. FIG. 3 is an explanatory diagram of a sector format. (1-track, (2)-recording audio input terminal, (3)-low-pass filter, (4)-noise reduction circuit, (5)-A/D conversion circuit, (7
), (IcI, (11), Q4), ■・
...Analog switch, (8), (9)...1st
．． 2nd main memory, α. (11... 1st. 2nd sub-memory, α needle... D/A conversion circuit, Q7) -°° flag, control code generation circuit, α mark... addition circuit, Ql... pre-emphasis circuit , 翰...FM modulation circuit, (211...recording amplifier,...recording audio output terminal, (241...data start point detection circuit, f...recording gate generation circuit, ■... Memory control circuit, (SO) ~ (S3)...
·sector.

Claims (1)

[Claims] (1) Each concentric track formed on a recording disk is divided into a plurality of sectors, and each sector has a
In a compressed audio signal recording device that time-compresses an audio signal and sequentially divides and records the audio signal, (N-1) of the audio data of one sector amount formed by dividing the audio signal into a predetermined amount of new recording in one sector. /N(
N>1), two main memories with a capacity to store the remaining 1/N, and the first (N-1)/N and the remaining 1 sector of audio data. Sequential writing to one of the main and secondary memories of 1/N of;
The first (N-1)/N of the next 1 sector amount of audio data
, write control means for alternately sequentially writing the remaining 1/N to the other main and secondary memories; and one of the secondary memories during writing to the one main memory;
Read control means for alternately performing sequential reading of the other main and secondary memories, sequential reading of the other secondary memory during writing to the other main memory, and sequential reading of the one main and secondary memories at a higher speed than writing. Based on the audio data read from each of the memories, 1
A compressed audio signal for recording of one sector consisting of the remaining 1/N of the one sector amount of audio data recorded in the previous sector and one sector amount of audio data to be recorded in the sector is 1. A compressed audio signal recording device comprising: recording/outputting means for sequentially outputting to the recording means of the disk during recording.