JP4193240B2 - Compressed encoded data decoding apparatus and karaoke apparatus using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable performing the decoding of data from the arbitrary position of compression-encoded data by reproducing the data received from the arbitrary frame of the compression-encoded data based on decode parameters in a header after receiving the header. SOLUTION: Decode parameters and Twin VQ frame data are transmitted from a sequencer 11 and when the decode parameters are transmitted, they are temporarily stored in a memory 60. An acoustic data processing part 8 receives the decode parameters, that is, data including a sampling rate and a bit rate before it reproduces karaoke music to store them in the memory. In this state, the part 8 decodes the Twin VQ frame data from the sequencer 11 based on these parameters. When the parameters are temporarily stored in the memory, even when the position of a pointer on a sound control track is changed by a fast forwarding operation or the rewinding operation or the like thereafter, the part 8 can perform the decoding of data from a frame data corresponding to that position.

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a compression-encoded audio data decoding apparatus and a karaoke apparatus using the same, and more particularly to an apparatus that compresses and encodes audio data such as audio data by a vector quantization method.
[0002]
[Prior art]
A vector quantization method has been proposed as a technique for compressing and encoding an acoustic signal such as an audio signal with high efficiency. For example, a vector quantization method called TwinVQ (Transform-domain Weighted Interleave Vector Quantization developed by NTT) cuts out a signal to be compressed and encoded in a certain interval. Each information pattern is interleaved to create a plurality of target vectors, and a code book search is performed on the target vectors to transmit the index of the nearest pattern vector. At the time of decoding, the original signal is restored using the same codebook used at the time of compression. In this vector quantization method, the code obtained by codebook search is used as a compression code, so the compression rate can be increased while maintaining a high sound quality level, and a fixed-length frame method that cuts out a signal in a certain interval is used. Because it is adopted, it is strong against code errors. Therefore, a compression coding method based on a vector quantization method typified by TwinVQ is used particularly for compression of an acoustic signal such as a music signal or a voice signal.
[0003]
Note that a compression coding method that performs vector quantization using a codebook is disclosed in, for example, Japanese Patent Laid-Open No. 10-112657.
[0004]
Conventionally, the data obtained by the above-described compression coding method is composed of a header 1 and a data body 2 as shown in FIG. 9, and the header 1 is a decoding composed of a sample rate and a bit rate necessary for decoding (decoding). The data body 2 is composed of compression-encoded data composed of a bit stream.
[0005]
At the time of decoding, first, the header 1 is obtained, the decoding parameters in this are interpreted, and then the decoding operation of the data body 2 is performed.
[0006]
[Problems to be solved by the invention]
However, in the conventional decoding device, the header 1 and the data body 2 are always recognized as one body, and when decoding, the decoding parameter in the header 1 is always first interpreted before the data body 2 is decoded. It was. For this reason, for example, the data body 2 is decoded halfway and temporarily stopped, and after a certain period of time, it is decoded again from an arbitrary part of the data body 2, or after being decoded partway through the data body 2, Operations such as starting decoding from the middle of the data itself could not be performed. That is, since decoding cannot be performed from the middle of the data, there is a disadvantage that the compressed encoded data cannot be used in an applied manner.
[0007]
An object of the present invention is to provide a compression-encoded data decoding apparatus capable of decoding from an arbitrary position of compression-encoded data.
[0008]
Another object of the present invention is to provide a karaoke apparatus that facilitates an applied operation of a voice signal, a musical instrument sound signal, or the like that MIDI is not good at by adopting the decoding device.
[0009]
[Means for Solving the Problems]
According to the first aspect of the present invention, a header including a decoding parameter for decoding compression-encoded data vector-quantized using a codebook, and the compressed-encoded data are stored in a state of being divided into fixed frames. A vector quantized data storage means; and a decoding means for reproducing the compressed encoded data stored in the vector quantized data storage means,
The decoding means includes
Decode parameter storage means for receiving and temporarily storing the decode parameter included in the header from the vector quantized data storage means when reproducing the compressed encoded data stored in the vector quantized data storage means And receiving the compressed encoded data stored in the vector quantized data storage means from the intermediate frame in a state where the decode parameters are stored in the decode parameter storage means, and receiving the received data And a playback unit for playing back based on the decode parameters stored in the decode parameter storage means .
[0010]
The compression-coded data of the present invention is obtained by vector quantization of a signal to be compressed using a code book. There is TwinVQ as a vector quantization method using a codebook. TwinVQ creates a plurality of target vectors by cutting out and compressing an acoustic signal or the like to be compressed at fixed intervals, selects a pattern vector closest to the target vector from the codebook, and transmits the code. It is a thing. At the time of decoding (decoding), the same signal as above is used to restore the target vector corresponding to the code, and the original signal is obtained by passing through the DAC. The vector quantized data storage means stores the compressed encoded data that has been vector quantized in this way, and is obtained via a storage medium such as a CD-ROM, or a communication line such as the Internet or ISDN. When decoding the vector quantized data, the decoding means first receives the header and temporarily stores the decoding parameters. Thereafter, it is received from an arbitrary frame of vector quantized data, and is decoded (decoded) based on the temporarily stored decoding parameters. In the present invention, the header and the compression encoded data are handled separately, and the compression encoded data is divided into fixed frames. It enables compression-coded data to be combined from the frame position. Therefore, once the header is received and the decoding parameters are interpreted and stored, it is possible to decode not only from the first frame of the compressed encoded data but also from the intermediate frame, and from the intermediate frame. After that, it is possible to jump to an arbitrary frame position for decoding.
[0011]
Note that the decode parameter is composed of a sample rate and a bit rate. Depending on the vector quantization method, other information may be included. In short, the header only needs to include all information necessary for decoding the compression-encoded data.
[0012]
The invention of claim 3 is a karaoke apparatus in which the compression-encoded data decoding device is used, and sequence data such as a musical sound track is reproduced by a sequencer unit.
The vector quantized data storage means stores a header including decoding parameters, and acoustic sequence data that has been vector quantized and divided into frames,
The decoding means acquires a decoding parameter of a header before karaoke reproduction and stores it in the decoding parameter storage means, and based on the decoding parameter from an arbitrary frame designated by a sequencer unit in acoustic sequence data at the time of karaoke reproduction And playing.
According to the present invention, in a karaoke apparatus using the above-described decoding apparatus for compression-encoded data, the object of compression-encoding is acoustic sequence data including voices, instrument sounds, and the like. Then, the header decoding parameters are acquired before karaoke reproduction, and reproduction is performed based on the decoding parameters from an arbitrary frame designated by the sequencer unit in the acoustic sequence data during karaoke reproduction.
[0013]
For example, by providing the karaoke device with a means for specifying fast forward and rewind in the remote control device or the like, when the fast forward or rewind is designated, the sequencer unit includes the acoustic sequence data based on the amount of fast forward or rewind. Specify a frame in the middle of. At this time, since the decoding parameters can be acquired in advance, it is possible to directly reproduce from the designated intermediate frame. In addition, since the decoding parameter is first acquired by providing means for designating a medley performance by a plurality of music data, when reproducing the second and subsequent music of the medley music, the acoustic sequence data of the second and subsequent music You can directly specify the middle frame of the medley part.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a schematic configuration diagram of a karaoke apparatus according to an embodiment of the present invention. In this karaoke device, song data for karaoke song performance and moving image data for background video are stored in a hard disk 12 and a DVD (digital video disc) 40, and the user uses the remote control device 3 to specify a song number. When input, the music data of the music number is read and a karaoke music is played. The remote control device includes a fast-forward key 3a, a rewind key 3b, and a pause key 3c. By turning on these keys, the karaoke song being played can be fast-forwarded, rewinded, and paused.
[0015]
The karaoke system includes a karaoke device main body 1, a control amplifier 2, a remote control device 3, a DVD changer 4, a speaker 5, a monitor 6, a microphone 7, and an acoustic data processing unit 8.
[0016]
The karaoke apparatus body 1 receives and decodes an infrared signal transmitted from a control unit 10 that controls the operation of the entire system, a sequencer 11 that performs karaoke performance, a hard disk (HDD) 12 that stores music data, and the remote control device 3. A remote control receiving unit 13, a sound source device 14 that generates musical sounds based on music data, a pattern development unit 16 that develops character patterns of lyrics, and a display control unit 17 that controls the display of lyrics and background video are provided.
[0017]
The sound data processing unit 8 processes sound signals such as sound that cannot be reproduced by MIDI data and instrument sounds. In this karaoke apparatus, TwinVQ data of a TwinVQ data file designated by an acoustic control track included in music data stored in the hard disk 12 is input. The acoustic data processing unit 8 decodes the TwinVQ data, performs DA conversion, and outputs it to the control amplifier 2.
[0018]
The musical sound generated by the sound source device 14, the voice such as the chorus sound decoded by the acoustic data processing unit 8, the instrument sound, and the like are input to the control amplifier 2. The control amplifier 2 mixes these musical sound signals, sound signals, instrument sound signals, singing sound signals from the microphone 7, etc., gives an effect, and outputs them to the speaker 5.
[0019]
The DVD changer 4 selects and reads / reproduces one of the plurality of DVDs 40. The DVD stores a moving picture background image compressed by MPEG. The background video includes an individual background video individually corresponding to a specific karaoke song, a general-purpose background video provided for each karaoke song type (for example, genre), and the like. The DVD changer 4 and the pattern development unit 16 are connected to the display control unit 17. The display control unit 17 superimposes the character pattern developed by the pattern development unit 16 on the background video input from the DVD changer 4 and displays it on the monitor 6.
[0020]
The hard disk 12 stores music data of a large number of karaoke songs, as well as still image data and the like.
[0021]
The karaoke apparatus shown in the functional block diagram of FIG. 1 is configured by a computer system including a CPU, and the sequencer 11 and the pattern development unit 16 are realized by software.
[0022]
FIG. 2 is a diagram showing the composition of music data stored in the hard disk 12 or DVD 40. As shown in FIG. In this karaoke apparatus, music data is composed of a header, a track group, and an acoustic data section. As described above, the acoustic data is composed of TwinVQ data. In the header, bibliographic data such as a music number and a title, and decoding parameters of TwinVQ data which is compression-encoded data are written. In addition to the musical sound track that controls the sound source device 14, the track group indicates the lyric track in which the lyric data developed by the pattern development unit 16 into the character pattern is written, and the timing at which the acoustic data divided for each phrase is reproduced. The control unit includes an acoustic control track to be controlled, an effect control track for controlling effects of the sound source device 14 and the control amplifier 2, and the like.
[0023]
The musical sound track is configured by arranging event data such as note event data and setting data and timing data indicating the read timing of each event data in time series. The timing data is described by data such as a duration indicating a time interval between the event data and an absolute time from when the music starts to when each event data is generated. The sequencer 11 reads the event data at the timing indicated by the timing data and inputs it to the sound source device 14. The tone generator 14 generates a musical sound according to the input event data.
[0024]
The acoustic control track is configured by arranging, in time series, TwinVQ data numbers that are event data and timing data indicating the read timing of each event data. The sequencer 11 reads the TwinVQ data number at the timing specified by the timing data, reads the TwinVQ data specified by this number from the file, and outputs it from the corresponding data frame in this data to the acoustic data processing unit 8. As will be described later, the sequencer 11 actually designates a TwinVQ frame, and data corresponding to the frame is output to the acoustic data processing unit 8.
[0025]
The lyrics track and the effect control track are configured by arranging event data and timing data indicating the read timing of each event data in chronological order, similarly to the musical tone track and the sound control track.
[0026]
As described above, the remote controller 3 includes the fast forward key 3a, the rewind key 3b, and the pause key 3c. By operating these keys, the karaoke song being played can be fast forwarded or rewinded. That is, when the fast-forward key 3a is operated while a karaoke song is being played, the karaoke song is fast-forwarded while the karaoke song is being operated. The fast-forwarding of the karaoke song is to increase the stepping speed in the forward direction (time passage direction) in the sequencer 11. Further, when the rewind key 3b is operated, the karaoke song is rewinded during the operation. The rewinding of the karaoke song means that the stepping direction in the sequencer 11 is reversed and the stepping speed is made faster than the normal stepping speed. FIG. 3 shows a stepping operation in the sequencer 11. The pointer P indicates the position on the track that is currently processed by the sequencer 11 and advances rightward (in the direction of time) at a constant speed during normal karaoke song playback. If the fast-forward key 3a is operated at the position a1 while the karaoke song is being reproduced, the pointer P moves rightward at the stepping speed set for fast-forwarding. When the fast-forward key 3a is released when the pointer P reaches the position a2, the reproduction of the karaoke song at the normal stepping speed is resumed from the position a2. Further, when the rewind key is operated when the pointer P reaches the position a3, it proceeds in the left direction (opposite to the passage of time) at the stepping speed set for rewinding. When the rewind key 3b is released when the pointer P reaches the position a4, the pointer P advances again at a normal speed rightward from this time. That is, normal karaoke music is resumed. In this way, when either the fast-forward key 3a or the rewind key 3b is operated, the position of the pointer P is advanced to an arbitrary position by controlling the step speed and the step direction in the sequencer 11. It can be freely returned to the original arbitrary position.
[0027]
When fast-forwarding or rewinding, it is possible to play back each track according to the stepping speed at that time, but in the case of rewinding, control such as reversing the interpretation of the event data is necessary. come. For example, in a musical tone track, it is necessary to execute note-off for note-on event data and to execute note-on for note-off event data. In this embodiment, the acoustic control track is not reproduced at the time of fast-forwarding and rewinding. Separately, it is also possible to make a pseudo sound such as “Kyurukurukuru”.
[0028]
FIG. 4 is a functional block diagram of the TwinVQ creation unit. The TwinVQ data is sent as a part of music data by a CD-ROM or ISDN communication line (not shown) and stored in the hard disk 12 or the like. Accordingly, the TwinVQ data itself is not created by the karaoke apparatus, but is implemented as a part of the music data. As shown in FIG. 4, in creating TwinVQ data, a digitalized target acoustic signal is first developed in the memory 50, and the vector quantization unit 52 controls the predetermined sampling rate and bit under the control of the control unit 51. Compression-encoded data that is vector-quantized based on the rate, that is, TwinVQ frame data is generated. A code book 53 is used for vector quantization. As described above, vector quantization for obtaining TwinVQ frame data is performed as follows. That is, the acoustic data stored in the memory 50 is cut out for every predetermined section and interleaved to create a plurality of target vectors, and the nearest pattern vector is selected with reference to the code book 53 for each target vector. . The code at that time is output as TwinVQ frame data. Note that TwinVQ frame data is framed. In this embodiment, invalid bits are added to the frame length determined from the bit rate or the like (padding processing) so that the frame size is in bytes. By making it possible to handle the frame length in units of bytes, it becomes easy to specify a frame when playback is started from an intermediate frame during decoding. Also, the sampling rate and bit rate used at this time for decoding are output as decoding parameters. The decoding parameter is inserted into the header portion of the music data, and the TwinVQ frame data subjected to vector quantization is used as acoustic data.
[0029]
FIG. 5 is a functional block diagram of an acoustic data processing unit for decoding TwinVQ data. Decode parameters and TwinVQ frame data are sent from the sequencer 11 and are temporarily stored in the memory 60 in the case of decode parameters. In the case of TwinVQ frame data, the vector inverse quantization unit 61 performs inverse quantization to perform decoding. That is, by using the same code book 62 as the code book 53 used when creating TwinVQ data, the vector inverse quantization unit 61 can decode the original signal. Reference numeral 63 denotes a part for performing these controls, and the DAC 64 converts the decoded data into an analog signal and creates a signal to be output to the control amplifier 2.
[0030]
The acoustic data processing unit 8 receives from the sequencer 11 the decoding parameters before playback of the karaoke song, that is, data including the sampling rate and bit rate (included in the header portion of the song data in FIG. 2). Remember. In this state, the TwinVQ frame data sent from the sequencer 11 is decoded based on the decoding parameter. Once the decoding parameter is stored in the memory 60, even if the position of the pointer P on the acoustic control track changes due to fast-forwarding or rewinding, it can be decoded from the frame data corresponding to that position. . That is, when the position of the pointer P changes, there is no need to receive it again from the header part.
[0031]
FIG. 6 shows the structure of TwinVQ data and shows an example of data frame reproduction when fast forward and rewind are performed. First, the sampling rate and bit rate of the header part are temporarily stored in the memory 60, then the data frame is reproduced from No. 1 to No. 2, and when the fast forward key is pressed to the data frame i-1, the data When the playback of the frame i is performed and the rewind key 3b is operated to the first position of the data frame 2 when the playback of the frame i is finished, the playback from the data frame 2 starts again. If the operation end position of the fast forward key or rewind key is in the middle of a frame, the frame number at which playback is started next is from the next frame of the frame. As described above, once the header sampling rate and bit rate are sent to the acoustic data processing unit 8, even if the fast-forward key 3a and the rewind key 3b are operated thereafter, the acoustic data is immediately reproduced from the end of the operation. It becomes possible. Note that at the time of fast forward or rewind, the playback start frame is calculated based on the duration data of the track and one frame time of TwinVQ data specified by the event. That is, when the pointer is moved by the time during which the fast forward key or rewind key is operated, the pointer is moved on the sequence track and on the TwinVQ data of one file, and the total movement is controlled to be the key operation time.
[0032]
FIG. 7 is a flowchart showing a partial operation of the karaoke apparatus. In particular, this flowchart shows control of acoustic data when the fast-forward key 3a and the rewind key 3b of the remote controller 3 are operated during a karaoke performance.
[0033]
At the timing of initial setting of the karaoke performance (n1), in addition to the initial settings of the sound source device 14 and the display control unit 17, the decoding parameters including the sampling rate and the bit rate are sent to the acoustic data processing unit 8. These data are stored in the memory 60 in the acoustic data processing unit 8.
[0034]
When the karaoke performance starts, the sequencer 11 enters the reproduction mode (n2) and performs the sequence processing of each track (n3). When the fast-forward key 3a is operated by the remote control key 3 during the karaoke performance (n4), the stepping speed of the pointer P is increased N times until the fast-forward is stopped (n6). When fast-forwarding stops, the process proceeds to n7, and the data frame designated by the pointer at that time is set as the start position. Further, when the rewind key 3b is operated (n8), reverse stepping is performed at a speed N times that of the pointer P until rewinding stops (n10), and when rewinding stops, the process proceeds to n11. The data frame specified by the pointer is set as the start position.
[0035]
With the above operation, since the decoding parameters are stored in the acoustic data processing unit 8 at the initial initial setting stage, the desired data can be obtained without retransmitting the data from the beginning when the fast forward key or the rewind key is operated. Regeneration can be done immediately. In the above embodiment, the configuration and the operation when the fast forward key and the rewind key are operated are shown. However, in the karaoke apparatus, the medley performance of a plurality of music is designated, and the acoustic data of each music is also specified. Can be played from the middle frame. For example, as shown in FIG. 8, data frames 1 and 2 are reproduced for the acoustic data of music 1, data frames 2 and 3 are reproduced for music 2, and data frames 2 and 3 are reproduced for music 3. By storing the sampling rate and bit rate of the header 1 in the memory 60 of the acoustic data processing unit, the music 2 can be immediately reproduced from the data frame 2 without re-reading the header. The reproduction of the music piece 3 can also be started immediately from the data frame 2. In this case, it is assumed that the acoustic data of each musical piece is quantized using the same codebook and the same sampling rate and bit rate.
[0036]
【The invention's effect】
According to the first aspect of the present invention, since the decoding parameter is received first, it can be reproduced from an arbitrary frame, and various applied uses of the compression-encoded data are possible.
[0037]
Further, by adopting this compression-encoded data decoding apparatus in a karaoke apparatus, it is possible to freely reproduce sound sequence data such as voice and musical instrument sound that cannot be processed by MIDI from the middle. For example, in a karaoke device that employs a fast-forward function and a rewind function, the acoustic sequence data can be immediately reproduced from the fast-forward position or the rewind position, and in the karaoke device that employs a medley performance function, The subsequent medley music sequence data can be reproduced immediately from the middle frame.
[Brief description of the drawings]
FIG. 1 is a functional block diagram of a karaoke apparatus according to an embodiment of the present invention. FIG. 2 is a block diagram of music data. FIG. 3 is a diagram showing a storage operation of a sequencer. FIG. 5 is a functional block diagram of an acoustic data processing unit. FIG. 6 is a diagram showing a configuration of TwinVQ data. FIG. 7 is a flowchart showing a partial operation of a karaoke apparatus. FIG. 8 is a TwinVQ data processing method during medley performance. FIG. 9 shows a configuration example of conventional compression-encoded data.

Claims (5)

A header including a decoding parameter for decoding the compression-encoded data vector-quantized using the codebook, and vector-quantized data storage means for storing the compressed-encoded data in a state of being divided into predetermined frames; A decoding means for reproducing the compressed encoded data stored in the vector quantized data storage means,
The decoding means includes
Decode parameter storage means for receiving and temporarily storing the decode parameter included in the header from the vector quantized data storage means when reproducing the compressed encoded data stored in the vector quantized data storage means And receiving the compressed encoded data stored in the vector quantized data storage means from the intermediate frame in a state where the decode parameters are stored in the decode parameter storage means, and receiving the received data A decoding device for compression-encoded data, comprising: a reproduction unit that reproduces data based on the decode parameter stored in the decode parameter storage means . 2. The apparatus for decoding compression-encoded data according to claim 1, wherein the decoding parameters are a sampling rate and a bit rate. A karaoke device that uses the decoding device for compression-encoded data according to claim 1 or 2 and reproduces sequence data such as a musical sound track by a sequencer unit,
The vector quantized data storage means stores a header including decoding parameters, and acoustic sequence data that has been vector quantized and divided into frames,
The decoding means acquires a decoding parameter of a header before karaoke reproduction and stores it in the decoding parameter storage means, and based on the decoding parameter from a frame in the middle designated by a sequencer unit in acoustic sequence data at the time of karaoke reproduction A karaoke device characterized by being played back. 4. The karaoke apparatus according to claim 3, further comprising means for designating fast forward and rewind, wherein the sequencer unit designates a halfway frame in the acoustic sequence data based on an amount of fast forward and rewind. 5. The karaoke apparatus according to claim 3, further comprising means for designating a medley performance based on a plurality of music data, wherein the sequencer unit designates an intermediate frame in the second and subsequent sound sequence data during the medley performance.

Images