JPH08186502A - Coded signal playback device - Google Patents

Abstract

(57) [Abstract] [Purpose] When a coded signal is played back at variable speeds, the pitch of the playback sound is made the same as during normal playback by thinning processing or repeat processing, and noise is not noticeable at the boundaries of discontinuous frames. . A demultiplexer 12 performs thinning processing and repeat processing in frame units during variable speed reproduction, outputs a subband signal switching command signal, and divides an encoded signal into bit allocation information, scale factor information, and quantized sample information. . The inverse quantizer 13 reproduces a subband signal from the bit allocation information, scale factor information and quantized sample information. The sub-band signal switcher 14 uses the sub-band signals of two discontinuous frames when the sub-band signal switching command signal is output, and sequentially sets different sub-bands for each sub-band sample group in one frame as boundaries. The playback subband signal is output after switching. The subband synthesis filter 15 band-synthesizes the reproduced subband signals and outputs an audio signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coded signal reproducing apparatus for reproducing a digitally recorded signal by sub-band coding or orthogonal transform coding, and particularly, a reproduced sound for reproducing at a speed different from that at the time of recording. Is related to the improvement of sound quality.

[0002]

2. Description of the Related Art In recent years, a coded signal recording device and a coded signal reproducing device for digitally recording voice using subband coding or orthogonal transform coding have been actively developed. In particular, recently, a device for recording or reproducing an audio signal accompanying an image together with the image signal has been actively developed. In such a device, it is desired to produce a sound that is easy to hear when fast-forwarding reproduction or slow reproduction is performed during reproduction. There is a request.

Generally, the pitch (pitch) of the reproduced sound rises in fast-forward reproduction, and the pitch (pitch) in slow reproduction.
However, the encoded signal reproduction device can reproduce the sound at a normal pitch (pitch) by thinning out the data of the reproduced sound or performing the interpolation processing to reproduce the sound. Is also being developed. A conventional coded signal reproducing device having such a function will be described below.

FIG. 8 is a block diagram showing the configuration of a conventional coded signal reproducing apparatus in subband coding when the number of subbands is M (a positive integer). The value of M is 32 or the like, for example. In FIG. 8, 81 is a synchronous extractor, 82 is a demultiplexer, 83 is an inverse quantizer, and 84.
Is a subband synthesis filter. The operation of the coded signal reproducing device configured as described above will be described below with reference to FIG.

The sync extractor 81 receives the encoded signal S801, detects a sync pattern, and outputs a sync signal S802. The demultiplexer 82 uses the encoded signal S801.
And a sync signal S802 output by the sync extractor 81 and a speed signal S803 representing the information of the reproduction speed are input. When the reproduction speed indicated by the speed signal S803 is faster than the standard reproduction speed, thinning processing is performed in frame units, If the reproduction speed indicated by the reproduction speed signal S803 is slower than the standard reproduction speed, repeat processing in frame units is performed, and then demultiplexing is performed, bit allocation information S804, scale factor information S805, and quantization sample information S806. And output.

The inverse quantizer 83 has the bit allocation information S80.
4, the scale factor information S805 and the quantized sample information S806 are input, and dequantized to obtain the subband signal S.
807 (1) to S807 (M) are output. The sub-band synthesis filter 84 includes the sub-band signals S807 (1) to S807 (1).
S807 (M) is input, band synthesis is performed, and a reproduced audio signal S808 is output.

By such an operation, the conventional coded signal reproducing apparatus reproduces the coded signal subjected to the thinning processing or the repeat processing during the variable speed reproduction, so that the pitch of the reproduced sound is the same as the pitch during the normal speed reproduction. Can be

[0008]

However, in the structure of the above-described conventional example, since the originally discontinuous frames are continuously reproduced at the time of variable speed reproduction, abnormal noise is generated at the joints of discontinuous frames. I had a problem. This problem will be described in detail with reference to FIG. FIG. 9 schematically shows the problems of the above-mentioned conventional example by taking the case of double speed reproduction as an example. In FIG. 9, S901 is a coded signal read at double speed, which is the same as the coded signal S801 in FIG. S902 is the demultiplexer 82 of FIG.
It is an encoded signal after the thinning processing is performed by.
The encoded signal S902 has a normal reproduction speed. The waveform W91 shows the waveform of the reproduced signal when the information contained in the encoded signal S901 is reproduced as it is, and the waveform W92 shows the waveform of the reproduced signal when the information contained in the encoded signal S902 is reproduced.

Since the encoded signal S902 is a combination of frames that are not originally continuous, the reproduced sound is discontinuous at the frame boundaries as shown by the waveform W92, and an abnormal sound is generated in the reproduced sound. . The present invention solves the above-mentioned conventional problems, and an object of the present invention is to provide a coded signal reproducing apparatus which does not change the pitch of reproduced sound and makes abnormal noise less noticeable when variable speed reproduction is performed. .

[0010]

In order to achieve this object, a coded signal reproducing device according to the invention of claim 1 obtains an audio signal by band-dividing it into M (positive integer) subbands. A sub-band signal to be encoded is encoded from a coded signal obtained by encoding K (positive integer) sub-band sample groups as one frame with a sub-band sample group consisting of signals of all data of all sub-bands as a basic unit. A signal is reproduced, a synchronization extraction means for inputting a coded signal, extracting a synchronization pattern and outputting a synchronization signal, a speed signal indicating reproduction speed information, a synchronization signal, and an encoded signal are input, When the playback speed indicated by the speed signal is faster than the speed during recording, the coded signal is thinned out in frame units, and when thinning out, a subband signal switching command signal is output. A demultiplexer that outputs the bit allocation information indicating the bit allocation of the quantized sample and the scale factor information indicating the level of the subband signal and the quantized sample information by demultiplexing the encoded signal other than the decimated one, Dequantization means for inputting bit allocation information, scale factor information, and quantization sample information, dequantizing and outputting a subband signal, and a subband signal are input, and a subband signal switching command signal is output. When there is no sub-band signal, the sub-band signal is output as it is as a reproduction sub-band signal, and when the sub-band signal switching command signal is output, the sub-band signal of the frame immediately before the thinned-out frame is held as the first sub-band signal, and The frame immediately after the decimated frame that is continuously output from the quantizer Input sub-band signal as the second sub-band signal, and outputs the first sub-band signal as the sub-band sample group number increases from 1st to Kth in one frame. Output the reproduced subband signal by switching between the first subband signal and the second subband signal so that the number gradually decreases and the number of subbands that output the second subband signal gradually increases. It is provided with a sub-band signal switching means and a sub-band synthesis filter for inputting a reproduction sub-band signal, band-combining it, and outputting a reproduction audio signal.

According to another aspect of the coded signal reproducing apparatus of the present invention, a subband signal obtained by band-dividing an audio signal into M (positive integer) subbands is divided into data of all subbands. An audio signal is reproduced from a coded signal obtained by coding K (positive integer) number of subband sample groups as one frame with a subband sample group consisting of signals as a basic unit. When a sync extraction means for extracting a sync pattern and outputting a sync signal, and a speed signal indicating the speed information for reproduction, a sync signal and an encoded signal are input, and the reproduction speed indicated by the speed signal is faster than the speed at the time of recording A subband signal switching command signal is output to, and the coded signal is demultiplexed to display bit allocation information indicating the bit allocation of the quantized sample and the level of the subband signal. A demultiplexer that outputs scale factor information and quantized sample information, an inverse quantizer that inputs bit allocation information, scale factor information, and quantized sample information, and inversely quantizes and outputs a subband signal, and a subband When a signal is input and the sub-band signal switching command signal is not output, the sub-band signal is output as it is as a reproduction sub-band signal, and when the sub-band signal switching command signal is output, the continuous two frames The sub-band signal of the previous frame is held as the first sub-band signal, and the sub-band signal of the latter frame of the two consecutive frames continuously output from the dequantization means is used as the second sub-band signal. The number of subband sample groups from the 1st to the Kth is large within one frame. As it becomes first
The first subband signal and the second subband signal so that the number of subbands that output the subband signal is gradually reduced and the number of subbands that output the second subband signal is gradually increased. And a sub-band signal switching unit for switching the signal to output a reproduction sub-band signal, and a sub-band synthesis filter for inputting the reproduction sub-band signal and performing band combination to output a reproduction audio signal.

According to the third aspect of the present invention, there is provided a coded signal reproducing apparatus, wherein a subband signal obtained by band-dividing an audio signal into M (positive integer) subbands is divided into data of all subbands. An audio signal is reproduced from a coded signal obtained by coding K (positive integer) number of subband sample groups as one frame with a subband sample group consisting of signals as a basic unit. When a sync extraction means for extracting a sync pattern and outputting a sync signal, and a speed signal indicating the speed information for reproduction, a sync signal and an encoded signal are input, and the reproduction speed indicated by the speed signal is faster than the speed at the time of recording Decimation processing is performed for each frame on the encoded signal, a subband signal switching command signal is output when decimation is performed, and the encoded signals other than the decimation are deselected. Demultiplexer that outputs bit allocation information indicating the number of bit allocation of quantized samples by multiplexing, scale factor information indicating the level of subband signal, and quantization sample information, bit allocation information, scale factor information, and quantization sample information And type
Dequantization means for dequantizing and outputting a subband signal, and a subband signal of a subband having a frequency lower than the N-th band (N is an integer 1 <N <M) are input, and a subband signal switching command is input. When the signal is not output, the subband signal is output as the reproduction subband signal as it is, and when the subband signal switching command signal is output, the subband signal of the frame immediately before the thinned frame is the first subband signal. The subband signal of the frame immediately after the decimated frame continuously output from the dequantization means is input as the second subband signal, and the subband samples from the 1st to the Kth subframes are included in one frame. As the group number increases, the number of subbands that output the first subband signal gradually decreases and the number of subbands decreases. Subband signal switching means for switching between the first subband signal and the second subband signal so as to gradually increase the number of subbands for outputting signals, and outputting a reproduction subband signal on the low frequency side; When a subband signal of a subband having a frequency higher than the N band is input and the subband signal switching command signal is not output, the subband signal is directly output as a reproduction subband signal, and the subband signal switching command signal is output. In the frame in which the reproduction subband signal on the low frequency side is output by switching the first subband signal and the second subband signal in the subband having a frequency lower than the Nth band, the reproduction subband on the high frequency side is output. Mute means that outputs a silent signal as a band signal, and a playback subband signal on the low frequency side and high frequency side And force, and a sub-band synthesis filter and band synthesis to output a reproduced audio signal.

According to another aspect of the coded signal reproducing apparatus of the present invention, a subband signal obtained by band-dividing an audio signal into M (positive integer) subbands is divided into data of all subbands. An audio signal is reproduced from a coded signal obtained by coding K (positive integer) number of subband sample groups as one frame with a subband sample group consisting of signals as a basic unit. When a sync extraction means for extracting a sync pattern and outputting a sync signal and a speed signal indicating the speed information for reproduction, a sync signal and an encoded signal are input, and the reproduction speed indicated by the speed signal is slower than the speed at the time of recording Repeats the coded signal in units of frames, outputs the subband signal switching command signal when the repeat processing is performed, and outputs the coded signal after the repeat processing. Demultiplexer that outputs bit allocation information indicating the bit allocation number of multiplexed quantized samples, scale factor information indicating subband signal level, and quantized sample information, bit allocation information, scale factor information, and quantized samples Dequantizing means for inputting information, dequantizing and outputting a subband signal, memory means for inputting a reproduction subband signal and holding for a certain period, and inputting a subband signal, and issuing a subband signal switching command When the signal is not output, the subband signal is output as it is as the reproduction subband signal,
When the subband signal switching command signal is output, the subband signal of the preceding frame of the two repeated frames is held as the first subband signal, and the repeated 2 stored in the memory means. The subband signal of the immediately preceding frame before the repeat processing of the subband signal of the preceding frame of the two frames is input as the second subband signal, and the first subframe signal within one frame is input.
As the number of the sub-band sample groups from the # 1 to the #K increases, the number of sub-bands that output the first sub-band signal gradually decreases and the number of sub-bands that output the second sub-band signal gradually increases. Sub-band signal switching means for switching the first sub-band signal and the second sub-band signal so as to increase and output a reproduction sub-band signal,
And a sub-band synthesis filter for inputting a reproduction sub-band signal, band-combining it, and outputting a reproduction audio signal.

According to another aspect of the coded signal reproducing apparatus of the present invention, a subband signal obtained by band-dividing an audio signal into M (positive integer) subbands is obtained from a signal for each data of all subbands. The basic unit is the sub-band sample group, and K (positive integer) number of sub-band sample groups is 1
Synchronous extraction means for reproducing an audio signal from an encoded signal encoded as a frame, inputting the encoded signal, extracting a synchronization pattern and outputting a synchronization signal, and synchronizing with a speed signal indicating reproduction speed information. A signal and an encoded signal are input, and when the reproduction speed indicated by the speed signal is faster than the recording speed, thinning processing is performed on the encoded signal in frame units, and when the reproduction speed is slow, the encoded signal is processed in frame units. Repeat processing, output a subband signal switching command signal when either thinning or repeating is performed, and demultiplex the coded signal after performing either thinning or repeat to quantize samples. Bitmapping information that indicates the number of bit allocations, scale factor information that indicates the level of the subband signal, and a hoax that outputs quantized sample information Type a mux, the bit allocation information and scale factor information and the quantization sample information,
Dequantizing means for dequantizing and outputting a subband signal, memory means for inputting a reproduced subband signal and holding it for a certain period, and inputting a subband signal and not outputting a subband signal switching command signal Sometimes the sub-band signal is output as it is as a reproduction sub-band signal, a sub-band signal switching command signal is output, and when the reproduction speed is faster than the recording speed, the sub-band signal of the frame immediately before the thinned frame is used as the first sub-signal. The sub-band signal of the frame immediately after the thinned frame which is held as a band signal and is continuously output from the dequantization means
The subband signal of the preceding frame of the two frames repeated when the reproduction speed is slower than the recording speed when the subband signal switching command signal is output as the first subband signal. The subband signal of the immediately preceding frame which is held as a signal and which is stored in the memory means before the repeat processing of the subband signal of the preceding frame of the two repeated frames is input as the second subband signal. As the number of the subband sample group increases from the 1st to the Kth subframe in one frame, the number of subbands that output the first subband signal gradually decreases and outputs the second subband signal. The reproduction subband signal is output by switching between the first subband signal and the second subband signal so that the number of subbands to be reproduced gradually increases. Includes a sub-band signal switching means inputs the reproduced sub-band signals, and a sub-band synthesis filter for outputting a reproduced audio signal band synthesis.

[0015]

According to the encoded signal reproducing apparatus of the present invention,
With the above configuration, when the demultiplexer outputs the subband signal switching command signal, the subband signal switching means increases the number of the subband sample group from the 1st to the Kth in one frame. As the number of subbands that output the first subband signal gradually decreases and the number of subbands that output the second subband signal gradually increases, the number of subbands is switched stepwise for each subband sample group. A playback subband signal is output, and the playback subband signal is band-synthesized by a subband synthesis filter to output a playback audio signal.

According to the coded signal reproducing apparatus of the present invention as defined in claim 2, the demultiplexer outputs the subband signal switching command signal according to the above configuration.
The subband signal switching means gradually decreases the number of subbands that output the first subband signal as the number of the subband sample group increases from 1st to Kth in one frame, and the second subband signal is output. The sub-band signal is output by switching in stages for each sub-band sample group so that the number of sub-band signals that are output gradually increases, and by outputting the playback sub-band signal by switching in stages for each sub-band sample group. A sub-band synthesis filter band-synthesizes the reproduced sub-band signal and outputs a reproduced audio signal.

According to the encoded signal reproducing apparatus of the present invention as defined in claim 3, the demultiplexer outputs the subband signal switching command signal by the above configuration,
The subband signal switching means gradually decreases the number of subbands that output the first subband signal as the number of the subband sample group increases from 1st to Kth in one frame, and the second subband signal is output. Of the sub-band signal is output stepwise for each sub-band sample group so that the number of sub-bands is gradually increased, and the reproduction sub-band signal on the low frequency side is output. The sub-band signal of the band is converted into a silent signal and output as a reproduction sub-band signal on the high frequency side, and both reproduction sub-band signals are band-combined by a sub-band synthesis filter to output a reproduction audio signal.

According to the coded signal reproducing apparatus of the invention described in claim 4, the demultiplexer outputs the subband signal switching command signal by the above-mentioned configuration,
The subband signal switching means gradually decreases the number of subbands that output the first subband signal as the number of the subband sample group increases from 1st to Kth in one frame, and the second subband signal is output. The sub-band synthesis filter outputs the reproduced sub-band signal by switching in stages for each sub-band sample group so that the number of sub-bands that output the sub-band signal gradually increases. And outputs the reproduced audio signal.

Further, the coded signal reproducing apparatus according to the invention of claim 5 has the functions of the inventions of claims 1 and 4 by the above-mentioned configuration.

[0020]

Embodiments of the present invention will be described below with reference to the drawings. Here, for example, the case of subband coding in units of frames in which the number of subbands is M (positive integer) and the number of subband sample groups is M (positive integer) will be described. The value of M is usually 3
2, etc., the subband signal switching means of the encoded signal reproducing device of the present invention performs processing by grouping M subbands into five subband groups. Also,
The value of K is 8, for example. One sub-band sample group is a sub-band signal obtained by band-dividing an audio signal into M sub-bands, and is composed of signals for each data of all sub-bands. A coded signal is created by the conversion. It is assumed that the subbands and subband groups have lower frequencies as the numbers are smaller.

[First Embodiment: Corresponding to Claim 1] FIG. 1 is a block diagram of a coded signal reproducing apparatus according to a first embodiment of the present invention. In FIG. 1, 11 is a synchronous extractor, 12 is a demultiplexer, 13 is an inverse quantizer, 14 is a subband signal switch, and 15 is a subband synthesis filter.

The operation of the coded signal reproducing apparatus of the present embodiment having the above-described structure will be described below by taking the case of reproducing at double speed as an example. Synchronous extractor 1
1 receives the encoded signal S101, detects the synchronization pattern, and outputs the synchronization signal S102. The demultiplexer 12 inputs the encoded signal S101, the synchronization signal S102, and the speed signal S103 indicating the information of the reproduction speed, and when the reproduction speed indicated by the speed signal S103 is faster than the standard reproduction speed, the demultiplexer 12 outputs the synchronization signal S102 as time. As a reference, thinning processing is performed in frame units and then demultiplexed to output bit allocation information S105, scale factor information S106, and quantized sample information S107, and at the same time output a subband signal switching command signal S104. Here, it is assumed that the subband signal switching command signal S104 indicates that the thinning processing is not performed when the subband signal is “L” and the thinning processing is performed when the subband signal switching command signal S104 is “H”. The subband signal switching command signal S104 described above is "H" in the frames immediately before and after the frame in which the thinning process is performed, and is "L" in the other frames.

The demultiplexer 12 has a speed signal S10.
When the reproduction speed indicated by 3 is equal to the standard reproduction speed, the input coded signal S101 is demultiplexed as it is, and bit allocation information S105 and scale factor information S
106, and quantized sample information S107 is output. However, the rate of decimation by the demultiplexer 12 at the time of high-speed reproduction is different from the decimation rate of the conventional coded signal reproduction apparatus in consideration of the decrease in the transfer rate by the subband signal switcher 14. For example, during double speed playback,
The demultiplexer 82 of the conventional coded signal reproducing device has two
Although one of the frames was thinned out, the demultiplexer 12 of this embodiment thins out one of the four frames.

The inverse quantizer 13 has the bit allocation information S10.
5, the scale factor information S106, and the quantized sample information S107 are input and inversely quantized to obtain the subband signal S.
108 (1) to S108 (M) are output. The sub-band signal switcher 14 uses the sub-band signal S108 (1).
~ S108 (M) and subband signal switching command signal S
104 and the subband signal switching command signal S
When 104 is “L” (a frame other than the frames immediately before and after the thinned frame, in FIG.
The input subband signal S in the D, H, and L frames)
After holding 108 (1) to S108 (M) for a certain period of time, the subband signals S109 (1) to S109 are kept as they are.
Output as (M). The operation of holding for a certain period of time and then outputting will be described in detail below with reference to FIG.

Further, the subband signal switch 14 is
When the subband signal switching command signal S104 is "H", the input subband signals S108 (1) to S1
08 (M) is held for a certain period of time, and the signal and the inverse quantizer 1
3 output sub-band signal S108 of the next frame
Subband signals S109 (1) to S109 (M) are output by sequentially switching (1) to S108 (M) with different subbands as boundaries for each subband sample group. More specifically, the subband signals S108 (1) to S108 of the frame immediately before the thinned frame are described.
(M) is held as the first subband signal, and the subband signals S108 (1) to S10 of the frame immediately after the decimated frame continuously output from the inverse quantizer 13 are held.
8 (M) is input as the second subband signal, and the number of subbands that output the first subband signal as the number of the subband sample groups from the first to the eighth increases in one frame Of the subband signals S109 (1) to S109 (1) to switch the first subband signal and the second subband signal so that the number of subbands that output the second subband signal gradually increases S109 (M) is output.

The sub-band synthesizing filter 15 inputs the sub-band signals S109 (1) to S109 (M), band-synthesizes them, and outputs a reproduced audio signal S110. Here, in this embodiment, the process of changing the reproduction signal and the subband switching performed by the subband signal switch 14 will be described with reference to FIG. In FIG.
S201 is a coded signal read at double speed, which is the same signal as S101 in FIG. The encoded signal S201 is thinned out by one frame every four frames by the demultiplexer 12, and each frame is expanded by 4/3 times on the time axis to become an intermediate encoded signal S202. The intermediate coded signal S202 is a signal temporarily generated in the demultiplexer. The subband signal switching command signal S1
04 is A, C, E, G, I, K of the signal S202 of FIG.
It becomes "H" during the period of, and becomes "L" during the periods of D, H, and L. In FIG. 2, out of the 12 frames A to L of the encoded signal S201, B, F, and J frames are thinned out to obtain A,
Nine frames C, D, E, G, H, I, K, and L become the intermediate coded signal S202. Intermediate coded signal S202
Is demultiplexed and inversely quantized into a subband signal. The subband signal switcher 14 sequentially changes the subframe signals S108 (1) to S108 (M) for two frames derived from the originally discontinuous frame A and frame C for each subband sample group as described above. The sub-band is used as a boundary and is switched and multiplied by 3/2 in the time axis direction to obtain the encoded signal S203 [sub-band signal S109
(1) to S109 (M)] are output. For frame D, the coded signal S203 [subband signals S109 (1) to S10 is simply multiplied by 3/2 in the time axis direction].
9 (M)]. Similar processing is performed for the frames E to L. By the above processing, the coded signal S101 is finally thinned to 1/2 and reproduced.
In the intermediate coded signal S202, A, C, E,
The subband signal switching command signal S104 becomes "H" in the G, I, K frames, and the subband signal switching command signal S104 becomes "L" in the D, H, L frames.

The above-mentioned operation of holding for a certain period of time and then outputting will be described. For example, D in the intermediate coded signal S202 and the coded signal S203,
Since the H and L frames have different time lengths, the data of the D, H, and L frames input at the time length of the intermediate coded signal S202 is held and the time axis is set to 3/2. This means that the double stretched encoded signal S203 is output from the time point of one frame delay of the intermediate encoded signal S202, and this operation is output after being held for a certain period of time. The A, E, and I frames are output from the time point of 1.5 frames delay of the intermediate coded signal S202, and the C, G, and K frames are output from the time point of 0.5 frame delay of the intermediate coded signal S202. Is output from.

The waveform W21 represents the waveform of the reproduced audio data when the encoded signal S201 is reproduced as it is. The waveform W22 represents the waveform of the reproduced audio signal when the encoded signal S203 is reproduced. In the waveform W22, since the reproduction signals of the respective frames are smoothly connected, the abnormal noise at the joint between the frames is inconspicuous. In this embodiment, one out of every two frames expands the read coded signal along the time axis and reproduces it as it is, so that the deterioration of the sound quality can be reduced.

The switching of the sub-band signals is performed, for example, as shown in the frames A and C of FIG. 2, the first sub-band sample group is all the data of the frame A, and the second sub-band sample group is the first sub-band group. Only the data of frame C is the data of the other subbands, the data of frame A,
Hereinafter, the ratio of the data of the frame C is sequentially increased, and all the final samples are the data of the frame C. In addition,
Here, an example is shown in which the ratio of the data of the frame C is increased from the first subband group side, but conversely, the ratio of the data of the frame C may be increased from the fifth subband group side.

[Second Embodiment: Corresponding to Claim 2] A second embodiment of the present invention will be described. The configuration of the coded signal reproducing device of the second embodiment of the present invention is the same as that of the first embodiment shown in FIG. 1, but the operations of the demultiplexer and the subband signal switcher are different. FIG. 3 shows the changing process of the reproduction signal and the subband switching performed by the subband signal switching unit 14 in this embodiment. The operation of the coded signal reproducing apparatus of this embodiment will be described below with reference to FIGS. 1 and 3. Here, a case where the encoded signal is reproduced at double speed will be described.

The sync extractor 11 receives the coded signal S101, detects a sync pattern, and outputs a sync signal S102. The demultiplexer 12 receives the encoded signal S101.
, A synchronization signal S102, and a speed signal S representing information on the reproduction speed
103 and the reproduction speed indicated by the speed signal S103 is twice the standard reproduction speed, demultiplexing is performed as it is, and bit allocation information S105, scale factor information S106, and quantized sample information S107 are output. , The subband signal switching command signal S104 is "H"
To output. In this embodiment, the reproduction speed is the recording speed 2
When it is double, the subband signal switching command signal S
104 is always "H".

The inverse quantizer 13 has the bit allocation information S10.
5, the scale factor information S106, and the quantized sample information S107 are input and inversely quantized to obtain the subband signal S.
108 (1) to S108 (M) are output. When the subband signal switching command signal S104 is "H", the input subband signals S108 (1) to S108 are input.
(M) is held for a certain time, and the subband signal S108 of the next frame output from the signal and the inverse quantizer 13 is held.
Subband signals S109 (1) to S109 (M) are output by sequentially switching (1) to S108 (M) for each subband sample group using different subbands as boundaries.
Specifically, the sub-band signals S108 (1) to S108 (M) of the previous frame of the two consecutive frames are held as the first sub-band signal and continuously output from the dequantization means. Sub-band signals S108 (1) to S108 of the subsequent frame of the two selected frames.
(M) is input as the second subband signal, and the number of subbands that output the first subband signal increases as the number of the subband sample groups from the first to the eighth increases in one frame. Subband signals S109 (1) to S109 are switched by switching between the first subband signal and the second subband signal so that the number of subbands that gradually decrease and output the second subband signal gradually increase. (M) is output.

The sub-band synthesizing filter 15 inputs the sub-band signals S109 (1) to S109 (M), band-synthesizes them, and outputs a reproduced audio signal S110. Here, in this embodiment, the process of changing the reproduction signal and the subband switching performed by the subband signal switch 14 will be described with reference to FIG. In FIG.
S301 is a coded signal read at double speed, which is the same signal as the coded signal S101 in FIG. The encoded signal S301 is demultiplexed by the demultiplexer 12 and inversely quantized by the inverse quantizer to become subband signals S108 (1) to S108 (M). The subband signal switcher 14 includes a frame A and a frame B.
2 frames of subband signal S108 derived from
As described above, (1) to S108 (M) are sequentially switched for different subbands for each subband sample group as boundaries, and doubled in the time axis direction to obtain the encoded signal S30.
2 [subband signals S109 (1) to S109 (M)]
Is output. Frame C, Frame D and Frame E
The same process is performed on the frame F and the frame F.

The coded signal S10 is obtained by the above processing.
1 is finally thinned out to 1/2 and reproduced. Waveform W3
Reference numeral 1 represents a waveform of reproduced audio data when the encoded signal S301 is reproduced as it is. Waveform W32
Represents a waveform of a reproduced audio signal when the encoded signal S302 is reproduced. In the waveform W32, since the reproduction signals of the respective frames are smoothly connected, the abnormal noise at the joint between the frames is inconspicuous. In this embodiment, since it is not necessary for the demultiplexer to perform the thinning-out process at the time of double speed reproduction, the demultiplexer can have a simple structure.

For switching the sub-band signals, for example, as shown in FIG. 3, the first sub-band sample group is all data of frame A, and the second sub-band sample group is only the first sub-band group of frame B data. The data of the sub-band group is the data of the frame A, and the ratio of the data of the frame B is sequentially increased. The final sample group is the data of the frame B. Although the example in which the ratio of the data of the frame B is increased in order from the first subband group side is shown here, the ratio of the data of the frame B may be increased in order from the fifth subband group side.

[Third Embodiment: Corresponding to Claim 3] FIG. 4 is a block diagram of an encoded signal reproducing apparatus according to a third embodiment of the present invention. In FIG. 4, 41 is a synchronous extractor, 42 is a demultiplexer, 43 is a dequantizer, 44 is a subband signal switcher, 45 is a mute device, and 46 is a subband synthesis filter.

The operation of the coded signal reproducing apparatus of this embodiment having the above-described structure will be described below by taking the case of reproducing at double speed as an example. Synchronous extractor 4
1 receives the encoded signal S401, detects the synchronization pattern, and outputs the synchronization signal S402. The demultiplexer 42 inputs the encoded signal S401, the synchronization signal S402, and the speed signal S403 indicating the information of the reproduction speed, and when the reproduction speed indicated by the speed signal S403 is faster than the standard reproduction speed, the decimation processing in frame units is performed. Then, demultiplexing is performed and bit allocation information S405, scale factor information S406, and quantized sample information S407 are output, and at the same time, a subband signal switching command signal S404 is output.

Here, it is assumed that the subband signal switching command signal S404 indicates that thinning processing is not performed when it is "L" and thinning processing is performed when it is "H". The demultiplexer 42 has a speed signal S40.
If the reproduction speed indicated by 3 is equal to the standard reproduction speed, the input encoded signal S401 is demultiplexed as it is, and bit allocation information S405, scale factor information S
406, quantized sample information S407 is output.

However, at the time of high speed reproduction, the demultiplexer 4
2 is the subband signal switching unit 4
In consideration of the decrease in the transfer rate due to No. 4, it is different from the thinning rate of the conventional coded signal reproducing apparatus. For example, 2
At the time of double speed reproduction, the demultiplexer 82 of the conventional coded signal reproducing apparatus thinned out one frame out of two frames, but the demultiplexer 42 of this embodiment thinned out one frame out of four frames.

The inverse quantizer 43 has the bit allocation information S40.
5, the scale factor information S406, and the quantized sample information S407 are input and inversely quantized to obtain the subband signal S.
408 (1) to S408 (M) are output. The subband signal switch 44 uses the subband signal S408 (1).
To S408 (I-1) and the subband signal switching command signal S404 are input, and when the subband signal switching command signal S404 is "L", the input subband signals S408 (1) to S408 (I-1). ) Is held for a certain period of time, and then the encoded signals S409 (1) to S409
Output as (I-1). When the subband signal switching command signal S404 is "H", the input subband signals S408 (1) to S408 (I-1) are held for a certain period of time and output from the inverse quantizer 43 together with the signal. Next frame encoded signal S408 (1) to S408
Similarly to the first embodiment, (I-1) is switched with subbands that sequentially differ for each subband sample group as boundaries, and subband signals S409 (1) to S409 (I-
1) is output.

The mute device 45 operates the subband signal S40.
8 (I) to S408 (M) and the subband signal switching command signal S404 are input, and when the subband signal switching command signal S404 is "L", the input subband signals S408 (I) to S408 (M). ) Is held for a certain period of time, and then the subband signals S409 (I) to S4
It outputs as 09 (M). Further, when the subband signal switching command signal S404 is "H", the silent signal is changed to the subband signals S409 (I) to S4 regardless of the input subband signals S408 (I) to S408 (M).
It outputs as 09 (M).

The subband synthesizing filter 46 inputs the subband signals S409 (1) to S409 (M), synthesizes the bands, and outputs the reproduced audio signal S410. Here, in this embodiment, the process of changing the reproduction signal and the subband switching performed by the subband signal switch 44 will be described with reference to FIG. In FIG.
S501 is a coded signal read at double speed, which is the same signal as the coded signal S401 in FIG. In S501, one frame is thinned out every four frames by the demultiplexer 42, and each frame is expanded 4/3 times on the time axis to become an intermediate coded signal S502. The intermediate coded signal S502 is a signal temporarily generated in the demultiplexer. In FIG. 5, 1 of A to L of the encoded signal S501
Of the two frames, the B, F, and J frames are thinned out, and the nine frames A, C, D, E, G, H, I, K, and L become the intermediate coded signal S502. The subband switching command signal S404 and the intermediate coded signal S502
The relationship with is the same as in FIG.

The intermediate coded signal S502 is demultiplexed and dequantized to generate subband signals S408 (1) -S408 (1).
It becomes S408 (M). Subband signal switch 44
Are sub-band signals S408 (1) to S4 for two frames derived from frame A and frame C which are originally discontinuous.
08 (I-1) is sequentially switched for each subband sample group using different subbands as boundaries, and 3 /
Double the encoded signal S503 [subband signal S409
(1) to S409 (I-1)] are output. For frame D, the encoded signal S503 [subband signals S409 (1) to S409] is simply multiplied by 3/2 in the time axis direction.
409 (I-1)].

The mute device 45 converts the two frames of subband signals S408 (I) to S408 (M) derived from the originally discontinuous frame A and frame C into a silent signal and encodes the encoded signal S503 [subband]. Signal S409
(I) to S409 (M)] are output. For frame D, the encoded signal S503 [subband signals S409 (I) to S40] is simply multiplied by 3/2 in the time axis direction.
9 (M)].

Similar processing is performed for the frames E to L. By the above processing, the encoded signal S
Finally, 501 is thinned to 1/2 and reproduced. A waveform W51 represents a waveform of reproduced audio data when the encoded signal S501 is reproduced as it is. Waveform W
Reference numeral 52 represents a waveform of a reproduced audio signal when the encoded signal S503 is reproduced. In the waveform W52, since the reproduction signals of the respective frames are smoothly connected, the abnormal noise at the joint between the frames is inconspicuous. In this embodiment, when the mute device 45 switches and outputs originally discontinuous frames, the high-frequency signal is made into a silent signal, so that high-frequency noise that is easily noticeable when reproducing discontinuous frames is generated. Can be suppressed.

In this example, the low band subbands from the first subband to the I-1th subband correspond to the first subband group to the third subband group, and the Ith subband to the Mth subband. It is assumed that the high band subbands of the subbands correspond to the fourth to fifth subband groups. The switching of the subband signals is performed, for example, as shown in the frames A and C of FIG. 5, the first subband sample group is all data of frame A, and the second subband sample group is only the first subband group is frame C. The data of the other sub-band group is the data of the frame A, and the ratio of the data of the frame C is sequentially increased, and all the final samples are the data of the frame C. Although the example in which the ratio of the data of the frame C is increased in order from the first subband group side is shown here, conversely, the ratio of the data of the frame C may be increased in order from the fifth subband group side.

[Fourth Embodiment: Corresponding to Claim 4] FIG. 6 is a block diagram of an encoded signal reproducing apparatus according to a fourth embodiment of the present invention. In FIG. 6, 61 is a synchronous extractor, 62 is a demultiplexer, 63 is an inverse quantizer, 64 is a subband signal switch, 65 is a subband synthesis filter, and 66 is a memory.

The operation of the coded signal reproducing apparatus of this embodiment having the above-described structure will be described below by taking the case of reproducing at 1/2 speed as an example. The sync extractor 61 receives the encoded signal S601, detects a sync pattern, and outputs a sync signal S602. The demultiplexer 62 inputs the encoded signal S601, the synchronization signal S602, and the speed signal S603 representing the information on the reproduction speed, and when the reproduction speed indicated by the speed signal S603 is slower than the standard reproduction speed, the repeat processing in frame units. Then, demultiplexing is performed and bit allocation information S605, scale factor information S606, and quantized sample information S607 are output, and at the same time, a subband signal switching command signal S60 is output.
4 is output. Here, it is assumed that the subband signal switching command signal S604 indicates that the repeat processing is not performed when it is "L", and that the repeat processing is performed when it is "H".

The demultiplexer 62 has a speed signal S60.
When the reproduction speed indicated by 3 is equal to the standard reproduction speed, the input coded signal S601 is demultiplexed as it is, and bit allocation information S605 and scale factor information S
606, Quantized sample information S607 is output. The dequantizer 63 receives the bit allocation information S605, the scale factor information S606, and the quantized sample information S607, dequantizes the subband signals S608 (1) to S608 (S).
608 (M) is output.

The subband signal switch 64 receives the subband signals S608 (1) to S608 (M), the subband signal switching command signal S604, and the subband signals S611 (1) to S611 (M) output by the memory 66. And the subband signal switching command signal S604 is "L", the subband signals S608 (1) to S6 are input.
After holding 08 (M) for a certain period of time, the sub band signals S609 (1) to S609 (M) are output as they are.

Further, the subband signal switching command signal S
When 604 is “H”, the input subband signal S
608 (1) to S608 (M) and the subband signals S611 (1) to S611 (M) output from the memory 66,
Subband signals S609 (1) to S609 (S) to S609
609 (M) is output. Specifically, the subband signals S608 (1) to S608 (M) of the preceding frame of the two repeated frames are held as the first subband signal, and the repeat stored in the memory 66 is stored. The sub-band signals S611 (1) to S611 (M) of the immediately preceding frame before the repeat processing of the sub-band signal of the front frame of the two frames are input as the second sub-band signal, and within one frame. The number of subbands that output the first subband signal gradually decreases as the number of the subband sample groups from the first to the eighth increases, and the number of subbands that output the second subband signal To switch between the first subband signal and the second subband signal so that
609 (1) to S609 (M) are output.

The memory 66 stores the subband signal S609.
Input (1) to S609 (M), hold for a certain period,
Output to the subband signal switch 64. The sub-band synthesis filter 65 has sub-band signals S609 (1)-
S609 (M) is input, band synthesis is performed, and a reproduced audio signal S610 is output. Here, in this embodiment, the process of changing the reproduction signal and the subband signal switch 6
Switching of subbands performed by No. 4 will be described with reference to FIG. 7. In FIG. 7, S701 is a coded signal read at 1/2 speed, which is the same signal as S601 in FIG. In step S701, the demultiplexer 62 performs a repeat process on a frame-by-frame basis, and each frame is compressed by 1/2 in the time axis, demultiplexed, and dequantized to generate a subband signal S702 [S608.
(1) to S608 (M)]. In FIG. 7, the subband signals reproduced from the frame A of the encoded signal S701 are repeats A1 and A2 of S702, and the same is B1, B2, C1 and C2. The subband signal switching command signal S604 and the intermediate coded signal S
The relationship of 702 is as follows. That is, the subband signal switching command signal S604 becomes “H” in the frames A1, B1, and C1 of the intermediate coded signal S702, and the subband signal switch 64 causes the intermediate coded signal S702.
Of that frame (for example, A1) and the frame immediately before that (for example, Z) are gradually switched to the encoded signal S703.
Is output. Since the frames Z and A2 were originally continuous, they are smoothly connected.

When the sub-band signal switch 64 receives the B1 sub-band signal, the memory 66 holds the A2 sub-band signal of the immediately preceding frame. At this time, the sub-band signal switching command signal S604.
Is "H", the subband signal switching means 64 holds the B1 subband signal and outputs the signal and the A2 subband signal input from the memory 66.
Subbands that are different for each subband sample group are sequentially switched and output with boundaries as boundaries. In the next frame, the subband signal switching command signal S604 is "L", and the subband signal switching unit 64 inputs the reproduced subband signal of B2 and outputs it as it is.

The waveform W71 represents the waveform of reproduced audio data when the encoded signal S701 is reproduced as it is. The waveform W72 represents the waveform of the reproduced audio signal when the encoded signal S703 is reproduced. In the waveform W72, since the reproduction signals of the respective frames are smoothly connected, the abnormal noise at the joint between the frames is inconspicuous.

For switching the sub-band signals, for example, as shown in FIG. 7, the first to fourth sub-band sample groups in the first half output B1 sub-band signals as they are, and the fifth sub-band sample group. Gradually from A
For example, the ratio of the reproduced subband signal of 2 is increased. In each of the above-described embodiments, the first and second sub-bands are sequentially demarcated for each sub-band sample group.
The sub-band signal of was switched, but not limited to this 1
As the number of the subband sample groups from the 1st to the Kth in the frame increases, the number of subbands that output the first subband signal gradually decreases and the second subband signal is output.
Any pattern may be used as long as the first subband signal and the second subband signal are switched so that the number of subbands that output the subband signal is gradually increased.

Further, although the above-mentioned embodiment is for performing repeat, a combination of this embodiment and the above-described thinning-out embodiment can also be mentioned as an embodiment.

[0057]

According to the coded signal reproducing apparatus of the invention described in claim 1, the coded signal switching means is the first in one frame.
As the number of the sub-band sample groups from the # 1 to the #K increases, the number of sub-bands that output the first sub-band signal gradually decreases and the number of sub-bands that output the second sub-band signal gradually increases. The playback subband signal is output by switching the subband sample groups step by step so that the playback sound is reproduced by band-synthesizing the playback subband signal with the subband synthesis filter and outputting the playback audio signal. When a discontinuous frame is generated by the thinning processing in frame units at the time of high-speed reproduction so as not to change the pitch of, the generation of abnormal noise at the boundary of the discontinuous frames can be suppressed.

In the coded signal reproducing apparatus according to the second aspect of the present invention, the coded signal switching means has the first sub-sample as the sub-band sample group number increases from 1st to Kth in one frame. The number of subbands that output the band signal gradually decreases and the number of subbands that outputs the second subband signal gradually increases, and the subband sample groups are output stepwise by switching. High-speed playback is achieved during high-speed playback by switching playback in stages for each band sample group, outputting a playback sub-band signal, and outputting a playback audio signal by band-synthesizing the playback sub-band signal with a sub-band synthesis filter. When performing, it is possible to prevent the pitch of the reproduced sound from changing without performing thinning processing in frame units, and it is possible to suppress the generation of abnormal noise.

In the coded signal reproducing apparatus according to the third aspect of the present invention, the coded signal switching means has the first sub-sample as the number of the sub-band sample group increases from the 1st to the Kth in one frame. The number of subbands that output the band signal gradually decreases, and the number of subbands that outputs the second subband signal gradually increases, so that the number of subband samples is changed stepwise for each subband sample group. The playback subband signal is output, and the mute means converts the subband signal of the high frequency subband into a silent signal and outputs it as the playback subband signal on the high frequency side, and the playback subband signal is output by the subband synthesis filter. By performing band synthesis and outputting the playback audio signal, discontinuity of frames is achieved by thinning processing in frame units during high-speed playback to prevent the pitch of the playback sound from changing. When the over-time has occurred, it is possible to suppress the generation of abnormal noise at the boundaries of discontinuous frames.
In particular, it is possible to suppress abnormal high-frequency noise that is easily noticeable.

In the coded signal reproducing device according to the fourth aspect of the present invention, the coded signal switching means makes the first sub as the number of the subband sample group increases from 1st to Kth in one frame. Playback sub-band signal that is switched stepwise for each sub-band sample group so that the number of sub-bands that output band signals gradually decreases and the number of sub-bands that outputs second sub-band signals gradually increases Is output, and the reproduced subband signals are band-combined by the subband synthesis filter to output the reproduced audio signal, which is discontinuous due to the repeat processing in frame units during low-speed reproduction so as not to change the pitch of the reproduced sound. When a frame occurs, it is possible to suppress the generation of abnormal noise at the boundary of discontinuous frames.

The coded signal reproducing apparatus according to the invention of claim 5 has the effects of the coded signal reproducing apparatus of the inventions of claims 1 and 4.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a coded signal reproducing device according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram showing changes in the reproduced signal of the encoded signal reproducing device in the first embodiment of the present invention.

FIG. 3 is a schematic diagram showing changes in a reproduced signal of the encoded signal reproducing device in the second embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of an encoded signal reproducing device according to a third embodiment of the present invention.

FIG. 5 is a schematic diagram showing changes in the reproduced signal of the coded signal reproducing device in the third embodiment of the present invention.

FIG. 6 is a block diagram showing a configuration of a coded signal reproducing device according to a fourth embodiment of the present invention.

FIG. 7 is a schematic diagram showing changes in the reproduced signal of the encoded signal reproducing device in the fourth embodiment of the present invention.

FIG. 8 is a block diagram showing a configuration of a conventional coded signal reproduction device.

FIG. 9 is a schematic diagram showing changes in a reproduction signal of a conventional encoded signal reproduction device.

[Explanation of symbols]

 11, 41, 61, 81 Sync extractor 12, 42, 62, 82 Demultiplexer 13, 43, 63, 83 Dequantizer 14, 44, 64, 84 Subband signal switcher 15, 46, 65 Subband synthesis Filter 45 Mute device 66 Memory

Claims (5)

[Claims] 1. A sub-band signal obtained by band-dividing an audio signal into M (positive integer) sub-bands, with a sub-band sample group consisting of one data signal of all sub-bands as a basic unit. A coded signal reproduction device for reproducing an audio signal from a coded signal obtained by coding (a positive integer) number of subband sample groups as one frame, wherein the coded signal is input to extract a synchronization pattern. Sync extraction means for outputting a sync signal, a speed signal indicating speed information for reproduction, the synchronization signal and the encoded signal are input, and the code is applied when the reproduction speed indicated by the speed signal is faster than the speed at the time of recording. The thinned signal is thinned out on a frame-by-frame basis, a subband signal switching command signal is output when thinning is performed, and the coded signals other than the thinned one are demulled. Demultiplexer for outputting bit allocation information indicating bit allocation of quantized samples by plexing, scale factor information indicating subband signal level, and quantized sample information, the bit allocation information, the scale factor information, and the quantization Dequantizing means for inputting sample information, dequantizing and outputting a subband signal, and inputting the subband signal, and when the subband signal switching command signal is not output, the subband signal as it is When output as a reproduction sub-band signal, and when the sub-band signal switching command signal is output, the sub-band signal of the frame immediately before the thinned-out frame is held as the first sub-band signal, and the sub-quantization unit continues to output the sub-band signal. Sub of the frame immediately after the thinned frame to be output The band signal is input as the second subband signal, and the number of subbands that output the first subband signal increases as the number of the subband sample group increases from 1st to Kth within one frame. The reproduction subband signal is output by switching between the first subband signal and the second subband signal so that the number of subbands that gradually decrease and output the second subband signal gradually increase. A coded signal reproducing device, comprising: a subband signal switching unit for performing the above, and a subband synthesizing filter for inputting the reproduced subband signal and performing band synthesis to output a reproduced audio signal. 2. A sub-band signal obtained by band-dividing an audio signal into M (positive integer) sub-bands, with a sub-band sample group consisting of signals for each data of all sub-bands as a basic unit. A coded signal reproduction device for reproducing an audio signal from a coded signal obtained by coding (a positive integer) number of subband sample groups as one frame, wherein the coded signal is input to extract a synchronization pattern. A sync extraction means for outputting a sync signal, a speed signal indicating speed information for reproduction, the synchronization signal and the encoded signal are input, and when the reproduction speed indicated by the speed signal is faster than the speed at the time of recording, the subband A signal switching command signal is output, the coded signal is demultiplexed, bit allocation information indicating the bit allocation of the quantized sample, and a scale indicating the level of the subband signal. Demultiplexer for outputting factor information and quantized sample information, inputting the bit allocation information, the scale factor information, and the quantized sample information, and dequantizing means for dequantizing and outputting a subband signal, When the sub-band signal is input and the sub-band signal switching command signal is not output, the sub-band signal is output as it is as a reproduction sub-band signal, and when the sub-band signal switching command signal is output, it is continuous. The sub-band signal of the previous frame of the two frames is held as the first sub-band signal, and the continuous 2 is continuously output from the dequantizing means.
The subband signal of the subsequent frame of one frame is input as the second subband signal, and the
The number of subbands that output the first subband signal gradually decreases as the number of the subband sample group increases from the #th to the Kth, and the number of subbands that outputs the second subband signal. Sub-band signal switching means for switching the first sub-band signal and the second sub-band signal so as to output a reproduced sub-band signal so as to increase gradually, and inputting the reproduced sub-band signal for band synthesis. And a subband synthesizing filter for outputting a reproduced audio signal. 3. A sub-band signal obtained by band-dividing an audio signal into M (positive integer) sub-bands, with a sub-band sample group consisting of signals for each sub-band as a basic unit, K A coded signal reproduction device for reproducing an audio signal from a coded signal obtained by coding (a positive integer) number of subband sample groups as one frame, wherein the coded signal is input to extract a synchronization pattern. Sync extraction means for outputting a sync signal, a speed signal indicating speed information for reproduction, the synchronization signal and the encoded signal are input, and the code is applied when the reproduction speed indicated by the speed signal is faster than the speed at the time of recording. The thinned signal is thinned out on a frame-by-frame basis, a subband signal switching command signal is output when thinning is performed, and the coded signals other than the thinned one are demulled. Demultiplexer for outputting bit allocation information indicating the number of bit allocation of quantized samples by plexing, scale factor information indicating the level of a subband signal, and quantized sample information, the bit allocation information, the scale factor information, and the quantum Dequantization means for inputting the dequantized sample information, dequantized and outputting a subband signal, and the subband signal of a subband having a frequency lower than the N-th band (N is an integer 1 <N <M) When the sub-band signal switching command signal is not output, the sub-band signal is directly output as a reproduction sub-band signal, and when the sub-band signal switching command signal is output, immediately before the thinned frame. Holding the sub-band signal of the frame as a first sub-band signal, The sub-band signal of the frame immediately after the decimated frame continuously output from is input as the second sub-band signal, and as the number of the sub-band sample group increases from 1st to Kth in one frame, The first subband signal and the first subband signal are output so that the number of subbands outputting the first subband signal gradually decreases and the number of subbands outputting the second subband signal gradually increases. Subband signal switching means for switching the second subband signal to output a reproduction subband signal on the low frequency side; and inputting the subband signal of a subband having a frequency higher than the Nth band, When the signal switching command signal is not output, the subband signal is directly output as a reproduction subband signal, and the subband signal is output. When the switching command signal is output, the first subband signal and the second subband signal are switched in a subband having a frequency lower than the Nth band to output a reproduction subband signal on the low frequency side. Mute means for outputting a silent signal as a reproduction subband signal on the high frequency side, and a subband composition for inputting the reproduction subband signals on the low frequency side and the high frequency side to output a reproduction audio signal in a frame A coded signal reproducing device comprising a filter. 4. A sub-band signal obtained by band-dividing an audio signal into M (positive integer) sub-bands, with a sub-band sample group consisting of signals for each data of all sub-bands as a basic unit. A coded signal reproduction device for reproducing an audio signal from a coded signal obtained by coding (a positive integer) number of subband sample groups as one frame, wherein the coded signal is input to extract a synchronization pattern. A sync extracting means for outputting a sync signal, a speed signal indicating speed information for reproduction, the sync signal and the encoded signal are inputted, and the code is applied when the reproduction speed indicated by the speed signal is slower than the speed at the time of recording. The repeat processing is performed for each frame on a framed signal, and when the repeat processing is performed, a subband signal switching command signal is output, and the coded signal after the repeat processing is demarcated. Demultiplexer that outputs bit allocation information indicating the number of bit allocation of quantized samples by chipplexing, scale factor information indicating the level of a subband signal, and quantized sample information, the bit allocation information, the scale factor information, and the quantum Dequantization means for inputting the dequantized sample information, dequantizing and outputting a subband signal, inputting the reproduction subband signal, holding the memory means for holding for a certain period, and inputting the subband signal, When the sub-band signal switching command signal is not output, the sub-band signal is directly output as a reproduction sub-band signal, and when the sub-band signal switching command signal is output, the front side of the two repeated frames is output. Holds the subband signal of the frame as the first subband signal , The sub-band signal of the immediately preceding frame before the repeat processing of the sub-band signal of the preceding frame of the two repeated frames stored in the memory means is input as the second sub-band signal, and 1 As the numbers of the subband sample groups from the 1st to the Kth in the frame increase, the number of subbands that output the first subband signal gradually decreases and outputs the second subband signal. Sub-band signal switching means for switching the first sub-band signal and the second sub-band signal so as to output a reproduced sub-band signal so that the number of sub-bands to be increased gradually, A coded signal reproducing apparatus comprising: a sub-band synthesizing filter for inputting, band-synthesizing, and outputting a reproduced audio signal. 5. A sub-band signal obtained by band-dividing an audio signal into M (positive integer) sub-bands, with a sub-band sample group consisting of signals of all data of all sub-bands as a basic unit, K A coded signal reproduction device for reproducing an audio signal from a coded signal obtained by coding (a positive integer) number of subband sample groups as one frame, wherein the coded signal is input to extract a synchronization pattern. Sync extraction means for outputting a sync signal, a speed signal indicating speed information for reproduction, the synchronization signal and the encoded signal are input, and the code is applied when the reproduction speed indicated by the speed signal is faster than the speed at the time of recording. The thinning process is performed for each frame on the encoded signal and the repeat process is performed on each frame for the encoded signal at a slow time,
A subband signal switching command signal is output when either decimation or repeat is performed, and the coded signal after decimation or repeat is demultiplexed to indicate the bit allocation number of quantized samples. Demultiplexer that outputs bit allocation information and scale factor information indicating the level of a subband signal and quantized sample information, and inputs the bit allocation information, the scale factor information, and the quantized sample information, and dequantizes them. Dequantizing means for outputting a subband signal, memory means for inputting the reproduced subband signal and holding it for a certain period of time, inputting the subband signal, and not outputting the subband signal switching command signal Sometimes the sub-band signal is output as it is as a reproduction sub-band signal, and the sub-band signal is output. When the reproduction speed is higher than the recording speed, the subband signal of the frame immediately before the thinned frame is held as the first subband signal and is continuously output from the dequantizing means. The sub-band signal of the frame immediately after the thinned frame is input as a second sub-band signal, the sub-band signal switching command signal is output, and the sub-band signal is repeated when the reproduction speed is slower than the recording speed. Holding the sub-band signal of the front frame of the two frames as the first sub-band signal, and storing the sub-band signal of the front frame of the two repeated frames stored in the memory means. The subband signal of the immediately preceding frame before the repeat processing is input as the second subband signal, As the number of the subband sample groups from the 1st to the Kth in the frame increases, the number of subbands that output the first subband signal gradually decreases and the second subband signal is output. Subband signal switching means for switching the first subband signal and the second subband signal so as to output a reproduction subband signal so that the number of subbands to be output gradually increases; and the reproduction subband signal. And a subband synthesizing filter for synthesizing and band-synthesizing and outputting a reproduced audio signal.