JPH0563584A - Signal recorder and/or reproducing device - Google Patents

Abstract

PURPOSE:To save the capacity of a recording medium by reducing the possibility of buffer overflow so as to enhance a bit compression efficiency. CONSTITUTION:Either equal length coding data from quantization sections 503,511 or unequal coding data from unequal code coding sections 504,512 are selected by changeover switches 521,523 and the selected data are recorded on a recording medium via a multiplexer 505 and a buffer 525. A discrimination section 520 compares a bit number of the equal length coding data from the quantization sections 503,511 with a bit number of the unequal coding data from the unequal code coding sections 504,512 and applies changeover control to the changeover switches 521,523 so as to select less bit number.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal recording device and / or reproducing device capable of recording and / or reproducing a signal in real time.

[0002]

2. Description of the Related Art First, let us consider a case where unequal length coding is adopted in a coding apparatus used in a signal recording apparatus and a decoding apparatus used in a signal reproducing apparatus. The unequal-length coding is also called variable-length coding, and is a coding method in which the length of the code to be allocated is changed according to the probability of occurrence of data. That is, by allocating short codes to those with a high probability of occurrence and long codes to those with a low probability of occurrence, it is possible to perform efficient coding with a short average code length. Huffman code is a typical method.

Generally, the data is treated as a statistical model and is modeled as a Gaussian distribution or a Laplacian distribution. Therefore, data with such a distribution is
Since the probability of occurrence at a small amplitude is high, a short code is assigned to data of a small amplitude, and the probability of occurrence at a large amplitude is low. Therefore, a long code is assigned to data of a large amplitude.

FIG. 9 shows an example of a conventional coding apparatus to which such an unequal length coding system is applied, and FIG. 10 shows an example of a decoding apparatus. In the encoding device shown in FIG. 9,
The input analog tone signal is A / D (analog / digital) converted by an A / D converter 301 to be a digital signal, which is collectively output for each predetermined number of samples for frame processing. This data is frequency-analyzed by the frequency analysis unit 302 and divided into, for example, signals of two bands (frequency bands). The signal of each band subjected to frequency analysis is supplied to the scale factor calculation units 307, 3
12, the scale factors for quantization are obtained, and the scale factor quantizers 308, 3
13 respectively. Scale factor quantizer 3
In 08 and 313, each scale factor is quantized by a predetermined number of bits, and the scale factor dequantization unit 309,
314 and multiplexer 305. The scale factor dequantization units 309 and 314 dequantize each quantized scale factor. These dequantized values are supplied to the quantizers 303 and 310, respectively, and also supplied to the quantization bit number determination unit 315. The quantization bit number determination unit 315 determines the number of bits for quantizing each band signal, and sends these bit numbers to the quantization units 303 and 310 and the unequal length code encoding units 304 and 311. The quantizers 303 and 310 quantize the data of the two band signals from the frequency analyzer 302, respectively, using the dequantized scale factors and the numbers of bits. The band data quantized by the quantizers 303 and 310 are input to the unequal length code encoders 304 and 311 and unequal length coded. These data are input to the multiplexer 305. Therefore, the bit stream data output from the multiplexer 305 has an unequal number of bits for each frame. Therefore, the bit stream data is stored in the buffer 306, and is sent to the recording path of a constant bit rate or a recording / reproducing system such as DAT (digital audio tape recorder) via the buffer 306.

Next, in the decoding apparatus shown in FIG. 10, since the bit stream data obtained from the transmission line or DAT has an unequal number of bits for each frame, it is stored in the buffer 401. The bitstream data is divided by the demultiplexer 402 into main data and scale factors. Each band data is input to the unequal length code decoding units 403 and 408, decoded respectively, and sent to the dequantization units 404 and 409. The quantized scale factor of each band is dequantized by scale factor dequantization units 407 and 410,
Inverse quantization units 404 and 409 and quantization bit number determination unit 4
11 respectively. Quantization bit number determination unit 411
Then, the number of quantization bits for each band is determined, and the unequal length code decoding units 403 and 408 and the inverse quantization units 404 and 40 are determined.
Output to 9 respectively. The dequantization units 404 and 409 dequantize the quantized data of each band based on the dequantized scale factor and the number of quantization bits for each band. The frequency synthesizer 405 synthesizes the dequantized data of each band, and
The A (digital / analog) converter 406 performs D / A conversion and outputs.

[0006]

By the way, in the conventional recording / reproducing apparatus adopting the unequal length encoding as shown in FIGS. 9 and 10, the bit rate for transmission or recording is constant. A buffer is provided to perform recording and reproduction in real time. That is,
The state of the buffer at a certain time is as follows: B (n) = B (n-1) + Bin (n) -Bout (n) [bits] (1) In this equation (1), n indicates the nth storage update cycle of the buffer, B (n) is the amount of data stored in the buffer, and Bin (n) is the amount of buffer input data.
Bout (n) indicates the amount of buffer output data. Bin (n) in the equation (1) has an unequal length, and Bout (n) is constant. In such a case, for example, when long codes are intensively allocated, the amount B (n) of data to be stored in the buffer exceeds the buffer capacity Bmax (B (n)> Bmax). When a so-called overflow condition occurs and short codes are intensively allocated, there is nothing to be sent or recorded from the buffer (B (n) <0), so-called underflow occurs. There is.

In order to solve this, conventionally, (a) a large memory capacity is assigned to a buffer. (B) When overflow or underflow is likely to occur, the scale factor for quantization is changed to a value in the direction of suppressing the occurrence. Etc. are taken. The measure (b) is to monitor the state of the buffer and increase the scale factor when overflow is likely to occur to reduce the quantized signal. Since a short code is assigned to a signal having a small amplitude, overflow can be prevented. If underflow is likely to occur, reduce the scale factor and increase the quantized signal. Since a long code is assigned to a signal having a large amplitude, underflow can be prevented.

However, in these measures, the measure (a) of increasing the buffer capacity increases the hardware scale, and the measure (b) of operating the scale factor has a desired characteristic. There is a problem that you can not get.

The present invention has been made in view of the above situation, and when the unequal length coding method is adopted, the possibility that the buffer overflows can be reduced and the recording capacity of the recording medium can be reduced. It is an object of the present invention to provide such a signal recording device and / or reproducing device.

[0010]

A signal recording apparatus and / or reproducing apparatus according to the present invention is a signal recording apparatus and / or reproducing apparatus capable of recording and / or reproducing a signal in real time. When recording with unequal length coding, if the number of bits per frame or band that has been unequal length code is longer than the number of bits when unequal length coding, select the code of equal length coding When the data is recorded and played back, the above problem is solved by decoding according to the selection result.

[0011]

Since the number of bits per frame or band in the case of unequal length encoding does not become longer than the number of bits in the case of equal length encoding, the possibility of buffer overflow is reduced, or The capacity can be reduced. Moreover, since the average number of bits recorded can be reduced, bit compression can be performed more efficiently, and the capacity of the recording medium can be saved.

[0012]

The basic principle of the embodiments of the present invention will be briefly described below. In a signal recording apparatus and / or reproducing apparatus capable of recording / reproducing in real time, quantized musical sound data or the like is encoded by an encoder. When recording with unequal length coding, if the number of bits per frame or band that has been unequal length code is longer than the number of bits when unequal length coding, then the code is equal length coded Is selected, and if the code is short, a code that has been encoded with unequal length is selected and recorded together with the selection result. When playing, read the above selection result, select the decoder to be decoded according to the selection result,
The above-mentioned musical tone data and the like are decoded.

Here, as the state of the buffer, the conventional system as shown in FIG. 9 is Ba (n) = Ba (n-1) + Bina (n) -Bout (n) [bits] (2) And the state of the buffer using the above technique is expressed as Bb (n) = Bb (n-1) + Binb (n) -Bout (n) [bits] (3), Binb (n The maximum value Binb (n) max of) is a fixed number of bits when equal-length coding is performed. That is, a code that is unequal-length coded with a fixed number of bits or more when equal-length coded is not adopted, but a code that is equal-length coded is adopted. That is, the designated code is adopted.

Therefore, since Bina (n) max ≥ Binb (n) max (4), it becomes possible to allocate long codes intensively (for example, when maximum values are continuously allocated). Is
The buffer status has a smaller number of bits than the conventional method.
Therefore, the possibility of overflow can be reduced.

The average bit rate is the average value of Bina (n) and Binb (n). As can be seen from the equation (4), the maximum value is suppressed, so that the average bit rate, which is the average value, is reduced.

In the equations (2), (3) and (4), the meaning of each term is B as in the case of the equation (1).
(n) indicates the amount of data stored in the buffer, Bin (n) indicates the amount of buffer input data, and Bout (n) indicates the amount of buffer output data. In addition, n in parentheses indicates the nth of the storage update cycle of the buffer, the attached character a attached to each item indicates the conventional method, and b indicates the case where the technique of the embodiment of the present invention is used.

The preferred embodiments of the present invention will be described below.
A description will be given with reference to the drawings. 1 is a block circuit diagram showing an embodiment of a signal recording apparatus according to the present invention. In the signal recording apparatus shown in FIG. 1, the input analog musical tone signal has an A / D (analog / digital) converter 50.
At 1, the signal is A / D converted into a digital signal, which is then collected and output for each predetermined number of samples of data for frame processing. This data is used as the frequency analysis unit 502.
Frequency analysis with, for example, two bands (frequency band)
The signal is divided into. The frequency analyzed signals of each band are
The scale factors are sent to the scale factor calculators 508 and 513, the scale factors for quantization are obtained, and the scale factors are sent to the scale factor quantizers 509 and 514, respectively. The scale factor quantizers 509 and 514 quantize each scale factor with a predetermined number of bits, and send the quantizers to the scale factor dequantizers 510 and 515 and the multiplexer 505, respectively. The scale factor dequantization units 510 and 515 dequantize each quantized scale factor. These dequantized values are supplied to the quantizers 503 and 511, respectively, and are also supplied to the quantization bit number determiner 516 to determine the number of bits for quantizing each band signal and quantize them. Send to the conversion units 503 and 511.

The quantizers 503 and 511 quantize the data of the two band signals from the frequency analyzer 502 by using the dequantized scale factors and the numbers of bits. These quantizers 503, 51
The data quantized by 1 are sent to the unequal length code encoding units 504 and 512, respectively, and to the selected terminals a of the changeover switches 521 and 523, respectively. The unequal-length coded data by the unequal-length code encoders 504 and 512 are sent to the selected terminals b of the changeover switches 521 and 523, respectively. In addition, the quantizer 50
3, 511 output equal-length coded total bit numbers, and unequal-length code encoding units 504, 512 output unequal-length encoded total bit numbers, which are sent to the determination unit 520. Sent. The determination unit 520 compares the total number of coded bits in the frame in the case of equal-length coding with the total number of coded bits in the case of unequal length coding, and determines which of the smaller coding methods is used. A 1-bit judgment result instructing selection is output. This determination result is supplied to the changeover switches 521 and 523, and either the unequal-length encoded code (data) or the equal-length encoded code (data) is switched and selected according to the determination result. And sends it to the multiplexer 505. The multiplexer 505 multiplexes the above-mentioned input data and sends the bit stream data to the buffer 525.
To a recording / reproducing system such as a transmission line or DAT (digital audio tape recorder).

Next, FIG. 2 shows an embodiment of the signal reproducing apparatus according to the present invention. In the signal reproducing apparatus shown in FIG. 2, the bit stream data read from the transmission path or DAT is stored in the buffer 401 in consideration of the fact that the number of bits is unequal length for each frame. The demultiplexer 603 first reads out the above determination result and controls the changeover switches 621 and 622 in accordance with the determination result in order to know the encoding system of the data to be read out. That is, the changeover switches 621 and 622 are
In the case where the above-mentioned determination results instruct the selection of the equal-length encoding method, the switching is controlled to the selected terminal a side in both cases, and when the determination results instruct the selection of the unequal-length encoding method. Both are switched and controlled to the selected terminal b side. The bitstream data is divided by the demultiplexer 603 into the quantized main data of each band and the quantized scale factor. Each main band data is sent to the changeover switches 621 and 622, the outputs from the selected terminals a of these changeover switches 621 and 622 are sent to the dequantizers 605 and 610, respectively, and the selected terminals b are sent. The output of is transmitted to the unequal length code decoding units 604 and 609, respectively. Therefore, when the determination result indicates the unequal length coding method, each band data from the demultiplexer 603 is changed.
After being sent to the unequal length code decoding units 604 and 609 through the selected terminals b of Nos. 21 and 622, respectively, and decoded,
It is sent to the inverse quantizers 605 and 610, respectively. When the determination result indicates the equal-length coding method, each band data from the demultiplexer 603 is transferred to the changeover switch 6
21 and 622 are respectively taken out from the selected terminals a, passed through the unequal length code decoding units 604 and 609, and sent to the dequantization units 605 and 610, respectively.

The quantized scale factor of each band from the demultiplexer 603 is dequantized by scale factor dequantization units 608 and 611, respectively,
Inverse quantization units 605 and 610 and quantization bit number determination unit 6
13 respectively. Quantization bit number determination unit 613
Then, the number of quantization bits for each band is determined and output to the inverse quantizers 605 and 610, respectively. Inverse quantizer 6
In 05 and 610, the quantized data of each band is dequantized by the dequantized scale factor and the number of quantization bits for each band. The frequency synthesizer 606 synthesizes the inversely quantized data of each band, and the D / A (digital / analog) converter 6
At 07, D / A conversion is performed and output.

According to the signal recording apparatus and the signal reproducing apparatus described above, the number of bits per frame in the case of unequal length coding does not become longer than the number of bits in the case of equal length coding, so that the buffer 525 is used. It becomes difficult for overflow to occur and the buffer capacity can be reduced. Moreover, since the average number of bits to be recorded can be reduced, the bit compression efficiency is improved and the capacity of the recording medium can be saved.

Here, FIGS. 3 and 4 show a signal recording device and a signal reproducing device, respectively, in the case of switching the coding method according to the number of bits for each band. The signal recording apparatus shown in FIG. 3 is different from the signal recording apparatus shown in FIG. 1 in that the number of bits encoded by the unequal-length code is compared with the number of bits encoded by the equal-length code to make a determination. One of the points is that each band is provided with a determination unit that does.
That is, the quantization unit 50 for one band data
The number of equal-length coded bits from 3 and the number of non-equal-length coded bits from the unequal-length code encoding unit 504 are sent to the determination unit 522 for comparison, and the smaller encoding method is selected. It is selected as the encoding method for the band. Also,
The number of equal-length coded bits from the quantizing unit 511 for the other band data and the unequal-length code encoding unit 51
The unequal length coded bit number from 2 is sent to the determination unit 524 of the band and compared and determined. Then, the changeover switch 521 is switch-controlled according to the determination result from the determination unit 522, and the changeover switch 522 is switch-controlled according to the determination result from the determination unit 524. In addition, these determination units 52
The determination results for each band from 2, 524 are recorded respectively. Since other configurations and operations are the same as those of the embodiment shown in FIG. 1 described above, the corresponding parts in the drawing are designated by the same reference numerals and the description thereof will be omitted.

Next, the signal reproducing apparatus shown in FIG. 4 is different from the signal reproducing apparatus shown in FIG. 2 above, because the judgment result for each band is recorded on the recording medium, and therefore, the signal is reproduced via the demultiplexer 603. The determination results for each band are read out, and the changeover switch 62 for each band is read.
That is, switching control is performed on each of No. 1 and 622. Since other configurations and operations are similar to those of the embodiment shown in FIG. 2 described above, the corresponding parts in the figure are designated by the same reference numerals and the description thereof will be omitted.

FIG. 5 and FIG. 6 are views showing the manner of processing for explaining the respective processing operations of the above-mentioned two types of embodiments. That is, FIG. 5 shows an example of the processing state of the embodiment shown in FIG. 1 and FIG. 2, where the horizontal axis shows the number of frames to be sequentially processed (frame processing number), and the vertical axis shows per frame. The number of bits (number of bits / frame) is shown. In FIG. 5, the solid line represents the number of bits in the case of unequal length coding, the broken line represents the number of bits in the case of equal length coding, and the thick line represents the number of bits selected in this embodiment. The 1-bit determination result indicates that "1" selects the unequal-length coding method and "0" selects the equal-length coding method. For example, in the first frame, the number of bits in the case of unequal length encoding is longer than the number of bits in the case of equal length encoding, so the equal length encoding method is selected.
The same applies to the second frame. In the third frame, the number of bits in the case of equal-length coding is longer than the number of bits in the case of unequal-length coding, so the unequal-length coding method is selected. In this way, the sequential processing is performed.

From the example of FIG. 5, the average bit rate is shorter in the case of unequal length coding than in the case of equal length coding, and the average bit rate is further shorter in the present embodiment than that. I understand that Further, since the maximum value of the number of bits assigned to each frame is limited to the number of bits in the case of equal length encoding, the number of bits in the case of unequal length encoding such as the case of the seventh frame is continuous. In this case, it can be seen that the buffer state is smaller in the case of the method of this embodiment.

Next, FIG. 6 shows an example of the processing state of the embodiment shown in FIGS. 3 and 4, in which the horizontal axis represents the band number and the vertical axis represents the bits per band in the order of the frame processing numbers. Each number is shown. The solid line represents the number of bits when unequal length coding is performed, and the broken line is the quantization bit number determination unit 5 described above.
16 indicates the number of bits in the case of equal-length encoding by the number of bits allocated to each band so that the number of bits per frame fits, and the thick line indicates the number of bits selected by the embodiments of FIGS. 3 and 4. Is represented. In the determination result, "1" corresponds to the unequal length coding method, and "0" corresponds to the equal length coding method. , Bold lines respectively indicate the number of bits selected in this embodiment. For example, in the first frame, the number of bits when unequal-length coding is performed in both the first band and the second band is shorter than the number of bits when unequal-length coding is performed. The method is selected. In the second frame, the unequal length coding scheme is shorter in the first band and the unequal length coding scheme is shorter in the second band. For the second band, the equal length coding method is selected. In this way, the processing is performed so that the optimum coding method is selected for each band in each frame.

Next, FIGS. 7 and 8 show the case where unequal length encoding is adopted in a recording / reproducing apparatus capable of recording / reproducing in real time using a semiconductor memory such as an IC memory card as a recording medium. An embodiment is shown in which the present invention is applied to a recording / reproducing apparatus that controls the writing / reading speed or time according to the number of bits per frame. That is, in the signal recording apparatus of FIG. 7, the number of bits per frame of the data of the selected encoding method from the determination unit 520 is used as the basic configuration of the embodiment shown in FIG. 1 (the total number of bits of each band). ) To the writing speed / time control unit 527, and the writing speed / time control unit 527
Then, the write (write) signal is controlled so that the writing speed of the band data is high when the number of bits is large and low when the number of bits is small according to the obtained number of bits. , Writing device 506
Output to. Or, always write at high speed,
The write signal time (write time) is controlled and output to the writing device 506. In the writing device 506, the bit stream data from the multiplexer 505 is transferred to the IC, which is a semiconductor memory, according to the write signal.
A process of writing to the memory card 507 is performed. Since the other structure and operation of this FIG. 7 are the same as those of the embodiment shown in FIG. 1 described above, the same reference numerals are given to the corresponding portions and the description thereof will be omitted.

As described above, since the writing speed or the writing time for the IC memory card 507 is controlled according to the number of bits per frame, recording by unequal length code can be performed without using the above buffer. .. Therefore, buffer processing and measures against overflow and underflow are not required, the hardware scale does not increase due to the buffer, and problems such as characteristic deterioration due to scale factor operation for measures against overflow and underflow do not occur.

Next, in the signal reproducing apparatus shown in FIG. 8, from a semiconductor memory, for example, an IC memory card 601, a reading device 602 and a demultiplexer 603,
First, the number of bits of data per frame to be read is read at a predetermined read speed. This value is sent to the reading speed / time control unit 612. The read speed / time control unit 612 reads the main data according to the total number of encoded bits so that the main data is read at high speed when the number of bits is large and at low speed when the number of bits is small ( The read signal is controlled and output to the reading device 602. Alternatively, the reading is always performed at high speed, and the time of the read signal is controlled and output to the reading device 602. The reading device 602 reads the bit stream data from the IC memory card 601 into the demultiplexer 603 according to the read signal. Since the other structure and operation of FIG. 8 are the same as those of the signal reproducing apparatus of the embodiment shown in FIG. 2 described above, the same reference numerals are given to the corresponding portions and the description thereof will be omitted.

According to such a signal reproducing device, the reading speed or the reading time is controlled according to the number of bits per frame of the data read from the IC memory card 601, so that the reproduction can be performed without using a buffer. .. Therefore, even if the unequal length code is adopted, buffer processing, overflow, and underflow countermeasures are not required, and the hardware scale does not increase, and overflow,
There is no problem such as characteristic deterioration due to scale factor operation or the like to prevent underflow.

The present invention is not limited to the above embodiment, and the number of divided bands may be three or more, for example. Further, it is needless to say that the present invention can be applied to any of a signal recording dedicated device, a signal reproduction dedicated device, and a signal recording / reproducing device. Further, the semiconductor memory serving as the recording medium is not limited to the IC memory card, and may be in the form of a memory cartridge, a memory pack, or the like. Furthermore, the present invention can be applied not only to recording / reproducing to / from a recording medium, but also to signal transmission via a transmission medium.

[0032]

According to the signal recording apparatus of the present invention, in a signal recording apparatus and / or reproducing apparatus capable of recording and / or reproducing a signal in real time, quantized digital data is encoded with unequal length. When recording, when the number of bits per unequal-length coded frame or band is longer than the number of bits when equal-length coding is performed, the equal-length coded code is selected, recorded, and reproduced. In this case, since the decoding is performed according to the selection result, the number of bits per frame or band when unequal length coding is
The number of bits does not become longer than that in the case of equal-length encoding, and the possibility of buffer overflow can be reduced or the buffer capacity can be reduced. Also,
Since the average number of bits recorded can be reduced, bit compression can be performed more efficiently, and the capacity of the recording medium can be saved.

[Brief description of drawings]

FIG. 1 is a block circuit diagram showing an embodiment of a signal recording device according to the present invention.

FIG. 2 is a block circuit diagram showing an embodiment of a signal reproducing device according to the present invention.

FIG. 3 is a block circuit diagram showing an embodiment of the signal recording apparatus according to the present invention in the case of switching the encoding method according to the number of bits for each band.

FIG. 4 is a block circuit diagram showing an embodiment of the signal reproducing apparatus according to the present invention in the case of switching the encoding method according to the number of bits for each band.

FIG. 5 is a diagram showing the manner of processing for explaining the operation of the embodiment shown in FIGS. 1 and 2;

FIG. 6 is a diagram showing a processing state for explaining the operation of the embodiment shown in FIGS. 3 and 4;

FIG. 7 is a block circuit diagram showing an embodiment of the signal recording apparatus according to the present invention, in which the writing speed or time is controlled according to the number of bits per frame.

FIG. 8 is a block circuit diagram showing an embodiment of a signal reproducing device according to the present invention, in which a reading speed or time is controlled according to the number of bits per frame.

FIG. 9 is a block circuit diagram showing an example of a conventional signal encoding device using an unequal length encoding method.

FIG. 10 is a block circuit diagram showing an example of a conventional signal decoding device using an unequal length coding method. 502 ... Frequency analysis unit 503, 511 ... Quantization unit 504, 512 ... Unequal length code encoding unit 505 ... Multiplexer 506 ... Writing device 507 ... IC memory card 508, 513 ... Scale factor calculation unit 509, 514 ... Scale factor quantization unit 510, 515 ... Scale factor inverse quantization unit 516. ... Quantization bit number determination unit 520, 522, 524 ... Judgment unit 521, 523 ... Changeover switch 525 ... Buffer 527 ... Writing speed / time control Unit 601 ... IC memory card 602 ... Readout device 603 ... Demultiplexer 604, 609 ... Unequal length code decoding unit 605, 6 0 ... Dequantization unit 606 ... Frequency synthesis unit 608, 611 ... Scale factor dequantization unit 612 ... Read speed / time control unit 613 ... -Quantization bit number determination unit 621, 622 ... Changeover switch 623 ... Buffer

Claims (1)

[Claims] 1. In a signal recording device and / or reproducing device capable of recording and / or reproducing a signal in real time, when the quantized digital data is unequal length encoded and recorded, it is unequal length encoded. When the number of bits per frame or band is longer than the number of bits in the case of equal-length encoding, the equal-length encoded code is selected and recorded, and when it is played back, decoding is performed according to the selection result. A signal recording device and / or a reproducing device characterized by being converted.