JP3443391B2 - Digital decoding method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital decoding method for sequentially decoding a coded bit stream of each scene so that a plurality of scenes of a moving image can be reproduced seamlessly.

[0002]

2. Description of the Related Art MPEG (Moving Picture Experts Group) is a typical example of a moving image compression coding technique. Japanese Unexamined Patent Publication No. 9-331524 discloses MPs.
A digital encoding method for seamlessly combining multiple bitstreams (multiple scenes) encoded according to the EG standard is disclosed. According to this method
Near the end of the first bitstream and the beginning of the second bitstream in order to seamlessly combine the second scene after the first scene so that decoder buffer corruption (overflow and underflow) does not occur. And the encoding bit rate (compression rate) of each is limited.

[0003]

In the above-mentioned conventional technique, the encoder side prepares for seamless reproduction of a plurality of scenes. Therefore, there is a problem that a scene to be combined is specified. There is also a problem that image quality varies within each scene.

An object of the present invention is to realize seamless reproduction of a plurality of scenes only on the decoder side without intervention of an encoder.

[0005]

In order to achieve the above object, a first method according to the present invention is such that a coded bit stream of a moving image read from a data storage medium is supplied to a decoder via a buffer. A digital decoding method configured to sequentially decode a bitstream of each scene so that a plurality of scenes can be reproduced seamlessly in a digital reproducing apparatus configured to as amount of data in the buffer becomes more than the specified amount in the bus
The amount of data in the buffer, the number of frames in the current scene, and the compression level.
The read rate of the current scene from the data storage medium is controlled on a scene-by- scene basis in accordance with the setting.

A second method according to the present invention is a digital reproducing apparatus configured so that a coded bit stream of a moving image is supplied to a decoder via a buffer,
A digital decoding method for sequentially decoding a bitstream of each scene so that a plurality of scenes are seamlessly reproduced, and predicting the amount of data in the buffer at the decoding start time of the first frame of each scene,
If the predicted amount of data at the start of decoding a certain scene is less than the specified amount, the data of the first scene is stored in the buffer by the amount obtained by adding the insufficient amount of data to the specified amount. And the decoding of the first frame of the first scene is started.

A third method according to the present invention is a digital reproducing apparatus configured so that a coded bit stream of a moving image is supplied to a decoder via a buffer,
A digital decoding method for sequentially decoding a bitstream of each scene so that a plurality of scenes are reproduced seamlessly, wherein the input rate of each scene to the buffer is controlled under a certain limit, and Predict the amount of data in the buffer at the start of decoding the first frame, and if the predicted amount of data at the start of decoding a scene is less than the specified amount, add the insufficient amount of data to the specified amount. Wait for the data of the first scene to be stored in the buffer by the amount obtained by
It is decided to start decoding the first frame of the first scene.

[0008]

1 is a block diagram of a first recording / reproducing apparatus using a digital decoding method according to the present invention. In FIG. 1, 11 is an encoding unit (encoder), 12 is a data storage medium, 13 is a data read control unit, 14 is a data input control unit, 15 is a bitstream storage unit (buffer), 16 is a data output control unit, 1
7 is a decoding unit (decoder), 18 is a transfer rate control unit,
BS1 to BS6 are coded bitstreams.

The apparatus shown in FIG. 1 is a so-called codec having an encoder 11 and a decoder 17. The encoder 11 generates a bitstream BS1 by compressing and encoding the moving image signal. The generated bitstream BS1 is recorded in the data storage medium 12. The data storage medium 12 is a recording medium such as a magneto-optical disk, a magnetic tape and a semiconductor memory. The data read control unit 13 can read the bitstream BS2 from the data storage medium 12 at a variable rate. The data input control unit 14 includes the data read control unit 13
The bit stream BS3 and the synchronization control signal are received from the bit stream BS3 and the bit stream BS4 is written in the buffer 15 in synchronization with the reading of the bit stream BS2. The buffer 15 is a memory for temporarily storing the bit stream BS4 supplied from the data input control unit 14.
The data output control unit 16 reads the bitstream BS5 from the buffer 15. The decoder 17 decodes the bit stream BS6 supplied from the data output control unit 16 and outputs the reproduced moving image signal obtained as a result.
The transfer rate control unit 18 uses the data input information (write address information of the buffer 15) obtained from the data input control unit 14 and the data output information (read address information of the buffer 15) obtained from the data output control unit 16 as a buffer. Data remaining amount in 15 is calculated, and based on the information of the remaining amount of data obtained by the calculation and the image compression information such as the number of frames of each scene and the compression rate, the data storage medium 1
The read rate of 2 is determined, and information relating to this rate is given to the data read control unit 13 as transfer rate information.

The bit rate (buffer input rate) of the bit stream BS4 input to the buffer 15 is equal to the read rate of the data storage medium 12 and variable. The bit rate (decoding rate) related to the decoding of the decoder 17 matches the encoding bit rate (compression rate) for each scene. Note that the buffer 15 is assigned a part of the memory resource shared by the encoder 11, the decoder 17, and other circuit blocks.

FIG. 2 shows a decoding buffer 1 in FIG.
5 shows the change over time in the amount of data. DI1,
DI2 and DI3 are the first and
It is an input period of the bitstreams of the second and third scenes. The input period DI2 of the second scene starts from time t20, and the input period DI3 of the third scene starts from time t30. Each scene is, for example, one GOP
(Group Of Picture). The broken line drawn by the one-dot chain line in FIG. 2 shows the locus of the amount of data in the buffer 15 when the buffer input rates of the respective scenes are set to the same value Ri, and the broken line drawn by the solid line in FIG. The loci of the data amount in the buffer 15 when the buffer input rates of the first, second and third scenes are set to Ri (1), Ri (2) and Ri (3) respectively are shown. The slope of the “upward line segment” of the polygonal line represents the buffer input rate, and the length of the “vertical line segment” of the polygonal line represents the data amount of the picture for each frame. Here, it is assumed that the withdrawal of each picture data from the buffer 15 is instantly completed.

Sv in FIG. 2 is the size of the virtual buffer used for controlling the code generation amount in the encoder 11.
That is, VBV (Video Buffer in) defined by MPEG
g Verifier) Indicates the buffer size. That is, for each scene, data of the specified amount Sv or more is stored in the buffer 1.
Once stored in 5, the decoding of the first frame can start. The specified data amount Sv is also the maximum amount of data that can be extracted from the buffer 15 at one time.

T in FIG. 2 represents one frame period. NTSC (National Television System Committe)
If the video rate of e) (30 frames / s), T =
33.3 ms.

According to FIG. 2, at time t11, which is one frame period T after time t10, the buffer 15
The data amount of the first scene in the number reaches the specified amount Sv.
Therefore, the decoding of the first scene can be started from time t11. Further, the buffer 15 at the decoding start time t21 of the first frame of the second scene
Since the amount of data in is equal to or greater than the specified amount Sv in both the dashed line and the solid line, it is possible to seamlessly combine the second scene with the first scene without causing an underflow in the buffer 15. You can Furthermore, the third
Since the amount of data in the buffer 15 at the decoding start time t31 of the first frame of the scene is the specified amount Sv on both the dashed line and the solid line, the buffer 1
The third scene can be seamlessly combined with the second scene without causing underflow in 5.

To describe the locus of the alternate long and short dash line in FIG. 2 in detail, at time t20, the prescribed amount Sv of data of the last frame of the first scene is extracted from the buffer 15 and output to the decoder 17. In the subsequent one frame period T, the data amount in the buffer 15 is restored to the specified amount Sv by the data input (buffer input rate Ri) of the second scene. Then, at time t21, the data of the specified amount Sv of the first frame of the second scene is stored in the buffer 1
As a result of being extracted from 5, and output to the decoder 17, the amount of data in the buffer 15 becomes zero. That is, time t
The data amount in the buffer 15 before data extraction in 20 is 2 × Sv−Ri × T. Therefore, even considering the worst case in which the specified amount of data Sv is extracted from the buffer 15 twice in a row, when the total capacity of the buffer 15 is Sb, Sb = 2 × Sv−Ri × T (1) As long as the buffer full capacity Sb satisfies the above condition, the second scene can be seamlessly combined with the first scene without causing failure (overflow and underflow) of the buffer 15. Of course, the total buffer capacity Sb may be larger than 2 × Sv−Ri × T.

Further, according to the locus of the alternate long and short dash line in FIG.
When the number of frames of the second scene is N (2) and the coding bit rate (compression rate) of the same scene is Re (2), in order to seamlessly combine the third scene with the second scene, In order to secure the specified data amount Sv at time t31, the data amount in the buffer 15 is set to 0 within the time T × N (2) from time t21 to time t31.
From this, it is understood that it is necessary to increase the specified amount Sv at the rate of Ri-Re (2). That is, the buffer input rate Ri may be set so as to satisfy Ri = Sv / (T × N (2)) + Re (2) (2).

Specifically, Sv = 1.835 Mbit,
T = 33.3 ms, N (2) = 15, Re (2) = 8 Mbps
Then, from the formula (2), Ri = 11.167 Mbps, and from the formula (1), Sb = 3.281 Mbps. That is, the total buffer capacity Sb is set to 3.281 Mbit.
Above, the buffer input rate (data storage medium 12
(Reading rate of each) Ri is set to 11.67 Mbps, thereby sequentially decoding the bitstreams of the respective scenes so that the three scenes can be reproduced seamlessly without causing the failure (overflow and underflow) of the buffer 15. be able to.

In the locus indicated by the solid line in FIG. 2, at time t20, a smaller amount of data than the prescribed amount Sv of the last frame of the first scene is withdrawn from the buffer 15 whose full capacity Sb is filled with data. It is supposed to be output to the decoder 17. In the subsequent one frame period T, the data amount in the buffer 15 is recovered by exceeding the specified amount Sv due to the data input of the second scene. Therefore, at the time point when the data of the specified amount Sv of the first frame of the second scene is extracted from the buffer 15 and output to the decoder 17 at time t21, α (>) is stored in the buffer 15.
Only 0) data remains. In accordance with the remaining data amount α, the transfer rate control unit 18 determines that the buffer 1 at the decoding start time t31 of the first frame of the third scene
Buffer input of the second scene according to Ri (2) = (Sv−α) / (T × N (2)) + Re (2) (3) so that the amount of data in 5 becomes the specified amount Sv. The rate (reading rate of the data storage medium 12) Ri (2) is controlled. Even in this case, since the specified amount of data Sv is secured in the buffer 15 at the decoding start time t31, underflow does not occur in the buffer 15 and
The third scene can be seamlessly combined with the second scene. Moreover, under the same conditions as the case of the locus indicated by the alternate long and short dash line, for example, if α = 0.5 Mbit, then Ri (2) = 10.67 Mbps is obtained from the equation (3).
The power consumption in the input period DI2 of the second scene can be reduced.

As described above, according to the configuration of FIG. 1, by controlling the read rate of the data storage medium 12 in accordance with the remaining amount of data in the buffer 15 for each connection point of scenes, a plurality of decoders alone can be used. It is possible to realize seamless playback of the scene.

Note that, for example, the remaining amount of data β at time t20 in FIG. 2, that is, the data of the last frame of the first scene is pulled out from the buffer 15 and the decoder 17 is operated.
The buffer input rate Ri (2) of the second scene (readout rate of the data storage medium 12) Ri (2) may be controlled according to the amount of data in the buffer 15 at the time of being output to.

FIG. 3 is a block diagram of a second recording / playback apparatus using the digital decoding method according to the present invention.
In FIG. 3, 11 is an encoding unit (encoder), 12 is a data storage medium, 13 is a data read control unit, 14
Is a data input control unit, 15 is a bit stream storage unit (buffer), 16 is a data output control unit, 17 is a decoding unit (decoder), 19 is a decoding control unit, and BS1 to BS6
Is a coded bitstream.

The apparatus shown in FIG. 3 is a so-called codec having an encoder 11 and a decoder 17. The encoder 11 generates a bitstream BS1 by compressing and encoding the moving image signal. The generated bitstream BS1 is recorded in the data storage medium 12. The data storage medium 12 is a recording medium such as a magneto-optical disk, a magnetic tape and a semiconductor memory. The data read control unit 13 reads the bitstream BS2 from the data storage medium 12 at a fixed rate. The data input control unit 14 receives the bitstream BS3 from the data read control unit 13. The buffer 15 is a variable capacity memory for temporarily storing the bit stream BS4 supplied from the data input control unit 14. A part of the memory resources shared by the encoder 11, the decoder 17, and other circuit blocks is allocated to the buffer 15. The data output controller 16 uses the buffer 1
5. Read the bitstream BS5 from 5. Decoder 1
7 decodes the bit stream BS6 supplied from the data output control unit 16 and outputs the reproduced moving image signal obtained as a result. The decoding control unit 19 stores in the buffer 15 at the decoding start time of the first frame of each scene based on the image compression information such as the number of frames and compression rate of each scene, and the reproduction information such as the reproduction order of the scene. The amount of data is predicted, and if the predicted amount of data at the start of decoding a certain scene is less than the specified amount (VBV buffer size) Sv, the buffer 15 is assumed to have no shortage of the amount of data. Buffer size information is given to the data input control unit 14 and the data output control unit 16 so as to increase the capacity Sb (see the equation (1)) by the insufficient data amount γ, and then the insufficient data amount γ is added to the specified amount Sv. The data of the first scene is the buffer 1
The decoding start information is given to the data output control unit 16 so that the decoding of the first frame of the first scene is started after the data is stored in the data 5. The decoding control unit 19 also receives the data input information (write address information of the buffer 15) obtained from the data input control unit 14 and the data output information (read address information of the buffer 15) obtained from the data output control unit 16. Therefore, the remaining amount of data in the buffer 15 can be grasped.

FIG. 4 shows the change over time in the amount of data in the buffer 15 shown in FIG. Buffer 15
The input bit rate (buffer input rate) Ri is constant regardless of the scene. The broken line drawn by the one-dot chain line in FIG. 4 indicates that the underflow occurs in the buffer 15 at time t31 ′ when the decoding of the first frame of the first scene starts at time t11. It is a locus of quantity. The input periods of the bitstreams of the first, second and third scenes to the buffer 15 in this case are DI1 ', DI2' and DI3 '. The input period DI2 ′ of the second scene is time t20 ′.
Therefore, the input period DI3 ′ of the third scene is time t30 ′.
Each starts with. Further, the broken line drawn by the solid line in FIG. 4 indicates that the underflow does not occur in the buffer 15 when the decoding of the first frame of the first scene is delayed until time t12. It is a trail. In this case, the input period of the bitstreams of the first, second and third scenes to the buffer 15 is DI.
1, DI2 and DI3. The input period DI2 of the second scene is the input period DI of the third scene from time t20.
3 starts at time t30.

According to FIG. 4, at time t11 when one frame period T has passed from time t10, the buffer 15
The data amount of the first scene in the number reaches the specified amount Sv.
In the locus of the alternate long and short dash line, the decoding of the first frame of the first scene starts at this time t11. Even in this case, at time t21 ′, which is one frame period T after time t20 ′, the specified amount of data Sv is secured in the buffer 15, and thus the decoding of the first frame of the second scene is immediately started. be able to. However,
Time t when one frame period T has passed from time t30 '
In 31 ′, since the data amount is insufficient γ with respect to the specified amount Sv in the buffer 15, the decoding of the first frame of the third scene cannot be immediately started, and the specified amount of Sv is not stored in the buffer 15. It is necessary to wait until time t32 'when data is stored. That is, time t31 '
From the time to the time t32 ', a so-called VBV delay (startup delay) occurs, and the third scene cannot be seamlessly combined with the second scene. For example, if the encoding bit rate of the second scene is considerably high, seamless playback may not be realized without taking measures such as increasing the buffer input rate of the second scene (equation ( See 3)).

The insufficient data amount γ is set to N (1) and N (2) for the number of frames in the first and second scenes, respectively, and the coding bit rate (compression rate) for both scenes is set to R.
When e (1) and Re (2), Can be estimated as. That is, the insufficient data amount γ is the difference Re (2) −Re (1) between the compression rates of two consecutive scenes.
And the ratio N (2) / N (1) of the number of frames of both scenes.

According to the solid line trajectory in FIG. 4, time t11
Until time t12 when a period DT has elapsed from that time, that is, until time t12 when the data of the first scene is stored in the buffer 15 by an amount obtained by adding the shortage amount γ predicted by the equation (4) to the specified amount Sv. Wait, start decoding the first frame of the first scene. As a result, the decoding start time t21 of the first frame of the second scene
Not only does the buffer 15 never underflow at the decoding start time t31 of the first frame of the third scene. That is, for the first scene, the second scene
The third scene can be seamlessly combined with the second scene. However, to prevent overflow in the buffer 15,
It is necessary to increase the total capacity of the buffer 15 to Sb + γ.

As described above, according to the configuration of FIG. 3, by delaying the start of decoding the first scene in accordance with the expected amount of insufficient data γ, seamless reproduction of a plurality of scenes can be realized only on the decoder side. be able to.

In the example of FIG. 3, the data storage medium 1
The read rate of 2 is fixed, and therefore the input bit rate of the buffer 15 (buffer input rate) is assumed to be constant regardless of the scene. However, when the reading rate of the data storage medium 12 can be changed to a certain upper limit rate, the input rate of each scene to the buffer 15 should be controlled to some extent according to the method described with reference to FIGS. The method described with reference to FIGS. 3 and 4 may be adopted. According to this, the capacity increase γ of the buffer 15 in FIG. 4 can be reduced.

Also, by considering the number of scenes to be combined, the coding bit rate of each scene, and the number of frames of each scene, the capacity of the buffer 15 is secured to be large enough for the data shortage, so that there are four or more scenes. It is also possible to seamlessly combine.

In the first and second recording / playback apparatuses, it is assumed that the next scene to be seamlessly combined with the current scene is known. However, a scene combination request may be suddenly given during the reproduction of a certain scene. The recording / playback apparatus described below can handle such a situation.

FIG. 5 is a block diagram of a third recording / playback apparatus using the digital decoding method according to the present invention.
The transfer rate control unit 18a in FIG. 5 differs from the transfer rate control unit 18 in FIG. 1 in that it is configured to receive a scene combination request. Other configurations in FIG. 5 are similar to those in FIG.

The transfer rate control unit 18a in FIG. 5 responds to a scene combination request given during the reproduction of the current scene, and stores the data in the buffer 15 at the start of decoding the first frame of the next scene to be combined. Predict the amount of data,
When the predicted data amount is less than the specified amount Sv, the data storage medium 12 is made up so as to compensate for the insufficient data amount γ.
Control the read rate of the remaining frames of the current scene from.

FIG. 6 shows a decoding buffer 1 in FIG.
5 shows an example of the temporal change in the data amount in No. 5. D
I1, DI2, and DI3 are input periods of the bitstreams of the first, second, and third scenes to the buffer 15, respectively. The input period DI2 of the second scene is time t2.
From 0, the input period DI3 of the third scene starts from time t30. Here, it is assumed that the request for combining the third scene is given during the reproduction of the second scene. The locus drawn by the alternate long and short dash line in FIG. 6 indicates that the buffer input rates Ri (1), Ri (2), and Ri (3) of each scene are maintained in the buffer 15 at the same value Ri at time t31. It shows that underflow occurs.
The locus drawn by the solid line in FIG. 6 indicates the buffer input rate Ri (2) of the second scene Te in response to the scene combination request.
It is shown that underflow does not occur in the buffer 15 when the maximum read rate Rimax of the data storage medium 12 is increased for a short period indicated by.

According to the locus of the alternate long and short dash line in FIG.
At time t31 when one frame period T has elapsed from 30, since the data amount is insufficient γ with respect to the specified amount Sv in the buffer 15, the decoding of the first frame of the third scene cannot be started immediately, Buffer 1
It is necessary to wait until the time t32 when the data of the specified amount Sv is stored in 5. That is, from time t31 to time t3
Start-up delay occurs up to 2,
The third scene cannot be seamlessly combined with the second scene. Here, the insufficient data amount γ can be estimated by the above equation (4).

According to the trajectory of the solid line in FIG. 6, of the one-frame period T starting from the time t22 immediately after the scene combination request is given, the second scene buffer is used for the period Te required to supplement the insufficient data amount γ. The input rate (readout rate of the data storage medium 12) Ri (2) is the maximum rate Rimax
Can be increased to. That is, the buffer input rate Ri (2) is increased during the period Te that satisfies γ = (Rimax−Ri) × Te (5). Then, after the supplement of the insufficient data amount γ is completed, the original rate Ri is restored. Therefore, with a simple control after receiving the scene combination request, the buffer 15
It is possible to seamlessly combine the third scene with the second scene without causing underflow to the second scene.

FIG. 7 shows a decoding buffer 1 in FIG.
5 shows another example of the temporal change of the data amount in No. 5.
As in the case of FIG. 6, when a combine request for the third scene is given during the reproduction of the second scene, the locus drawn by the solid line in FIG. It is shown that underflow does not occur in the buffer 15 when the buffer input rate Ri (2) of the remaining frames of the scene (readout rate of the data storage medium 12) Ri (2) is increased to a certain rate Rie on average.

According to the trajectory of the solid line in FIG. 7, the buffer input rate Ri (2) of the second scene is insufficient data amount γ over a plurality of frame periods starting from time t22 immediately after receiving the scene combination request. It is increased to the rate Rie required to replenish. That is, when the number of remaining frames of the current scene (second scene) at the time when the scene combination request is given is Nr (2), the remaining time available for reading the current scene from the data storage medium 12 is T × Nr. (2)
In view of this, the rate Rie that satisfies Rie = Ri + γ / (T × Nr (2)) (6) is adopted as the read rate of the remaining frames. Therefore, as in the case of FIG.
2nd without causing underflow in the buffer 15
It is possible to seamlessly combine the third scene with the second scene.

As described above, according to the configuration of FIG. 5, the read rate of the remaining frames of the current scene from the data storage medium 12 is determined according to the amount of insufficient data γ expected at the time when the scene combination request is given. By controlling, it is possible to realize seamless reproduction of a plurality of scenes only on the decoder side.

Since Te <T in the example of FIG. 6, the read rate of only one frame of the remaining frames in the second scene is controlled. For example, T <Te <2T
In that case, the reading rate of two frames may be controlled.

[0040]

As described above, according to the present invention, seamless reproduction of a plurality of scenes can be realized only on the decoder side without intervention of an encoder.

[Brief description of drawings]

FIG. 1 is a block diagram of a first recording / playback apparatus using a digital decoding method according to the present invention.

FIG. 2 is a bitstream storage unit (buffer) in FIG.
It is a time chart figure showing the time change of the amount of data in.

FIG. 3 is a block diagram of a second recording / playback apparatus using the digital decoding method according to the present invention.

FIG. 4 is a bitstream storage unit (buffer) in FIG.
It is a time chart figure showing the time change of the amount of data in.

FIG. 5 is a block diagram of a third recording / playback apparatus using the digital decoding method according to the present invention.

FIG. 6 is a bitstream storage unit (buffer) in FIG.
FIG. 6 is a time chart diagram showing an example of temporal changes in the amount of data in FIG.

FIG. 7 is a bitstream storage unit (buffer) in FIG.
FIG. 6 is a time chart diagram showing another example of the temporal change of the data amount in FIG.

[Explanation of symbols]

11 Encoding unit (encoder) 12 Data storage media 13 Data read controller 14 Data input control section 15 Bitstream storage (buffer) 16 Data output controller 17 Decoding unit (decoder) 18, 18a Transfer rate control unit 19 Decoding control unit BS1 to BS6 bitstream

─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-9-182024 (JP, A) JP-A-9-331524 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04N 5/76-5/956 H04N 7/24-7/68

Claims (14)

(57) [Claims] 1. A digital reproducing apparatus configured to supply a coded bit stream of a moving image read from a data storage medium to a decoder via a buffer so that a plurality of scenes can be reproduced seamlessly. the bit stream of each scene a sequential digital decoding method for decoding, as the data amount in the buffer at the decoding start time of the first frame of the next scene is more than the specified amount, the data in the buffer Amount and number of frames in the current scene
And a decoding rate of the current scene from the data storage medium is controlled in units of one scene depending on the compression rate . 2. The digital decoding method according to claim 1, wherein the data storage medium is dependent on the amount of data in the buffer at the time when the data of the first frame of the current scene is output from the buffer to the decoder. A digital decoding method characterized by controlling the reading rate of the current scene from. 3. The digital decoding method according to claim 1, wherein the data of the last frame of the scene immediately preceding the current scene is output from the buffer to the decoder according to the amount of data in the buffer. A digital decoding method, characterized in that the reading rate of the current scene from the data storage medium is controlled. 4. The digital decoding method according to claim 1, wherein in response to a scene combination request given during reproduction of the current scene, the buffer at the time of starting decoding of the first frame of the next scene to be combined. Predict the amount of data inside, if the predicted data amount at the time of decoding start of the next scene is less than the specified amount, the current scene from the data storage medium to compensate for the lacking amount of data. A digital decoding method characterized by controlling the read rate of the remaining frames. 5. The digital decoding method according to claim 4, wherein the reading rate from the data storage medium is maximized until the insufficient data amount is supplemented. 6. The digital decoding method according to claim 4, wherein the increment of the read rate is obtained by dividing the data amount of the shortage by the remaining time that can be used for reading the current scene from the data storage medium, A digital decoding method, characterized in that the reading rate of the remaining frames of the current scene from the storage medium is increased by the increment. 7. A digital reproducing apparatus configured so that a coded bit stream of a moving image is supplied to a decoder via a buffer, and a bit stream of each scene is sequentially reproduced so that a plurality of scenes can be reproduced seamlessly. A digital decoding method for decoding, which predicts the amount of data in the buffer at the decoding start time of the first frame of each scene, and when the predicted data amount at the decoding start time of a certain scene is less than the specified amount Waiting for the data of the first scene to be stored in the buffer by an amount obtained by adding the insufficient amount of data to the specified amount, and then starting decoding the first frame of the first scene. Characteristic digital decoding method. 8. The digital decoding method according to claim 7, wherein the capacity of the buffer, which is determined on the assumption that the data amount is not insufficient, is increased by the amount of the insufficient data. And a digital decoding method. 9. The digital decoding method according to claim 8, wherein the capacity increment of the buffer is determined according to a difference in compression rate between two consecutive scenes and a ratio of the number of frames. . 10. A digital reproducing apparatus configured so that a coded bitstream of a moving image is supplied to a decoder via a buffer, and the bitstream of each scene is sequentially reproduced so that a plurality of scenes can be reproduced seamlessly. A digital decoding method for decoding, in which the input rate of each scene to the buffer is controlled under a certain limit, and the amount of data in the buffer at the time of starting decoding of the first frame of each scene is predicted, If the predicted amount of data at the start of decoding a certain scene is less than the specified amount, the data of the first scene is stored in the buffer by the amount obtained by adding the insufficient amount of data to the specified amount. A digital decoding method, characterized in that, after waiting for, the decoding of the first frame of the first scene is started. 11. The digital decoding method according to claim 10, wherein the data of the first frame of each scene is stored in the buffer in accordance with the amount of predicted data in the buffer at the time when the data is output from the buffer to the decoder. A digital decoding method characterized by controlling the input rate of each scene of. 12. The digital decoding method according to claim 10, wherein the data of the last frame of the immediately preceding scene is output from the buffer to the decoder according to the predicted data amount in the buffer, A digital decoding method characterized by controlling an input rate of each scene to the buffer. 13. The digital decoding method according to claim 10, wherein the capacity of the buffer, which is determined on the assumption that the data amount is not insufficient, is increased by the amount of the insufficient data. And a digital decoding method. 14. The digital decoding method according to claim 13, wherein the capacity increment of the buffer is determined according to the difference in the compression rates of two consecutive scenes and the ratio of the number of frames. .