KR20080049632A - Playback method and apparatus, program, and recording medium - Google Patents

Abstract

A play method and a play apparatus, a program, and a recording medium are provided to output a stereo audio signal without abnormal sound regardless of the supply of necessary stereo process information. A play apparatus includes a band division unit(44), a stereo processing unit(45), a band synthesis unit(52), and a control unit. The control unit carries out a stereo process based on stereo process information obtained by necessary stereo process information. A playback method comprises the steps of: outputting stereo audio signals using the monaural audio signal if no necessary stereo process information is supplied; starting updating stereo variables within filters, and outputting the stereo audio signals using the monaural audio signal until all the state variables are updated, if necessary stereo process information is supplied; and performing the stereo process based on stereo process information acquired by the necessary stereo process information, on the monaural audio signal to generate and output stereo audio signals, if all the state variables within the filters are updated.

Description

Playback method and apparatus, program and recording medium

The present invention decodes coded audio data to which stereo peocess information is intermittently multiplexed as coded information of a monaural audio signal. A playback method and apparatus, a program and a recording medium for processing and playing back.

This specification discloses the subject matter related to Japanese Patent Documents JP 2006-324775 and JP 2007-272856, filed on November 30, 2006 and October 19, 2007, respectively, with the Japan Patent Office. Include.

BACKGROUND ART A reproducing apparatus for supplying a monaural audio signal and stereo processing information and generating a stereo audio signal by stereo processing the monaural audio signal based on the stereo processing information is known.

As described above, general stereo processing based on a monaural audio signal and stereo processing information will be described with reference to the drawings. FIG. 6 is a block diagram showing a configuration example of a general stereo processing apparatus, and FIG. 7 is a diagram showing an example of a signal supplied to the stereo processing apparatus of FIG. Stereo processing information may be multiplexed and transmitted.

In Fig. 6, a monaural audio signal is supplied to an input terminal 41, and stereo processing information is supplied to an input terminal 42, respectively. The monaural audio signal from the input terminal 41 is sent to a band divider 44 through a selector switch 43, and is band-divided, and the band-divided monaural audio signal is a stereo processor. Is sent to (45). The stereo processing unit 45 is supplied with stereo processing information from the input terminal 42, and performs stereo processing of the band-divided monaural audio signal into the left channel Lch and the right channel Rch. The stereo signals of the left and right channels Lch and Rch are sent to the left channel Lch band synthesizer 51 and the right channel Rch band synthesizer 52, respectively. The Lch audio signal from the band synthesizer 51 is sent to the conversion switch 53, and one of the Lch audio signal and the signal supplied from the conversion switch 43 through the delay unit 46 is selected to select the selected signal. Is sent to the changeover switch 54 and the output terminal 55. The Rch audio signal from the band synthesizer 52 is sent to the conversion switch 54, one of the Rch audio signal and the signal from the conversion switch 53 is selected, and the selected signal is sent to the output terminal 56. Lose.

FIG. 7 shows an example of a signal input to the stereo processing apparatus of FIG. 6. The signals are numbered # 0, # 1, # 2, ... as in units of frames or blocks, for transmission units of encoded audio data. In the figure, symbol M denotes a monaural audio signal, and symbol S denotes stereo processing information. In the example of FIG. 7, the monaural audio signal M is always transmitted, while the stereo processing information S is multiplexed and transmitted at one rate every five times. In this case, the stereo processing information (S) sent in transmission unit # 0 is used for stereo processing for the period corresponding to transmission unit # 0 to # 4, and the following stereo processing information is performed at the timing corresponding to transmission unit # 5. Switch to (S). The stereo processing information S sent at the timing corresponding to this transmission unit # 5 is used for a period corresponding to the transmission unit # 5 to # 9. Thereafter, the previously sent stereo processing information S is likewise used until the next stereo processing information S is sent.

In the configuration of Fig. 6, when stereo processing information is supplied, each conversion switch 43, 53, 54 is switched to the selectable terminal B side. That is, the monaural audio signal input from the input terminal 41 is band-divided by the band dividing section 44, and the stereo signal is produced by the stereo processing section 45 based on the stereo processing information. The generated stereo signal is band synthesized by the band synthesizers 51 and 52 of each channel, and output as stereo audio signals of the left and right channels Lch and Rch from the output terminals 55 and 56, respectively.

In contrast, in discontinuous frame playback, such as fast-forward playback based on playback by frame (transmission unit) decimating, or in playback from any frame. In some cases, multiplexed encoding information may be missing. When encoded audio data is supplied from an arbitrary frame (transmission unit) due to such discontinuous frame reproduction or the like, an absence of usable stereo processing information may occur. For example, when an input is started at a position corresponding to the transmission unit # 2 of FIG. 7, the stereo processing information S included in the transmission unit # 0 is absent due to frame decimation or the like. There is no stereo processing information available for the period up to ＃ 4.

In the apparatus of FIG. 6, in order to prevent the number of channels of the output audio signal from changing due to the presence or absence of stereo processing information, the absence of usable stereo processing information (for example, corresponds to transfer unit # 2 to # 4 in FIG. 7). The monaural audio signal is output to both stereo left and right channels. Specifically, the same monaural audio signal is output from the output terminals 55 and 56 by switching the respective conversion switches 43, 53 and 54 to the selection terminal A side. Here, when the conversion switch 43 is switched to the selection terminal A side, the monaural audio signal from the input terminal 41 is sent to the delay unit 46. This means that a delay occurs whenever the band splitting section 44 holds a state variable, updates a state variable, and performs processing, as in, for example, an FIR filtering process. In order to give a delay occurring in the band dividing section 44 to the supplied monaural audio signal. Since the band synthesis unit performs band synthesis so that a delay does not occur, the delay unit 46 considers only the delay in the band division unit 44. The monaural audio signal from the delay unit 46 is output from the Lch output terminal 55 through the conversion switch 53, and also from the Rch output terminal 56 through the conversion switch 54. In addition, when there is no stereo processing information available for the period corresponding to the transmission units # 2 to # 4 in FIG. 7, an internal state variable such as the band dividing section 44 is initialized.

Therefore, when data is supplied at a position corresponding to the transmission unit # 2 of FIG. 7, in the stereo processing apparatus of FIG. 6, the internal state variable is initialized for the period of the transmission units # 2 to # 4 described above. , Each changeover switch 43, 53, 54 is also switched to the selection terminal A side. Thereafter, when data at a position corresponding to the transmission unit # 5 is input, each of the conversion switches 43, 53, 54 is switched to the selection terminal B side, and the internal state variable is also updated. In addition, the switching operation of each conversion switch 43, 53, 54 and the processing operation of each part are controlled by the control section which is not shown in accordance with the content, internal state, etc. of input data.

Here, a specific example of a coding system in which a part of encoded information for stereo processing or the like is multiplexed and transmitted as a monaural audio signal will be described below.

For example, audio data encoded by HE AAC (High Efficiency Advanced Audio Coding, International Standard ISO / IEC 14496-3) coding scheme, in particular, HE AAC v2 (version 2) coding scheme, may be decoded ( A part of encoding information necessary for decoding is multiplexed therein and transmitted. The HE AAC v2 coding scheme is formed by combining three techniques, for example, Advanced Audio Coding (AAC) processing, Spectral Band Replication (SBR) processing, and Parametric Stereo (PS) processing. Encoded information for SBR processing and PS processing is partially multiplexed and transmitted.

The AAC process is a coding process of a speech compression method standardized by moving picture experts group (MPEG) audio. SBR processing is an encoding process that divides an input signal into a plurality of subbands and performs band extension by replicating a high sound frequency band from a lower frequency band. The PS process is a coding process that performs spatial coding using spatial information or the like necessary for generating a stereo signal from a monaural signal.

The coded audio data encoded by the HE AAC v2 scheme is coded information of an AAC core equivalent to monaural audio data encoded by the AAC encoding scheme, and the encoded information of the SBR process. And encoding information of the above-described PS processing. The encoded information of the SBR process includes encoded information (sbr header) that is multiplexed and transmitted intermittently and encoded information (sbr data) that is always transmitted. To decode sbr data, an sbr header is required. The SBR header may change its contents under certain rules, and its transmission timing depends on operational practice. The coding information (ps data) of the PS processing is included in the extended area of the sbr data (SBR data) and transmitted. Therefore, sbr header information is required for decoding PS data as well. That is, the sbr header is necessary stereo process information required for obtaining ps data for stereo processing. 8 shows an example of audio data encoded by the HE AAC v2 coding scheme. In FIG. 8, code AC represents encoding information of an AAC core, code SH represents the sbr header, and code SD represents the sbr data.

As shown in Fig. 8, in order to decode the SBR data SD and the PS data included in the extended area, an SBR header SH intermittently transmitted is required. However, in case of reproduction from any frame as described above, there may be a case where the multiplexed SBR header SH is missing. Here, if the multiplexed frames are not always monitored, especially by a higher-level system or the like, AAC to generate an output audio signal until reaching a frame from which the multiplexed SBR header (SH) can be obtained. The decoding process is performed using the encoding information AC of the core. The decode process in this case includes the AAC decode process and an up-sampling process based on the SBR process for band division and band synthesis.

When the frame including the multiplexed SBR header SH is reached, the SBR data SD or the PS data included in the extended area are decoded using the SBR header SH. Then, a "complete" decoding process (including stereo processing) is performed to generate an output stereo audio signal using this SBR data and PS data. In the decoding processing of the HE AAC v2 encoded audio data, the AAC decoding processing is performed, and then, in the SBR processing, band division and generation of a high frequency component (HF) are performed. After the stereo signal is generated from the monaural signal band-divided based on the spatial information encoded in the PS process, the output stereo audio signal is finally generated by the band synthesis process of the SBR process.

Fig. 9 is a block diagram showing an example of the configuration of a playback device for encoded audio data encoded by the HE AAC v2 system. A coded audio stream is supplied to the input terminal 11 of FIG. 9 by transmission. The encoded audio stream includes encoding information of the AAC core, HF generation encoding information (SBR data), and PS encoding information (PS data). Some of the encoded information is multiplexed and transmitted. In order to decode HF generated encoded information (SBR data) and PS encoded information (PS data), as described above, an SBR header SH which is multiplexed and transmitted is required.

In the HE AAC v2 coding scheme, when a part of the SBR header SH is different from that included in the previous frame, it is necessary to initialize the SBR process. By the initialization of the SBR process, state variables (delay signals) such as a QMF analyzer / synthesizer, a hybrid analyzer, and the like described later are initialized. Here, the state variable (delay signal) means data (signal) held in a delay element in the filter. In the filter process, a delay occurs within a period from the input to the output of the signal according to the filter length, and the state variable means this delay signal.

However, monaural audio data obtained by decoding AAC encoded information encoded by the HE AAC v2 coding scheme is upsampled by performing QMF analysis and QMF synthesis processing of the SBR process. For example, the apparatus performs SBR processing of monaural audio data after sampling AAC at a sampling rate of 24 kHz, thereby outputting audio data having a sampling rate of 48 kHz.

In FIG. 9, the encoded audio data from the input terminal 11 is sent to the payload deformatter 12, to the AAC core decoder 13, to the AAC core encoded information and the HF generated encoded information (SBR). Data) / PS encoded information (PS data). The AAC core decoding unit 13 decodes the supplied AAC core encoding information, generates a monaural audio signal of the AAC core, and sends the generated signal to the SBR processing unit 20. The parser 14 of the SBR processing unit 20 acquires multiplexed information, such as HF generation coded information from the payload deformatter 12, and checks the contents thereof, and checks the contents thereof. Judge whether processing should be initiated. When initialization is required, the parser 14 outputs an initialization control signal from the terminal 14t so that initialization of the SBR process is performed in each relevant section as will be described later. The monaural audio signal sent from the AAC core decoder 13 to the SBR processor 20 is band-divided by the QMF analyzer 21, and each band-divided signal is sent to the conversion switch 22. When the HF generation encoding information (SBR data) is supplied, the conversion switch 22 is switched for connection to the selection terminals B and C so that a signal from the QMF analyzer 21 is sent to the HF generator 23. To do that. The HF generator 23 generates an HF signal. Envelope adjuster 24 performs envelope adjustment. The resulting signal is sent to conversion switch 25.

When stereo processing information is obtained from the PS coded information (PS data), the conversion switches 22 and 25 are switched for the connection to the selection terminal C. The signal from the selection terminal C of the conversion switch 25 is sent to the hybrid analyzer 27. The hybrid analyzer 27 further divides the low frequency (LF) signal of the supplied band split signal, and divides the resulting signal into a signal de-correlator 29 and a stereo processor 30. Supplies). The signal analyzer 29 analyzes the supplied signal, makes acoustic adjustments to it, and supplies the resulting signal to the stereo processor 30. The stereo processor 30 generates stereo signals of left and right channels Lch and Rch from the supplied band split signal and the stereo processing information. With respect to the stereo signals of the generated left and right channels Lch and Rch, the hybrid synthesizing units 31 and 32 of each channel band synthesize the band-divided signals obtained by the hybrid analyzing unit 27, and the QMF synthesizing unit. Reference numerals 33 and 34 band synthesize the band-divided signals obtained by the QMF analyzer 21 to generate stereo output audio signals of left and right channels Lch and Rch. The Lch audio signal from the QMF synthesizing section 33 is sent to the conversion switch 36 and the output terminal 37. The audio signal of Rch from the QMF synthesizing section 34 is sent to the conversion switch 36, one of the signal from the QMF synthesizing section 33 and this Rch audio signal is selected, and the selected signal is output to the output terminal 38. Is sent to.

When multiplexing information such as the stereo processing information is not transmitted, the conversion switches 22, 25, 35, 36 in Fig. 9 are switched for connection to the selection terminal A or B. In order to make the sampling frequency of the output audio signal constant, only upsampling is performed using the QMF analyzer 21 and the QMF synthesizer 33. Also, in order to make the number of output channels constant, the audio signal of Lch is copied to the audio signal of Rch in order to generate the output signal.

FIG. 10 is a flowchart showing a decoding operation as described above in the configuration of FIG. 9, for example.

10, decoding (deformatting) of data encoded by the HE AAC v2 method is performed on the encoded information such as the encoded audio stream supplied to the input terminal 11, and multiplexed. As the encoded information, the HF generated encoded information and the spatial encoded information are extracted as described above. Also, for the AAC core information, AAC signal processing is performed in step S102. In the next step S103, it is determined whether or not the SBR process is performed. If YES, the process proceeds to step S104, whereas if NO, the process goes to step S114. Proceed. This process corresponds to, for example, the process performed by the payload deformatter 12 and the AAC core decoder 13 of FIG.

In step S104, for example, the QMF analyzer 21 performs QMF band division processing. In the next step S105, it is determined whether or not the multiplexed encoded information has already been decoded, and if yes (YES), the processing proceeds to step S106, while if no (NO), the processing proceeds to step S113. Proceeds to). In step S106, for example, the HF generation unit 23 performs an HF signal generation process using the multiplexed HF generation encoding information (already decoded information), and then, the next step ( In S107, it is determined whether or not the PS process is performed.

If it is determined in step S107 that YES (PS processing is performed), then control proceeds to step S108, where a hybrid analysis process is performed. Then, in step S109, a stereo signal generation process is performed based on the spatial information, and then in step S110, a hybrid synthesis process is performed. Thereafter, the control proceeds to step S111. Such a process corresponds to, for example, a process performed by the hybrid analyzer 27 of FIG. 9 to a process performed by the hybrid synthesis units 31 and 32. In step S107, if NO is determined (NO PS processing is not performed), control proceeds to step S111.

In step S111, QMF band synthesis processing of Lch is performed, and in step S112, QMF band synthesis processing of Rch is performed, and the resultant audio signal is output. Further, in step S113, QMF band synthesis processing of Lch is performed, and in step S114, a monaural signal is duplicated as necessary to generate a stereo signal, and as a result, an audio signal is output. Such a process corresponds to, for example, a process performed by the QMF synthesizing units 33 and 34 through the conversion switches 22, 35, and 36 shown in FIG.

As a prior art, Published translation of International Patent Application (KHOYO) No. 2004-535145 (Reference Patent 1) and Japanese Patent Application Publication (KOKAI) No. 2006-085183 (Reference) Patent 2) discloses a technique for generating a stereo audio signal by stereo processing a monaural audio signal based on stereo processing information. ISO / IEC 14496-3: 2005, Information technology-Encoding of audio-visual objects- Part 3: Audio (Information Technology-Coding of audio-visual objects-Part 3: Audio) (Non-Patent Document 1) discloses a specification of the HE AAC (High Efficiency Advanced Audio Coding) coding scheme.

By the way, for example, when the internal state variable is initialized in playback from an arbitrary frame by playback of discontinuous frames such as the above-described frame decimation playback, thereafter, partially multiplexed encoding such as stereo processing information is then performed. When information is supplied, the update of the state variable is started. Therefore, an abnormal sound is generated due to the influence of the filter delay.

For example, in the configuration of FIG. 6, the input starts from a position corresponding to the transmission unit # 2 of FIG. 7 and is usable for a period corresponding to the transmission unit # 2 to # 4. When stereo processing information is included in the transmission unit # 5 from the state where there is no processing information, each conversion switch 43, 53, 56 is switched to the selection terminal B side. The band divider 44 generates a band-division signal for the first time after these switches are switched to the selection terminal B side. At this time, since the state variable of the band dividing unit 44 is the initial state, the influence of this state is affected on the output corresponding to the transmission unit # 5. For example, the influence may include damping of the output signal or the like, thereby causing an irregular sound.

In the case of the configuration of FIG. 9, when the frames are reproduced discontinuously, such as when fast-forwarding reproduction of audio data encoded by the HE AAC v2 method is performed by frame decimation, the multiplexing is performed. Sometimes the sbr header is missing. For example, in the case of the example of Fig. 8, when playback starts from frame (transmission unit) # 1, the SBR header SH is first transmitted at a timing corresponding to frame # 5. In this case, the SBR encoding information and the PS encoding information of the SBR data SD cannot be decoded until the frame in which the SBR header SH is obtained can be decoded, so that the conversion switch 22 selects the selection terminal A. ), The changeover switch 35 is connected to the selection terminal A side, and the changeover switch 36 is connected to the selection terminal B side, respectively. Therefore, the monaural audio signal of the AAC core is upsampled using the SBR QMF analyzer 21 and the Lch QMF synthesizer 33, and an identical output audio signal is generated in the stereo left and right channels.

When the frames are discontinuously reproduced in this manner, the state variables (delayed signals) of the filters in the reproducing apparatus and the input audio data encoded by the HE AAC v2 coding scheme become discontinuous. Therefore, the playback apparatus needs to be initialized (including SBR processing initialization) to initialize these internal state variables. The state variables (delay signals) in these reproducing apparatuses include the state variables of the QMF analyzing section 21, the QMF synthesizing sections 33 and 34, and the hybrid analyzing section 27, and these state variables are set to zero at initialization. do. Since the SBR encoded information / PS encoded information cannot be decoded until the SBR header SH is transmitted, the playback apparatus switches its conversion switches 22, 35, and 36 to their selection terminal A side, so that the AAC core decoder The monaural audio signal from (13) is upsampled through the processing by the QMF analyzer 21 and the Lch QMF synthesizer 33, and outputs the resulting output audio signal to the stereo left and right channels. When the SBR header SH is transmitted, the SBR encoding information and the PS encoding information are first decoded after initialization of the playback apparatus, and the SBR processing and the PS processing are executed. Since the QMF analyzer 21 and the Lch QMF synthesizer 33 perform their upsampling process even before the SBR header SH is transmitted, their state variables are continuously updated. On the other hand, the state variables of each hybrid analyzer 27 and Rch QMF synthesizer 34 is in an initialized state. This state affects subsequent processing, whereby an irregular sound is generated in the output audio signal. 11A and 11B show examples of stereo output audio signals of left and right channels Lch and Rch at this processing point.

11A and 11B show a state in which no multiplexed encoded information (stereo information, etc.) is available, for example, only AAC-LC (Low Complexity) encoded information signals are supplied, and only upsampling processing is performed in the SBR process. From the state, the multiplexed encoded information including stereo processing information at time t1 becomes valid (usable), indicating a state in which AAC processing, SBR processing, and PS processing are started. Fig. 11A shows the Lch output audio signal, while Fig. 11B shows the Rch output audio signal, respectively.

11A and 11B, at time t1, the playback apparatus recognizes encoded information multiplexed for the first time after initialization of the internal state variable. However, since the state variable changes from the initial state, the influence of the state variable of the band synthesizing unit (Rch QMF synthesizing unit 34) for the SBR processing between time t1 and t2 affects the output audio signal of Rch. On the other hand, the influence of the state variables of the hybrid filter (hybrid analyzer 27) for PS processing between the times t2 and t3 affects both audio signals of Lch and Rch. As a result, irregular sounds are generated in the output audio signal.

In order to avoid the disadvantage, it is considered to always monitor the multiplexed encoded information. In this case, the multiplexed information is transmitted simultaneously with the normal encoded information. Therefore, all encoded information needs to be decoded, which hinders a decrease in throughput.

In view of the above situation, when reproduction is performed from an arbitrary position due to the intermittent transmission of multiplexed encoding information and information necessary for decoding (such as an SBR header), necessary encoding information is supplied from an internal state variable initialized state. It is desirable to provide a reproducing apparatus and method, a program, and a recording medium which can effectively prevent negative influence (irregular sound generation, etc.) on an output audio signal due to filter delay occurring when the filter is generated.

In one embodiment of the present invention, when the necessary stereo processing information required for stereo processing is intermittently multiplexed into encoded information of a monaural audio signal and decoded and reproduced encoded audio data transmitted, the required stereo processing information is not supplied. Outputting stereo audio signal using monaural audio signal, and when necessary stereo processing information is input, update of state variable in filter is started, and stereo audio signal is output using monaural audio signal until all state variables are updated, If all the state variables in the filter have been updated, stereo processing is performed on the monaural audio signal based on the stereo processing information obtained by the necessary stereo processing information to generate and output a stereo audio signal.

In this case, the stereo processing is preferably performed on the monaural audio signal extended in the band.

When the necessary stereo processing information is not supplied, the monaural audio signal is divided into at least two sub-bands by band division filtering processing, and the monaural divided by band synthesis filtering processing. The audio signal is upsampled and a stereo audio signal is output using the monaural audio signal. When the necessary stereo processing information is supplied, it is preferable to process the state variable in the filter for the monaural audio signal as the filtering state variable for the stereo audio signal.

The encoded audio data may include encoding information of an AAC core corresponding to monaural audio data based on HE AAC (High Efficiency Advanced Audio Coding) coding scheme, encoding information for SBR (Spectral Band Replication) processing, and a parametric (PS). Stereo) has encoding information for processing. The encoded information for SBR processing includes an SBR header, which is multiplexed and intermittently transmitted information, and SBR data, which is always transmitted encoded information. PS data, which is encoding information for the PS processing, is included in the extended area of the SBR data and transmitted. The SBR header is the necessary stereo processing information required for decoding the SBR data.

According to the embodiment of the present invention, even from when the necessary stereo processing information is not supplied to when the necessary stereo processing information is supplied, a good stereo audio signal free from irregular sound generation can be reproduced.

Other features and aspects of the present invention are described in detail in the following detailed description with reference to the accompanying drawings.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram showing a schematic configuration example of a stereo processing apparatus used in a playback apparatus or a playback method according to an embodiment of the present invention. In Fig. 1, the parts corresponding to Fig. 6 described above are denoted by the same reference numerals.

A monaural audio signal is supplied to the input terminal 41 of FIG. 1, and stereo processing information is supplied to the input terminal 42 of FIG. The monaural audio signal from the input terminal 41 is sent to a switch 43X and a delay section 46. The monaural audio signal from the switch 43X is sent to the band divider 44 and band-divided so that the band-divided monaural audio signal is sent to the stereo processor 45. The stereo processing unit 45 is supplied with stereo processing information from the input terminal 42, and performs stereo processing of the band-divided monaural audio signal into left channel Lch and right channel Rch stereo signals. Then, of the resulting Lch and Rch stereo signals, the Lch signal is sent to the band synthesizer 51 through the switch 61, and the Rch signal is passed through the switch 62 to the band synthesizer 52. Is sent). The Lch audio signal from the band synthesis section 51 is sent to the conversion switch 53X, and one of the Lch audio signal and the signal supplied through the delay section 46 is selected, and the selected signal is converted to the conversion switch 54X. And output terminal 55. The Rch audio signal from the band synthesizer 52 is sent to the conversion switch 54X, and one of the Rch audio signal and the signal supplied from the conversion switch 53X is selected, and the selected signal is sent to the output terminal 56. Is sent. Here, the switching operation of each of the conversion switches 43X, 53X, 54X, the on / off operation of the switches 61, 62, and the processing operation of each part are performed by the control unit (not shown), and the contents of the input data, It is controlled according to the internal state or the like.

When the input signal (monaural audio signal M and the intermittent stereo processing information S) as shown in FIG. 7 is supplied to the stereo processing apparatus as shown in FIG. 1, it is included in the transmission unit # 0 and sent. The stereo processing information S is used for stereo processing for a period corresponding to the transmission unit # 0 to # 4, and is then switched to the next stereo processing information S at the timing corresponding to the transmission unit # 5. This stereo processing information S sent at the timing corresponding to transmission unit # 5 is used for a period corresponding to # 9 in transmission unit # 5 as already mentioned.

When this usable stereo processing information is available, the switch 43X is connected to the selection terminal B, and the switches 61 and 62 are connected to the selection terminal C, and the conversion switch ( 53X and 54X are switched to be connected to the selection terminal C. FIG. In this state, the monaural audio signal input from the input terminal 41 is band-divided by the band dividing section 44, and a stereo signal is generated by the stereo processing section 45 based on the stereo processing information. The generated stereo signal is then band synthesized by the band synthesis processing sections 51 and 52 of each channel, and the resulting Lch and Rch stereo audio signals are output from the output terminals 55 and 56, respectively.

On the other hand, when encoded audio data is supplied from an arbitrary frame (transmission unit) due to discontinuous frame reproduction such as fast-forward reproduction or the like, a state in which no stereo processing information is available may occur. For example, when the input starts from the position corresponding to the transmission unit # 2 of FIG. 7, the stereo processing information S included in the transmission unit # 0 is not supplied due to frame decimation, and as a result, the transmission unit # There is no stereo processing information available for the period 2 to 4. In the stereo processing apparatus of FIG. 1, for example, the internal state variables such as the band dividing unit 44 are initialized during the period corresponding to the transmission units # 2 to # 4 where no stereo processing information is available. In addition, each conversion switch 53X, 54X is connected to the selection terminal A. As shown in FIG. Accordingly, the monaural audio signal supplied from the input terminal 41 through the delay unit 46 is output from the Lch output terminal 55 through the conversion switch 53X, and also through the conversion switch 54X. It is also output from 56). This configuration prevents the channel number of the output audio signal from changing due to the presence or absence of stereo processing information. Here, the delay unit 46 is provided in consideration of the delay caused by the band division unit 44, for example, generated by the FIR filtering process or the like.

Thereafter, when data is supplied at the position corresponding to the transmission unit # 5 in FIG. 7 and the usable stereo processing information S is also supplied, first, the switch 43X is connected to the selection terminal B, whereby the bandwidth is divided. The monaural audio signal is supplied to the installment 44. However, the switches 61 and 62 and the conversion switches 53X and 54X are not connected to their selection terminal C until the state variable of this band dividing section 44 is completely updated. For this reason, when stereo processing information is supplied for the first time since no stereo processing information is available and the internal state variable is initialized, the switch 43X is connected to the selection terminal B, whereby the state variable of the band splitting section 44 is provided. The monaural audio signal supplied through the delay section 46 is output from the respective output terminals 55 and 56 through the selection terminal A of each of the conversion switches 53X and 54X while updating the. After that, when the state variable of the band dividing section 44 is completely updated, the switches 61 and 62 are connected to the selection terminal C, and the conversion switches 53X and 54X are connected to the selection terminal C. The signals which have been switched and stereo processed as described above are output from the output terminals 55 and 56 as output audio signals, respectively. Therefore, the output audio signal is free from the influence of the state in which the state variable of the band dividing unit 44 is initialized, and thus an audio signal in which irregular sound generation is prevented can be obtained.

That is, in the embodiment of the present invention, when the encoded audio data transmitted by intermittently multiplexing the stereo processing information into the encoded information of the monaural audio signal is decoded and reproduced, the monaural audio signal is not supplied. And output stereo audio signal using stereo processing information, if the stereo processing information is supplied, to start updating the state variable in the filter and output the stereo audio signal using the monaural audio signal until all state variables are updated. It is composed. Thereafter, when all the state variables in the filter are updated, stereo processing is performed on the monaural audio signal based on the stereo processing information to generate and output a stereo audio signal.

Next, the present invention reproduces the encoded audio data encoded by the above-described HE AAC (High Efficiency Advanced Audio Coding, ISO / IEC 14496-3) encoding scheme, in particular, HE AAC v2 (version 2) encoding scheme. A configuration example of a playback apparatus to which the embodiment is applied will be described with reference to FIG. In FIG. 2, parts corresponding to the above-described elements of FIG. 9 are denoted by the same reference numerals.

The coded audio stream is supplied to the input terminal 11 of FIG. 2 by transmission. The encoded audio stream includes encoding information of the AAC core, HF generation encoding information (band extension encoding information for SBR processing), and PS encoding information (spatial information for stereo processing). Some encoded information is multiplexed and transmitted. That is, as described with reference to FIG. 8, the encoded information SD (SBR data) for the SBR processing is always multiplexed with the encoded information AC of the AAC core, while decoding the SBR data SD. The SBR header SH required for this purpose is intermittently multiplexed with the encoding information AC. PS data for the PS processing is included in the extended area of the SBR data SD and transmitted. Since the SBR header SH is also required to obtain PS data, this SBR header SH is necessary stereo processing information.

When the HF generation encoding information (SBR data) and the PS encoding information (PS data) are included, the audio signal decoded by the AAC core decoder 13 has a sampling rate of half of the final output audio signal. Output at (half sampling rate). Therefore, by combining the QMF analyzer 21 and the QMF synthesizers 33 and 34, the audio signal is upsampled. For example, when the sampling frequency of the output signal from the AAC core decoder 13 is 24 kHz, the sampling frequency of the output audio signal from the QMF synthesizing units 33 and 34 is 48 kHz.

The encoded audio data from the input terminal 11 is a bitstream payload deformatter, i.e., a pay separated by AAC core encoding information to the AAC core decoder 13 and HF generation encoding information / PS encoding information. It is sent to the load deformatter 12.

The HF generation encoding information / PS encoding information is sent to the SBR processing unit 20, and then through the bitstream parser, that is, the parser 14 of the SBR processing unit 20, Huffman decoding / dequantization unit (Huffman). decoder / dequantizer). The HFman decoding / dequantization unit 15 extracts HF signal generation information, envelope adjustment information, and stereo processing information. The former two items of potentials of the extracted information are sent to the HF generator 23 and the envelope adjuster 24, respectively, while the latter one item is It is sent to the stereo processor 30 through the Lch replication process judgment section 16. The parser 14 of the SBR processing unit 20 obtains multiplexed information, such as HF generation encoding information, from the payload deformatter 12, checks the contents thereof, determines whether initialization of the SBR processing is necessary, and initiates initialization. If necessary, the initialization control signal is output from the terminal 14t so that the initialization of the SBR process is performed for each unit as described later. In addition, the Lch duplication processing judging unit 16 determines the multiplexed encoded information obtained after the SBR process initialization for the first time, outputs a determination output from the terminal 16t, and the Lch QMF synthesizing unit 34 described later in Lch QMF synthesis. The process of copying the state variable (delay signal) of the unit 33 is performed.

The AAC core decoder 13 decodes the supplied AAC core encoding information and generates a monaural audio signal of the AAC core. The decoder 13 sends the generated monaural audio signal to the QMF analyzer 21 of the SBR processor 20. The QMF analyzer 21 divides the monaural audio signal into 64 bands and, as a result, sends the band-divided signal to the conversion switch 22X. When the HF generation encoding information (SBR data) is supplied, the conversion switch 22X is switched to be connected to the selection terminals B and C, so that the signal from the QMF analyzer 21 is converted to the HF generator (HF generator). 23). The HF generator 23 generates an HF signal, and the envelope adjuster 24 performs envelope adjustment. The envelope adjusting unit 24 sends the resulting signal to the hybrid analyzer 27 and the conversion switch 35X.

When stereo processing information is obtained from the PS encoded information (PS data), the conversion switch 22X is switched to be connected to the selection terminal C. The hybrid analyzer 27 further divides the LF signal of the supplied band-divided signal, and the signal de-correlator 29 together with the resulting signal along with the HF signal of the previous band-divided signal. And a stereo processor 30. The signal analyzing unit 29 analyzes the supplied signal, makes acoustic adjustment thereto, and supplies the resulting signal to the stereo processing unit 30. The stereo processor 30 generates Lch and Rch stereo signals from the supplied band-divided signal and stereo processing information. The generated stereo signal of each channel is sent to the hybrid snthesizers 31 and 32 of each channel through the switches 17 and 18, respectively. The hybrid synthesis units 31 and 32 band synthesize the bands obtained by the hybrid analyzer 27 described above. The resultant signal from the hybrid synthesis section 31 is sent to the QMF synthesis section 33 and the conversion switch 19 through the conversion switch 35X, while the result signal from the hybrid synthesis section 32 is used. Is sent to the QMF synthesizing section 34 via the conversion switch 19. The QMF synthesizing units 33 and 34 of each channel band synthesize the band-divided bands obtained by the QMF analyzing unit 21 to generate Lch and Rch stereo output audio signals, respectively. The Lch audio signal from the QMF synthesizing section 33 is sent to a conversion switch 36X and an output terminal 37. The Rch audio signal from the QMF synthesizing section 34 is sent to the conversion switch 36X, and one of the Rch audio signal and the signal from the QMF synthesizing section 33 is selected, and the selected signal is sent to the output terminal 38. Is sent.

Here, the operation of each unit including the switching of the playback device in Fig. 2 is controlled by control means (not shown) according to the contents of the input encoding information, the state of each unit, and the like.

Compared with the configuration of the playback apparatus shown in FIG. 9 described above, the playback apparatus shown in FIG. 2 differs in the following points. The switching configuration downstream of the QMF analyzer 21 and the switching configuration downstream of the envelope adjustment unit 24 are modified. Switches 17 and 18 and conversion switch 19 were added. The state variable of one of the QMF synthesizing units 33 and 34 is duplicated to the other.

In the playback device of FIG. 2, a case where encoded audio data is supplied from an arbitrary frame (transmission unit) as described above will be described. For example, when the input starts at the position corresponding to the transmission unit # 2 of FIG. 8, the SBR header SH, which is necessary stereo processing information included in the transmission unit # 0, is not supplied. Therefore, the device cannot decode the SBR data SD while receiving the transmission units # 2 through # 4, and thus cannot obtain usable stereo processing information (PS data). Therefore, the apparatus initializes internal state variables (delay signals) such as the QMF analyzer 21, the hybrid analyzer 27, and the QMF synthesizers 33 and 34 of the SBR processor 20. Next, when data is supplied at the position corresponding to the transmission unit # 5 of FIG. 8 and the SBR header SH which is necessary stereo processing information is supplied accordingly, the device can decode the SBR data SD. Obtain possible stereo processing information (PS data). As a result, the device updates the internal state variables (delay signals) of the QMF analyzing unit 21, the hybrid analyzing unit 27, the QMF synthesizing units 33 and 34 of the SBR processing unit 20, and the like. These state variables (delay signals) mean data (signals) held by delay elements in the filter, respectively. In the filtering process, a delay occurs during the period from the input to the output of the signal according to the filtering length, and the state variable means such a delay signal.

Here, in the state where the usable stereo processing information (PS data) cannot be obtained and the internal state variable is initialized, the conversion switches 22X, 35X, 36X are switched to be connected to the selection terminal A side. In this state, the QMF analyzer 21 band-segments the monaural audio signal from the AAC core decoder 13, and the LMF QMF synthesizer 33 band-synthesizes the band-divided signal so that the left and right channels are the same audio. Output the signal.

Then, when the multiplexed encoded information is transmitted, the conversion switches 22X, 35X, 19, 36X are switched to be connected to the selection terminals B, C. In this case, the selection terminal B is selected when the encoding information includes only the extended encoding information. On the other hand, the selection terminal C includes the extension information (HF generation information) and the stereo processing information. When it is selected.

The following describes a case in which the SBR header SH which is necessary stereo processing information is transmitted, whereby the playback apparatus decodes the SBR data SD to obtain stereo processing information (PS data). When encoding information (SBR data) and stereo processing information (PS data) for SBR processing are obtained, the apparatus is ready to send a signal to the QMF synthesis section 34 of Rch for the first time. Therefore, if an output audio signal is generated without considering the state variable (delay signal), the device outputs the state variable initialization signal to the Rch audio signal, thereby generating an irregular sound. In view of this, in the embodiment of the present invention, the determination output from the Lch duplication processing judging unit 16 outputs the state variable (delay signal) of the QMF synthesizing unit 33 of the Lch at this timing in the state variable duplication process. It is used to replicate to the QMF synthesis section 34 of. Through this operation, the state variable of the QMF synthesizing unit 33 of Lch, regardless of whether the playback device is playing the encoded audio signal with the conversion switches connected to the selection terminal A until the stereo processing information is transmitted. The same state variable as is set in the QMF combining section 34 of Rch. When the above duplication process is executed, each conversion switch 22X, 35X, 19, 36X is switched to be connected to the selection terminal F. As shown in FIG.

In general, if an irrelevant, arbitrary signal is used as the delay signal during the band synthesis process, unexpected amplification / damping occurs during the band synthesis process, Thereby an irregular sound is produced. In the method according to the present invention, since the frame from which the multiplexed encoded information is obtained for the first time after initialization indicates a switching point from the monaural output to the stereo output, the state variable (delayed signal) of the QMF synthesizer 33 of Lch is delayed. Is used as a state variable (delay signal) of the QMF synthesis section 34 of Rch, so that irregular sounds are not generated.

In addition, in stereo processing (PS processing), in order to apply spatial coded information, the reproduction apparatus is band-split by the hybrid analyzer 27, analysis results from the signal analyzer 29, and transmitted. Stereo signal generation and hybrid synthesis are performed based on the spatial information. Since the hybrid analyzer 27 requests a delay and performs the processing for the first time after the multiplexed encoded information is decoded, the state variable (delayed signal) is obtained when the multiplexed encoded information is obtained for the first time after initialization of the variable in the decoder. ) Is initialized, which affects the signal analysis by the signal analyzing section 29, whereby an irregular sound is generated. That is, the signal band-divided by the QMF analyzer 21 is supplied to the hybrid analyzer 27, and the state variable (delay signal) of the hybrid analyzer 27 is initialized, so that subsequent processing is not performed correctly. .

In view of this, in the present embodiment, in order to eliminate this effect, when the hybrid analyzer 27 performs the processing for the first time after initialization, the playback apparatus performs the hybrid analyzer 27 and to update the delay signal. The Lch and Rch stereo signal generation coefficients of the stereo processing unit 30 are updated. With respect to the output, the switches 35X and 19 are switched to the selection terminals F, respectively, and the signal is branched before the hybrid analyzer 27 is output to the QMF synthesizers 33 and 34 of each channel.

Specifically, the stereo signal is disconnected (switches 17 and 18 are all turned off) by the switches 17 and 18 until the state variable (delay signal) of the hybrid analyzer 27 is completely updated. . Instead, the signal transmitted through the selection terminal F of each of the conversion switches 22X and 35X is transmitted through the QMF synthesis section 33 of the Lch and the QMF synthesis section of the Rch through the selection terminal F of the conversion switch 19. Is sent to 34. The result signal from the QMF synthesizing section 33 of Lch is output from the output terminal 37, while the result signal from the QMF synthesizing section 34 of Rch having the same state variable as the QMF synthesizing section 33 of Lch. Is output from the output terminal 38 via the selection terminal F of the conversion switch 36X.

The state variable (delay signal) of the hybrid analyzer 27 has a delay by 6 QMF samples, as specified in Section 8.6.4 of Non-Patent Document 1 above. The update processing of the Lch and Rch stereo signal generation coefficients of the stereo processing unit 30 needs to be performed since the coefficients are transmitted as difference information as described in Section 8.6.4.4 of Non-Patent Document 1 above. There is.

When the state variable (delay signal) of the hybrid analyzer 27 is completely updated, the switches 17 and 18 are both turned on (connected to the selection terminal E), and the Lch and Rch stereo from the stereo processor 30 are The signals are sent to the hybrid combining sections 31 and 32, respectively. The conversion switches 35X, 19, 36X are respectively switched to be connected to the selection terminal E side, so that the signal from the hybrid synthesis section 31 is processed in the QMF synthesis section 33, and the resulting signal is an Lch stereo audio signal. The signal from the hybrid synthesizing section 32 is processed by the QMF synthesizing section 34, and the resultant signal is output from the output terminal 38 as an Rch stereo audio signal. Here, the reproducing apparatus updates the state variables of the QMF synthesizing unit 34 of the Rch while the state variables of the hybrid analyzer 27 are updated, thereby switching the switches 17 and 18 and the conversion switches 35X, 19 and 36X. It is possible to connect to the selection terminal E, whereby the playback apparatus can switch these switches without generating irregular sounds in the processing of a single frame.

3 to 5 are flowcharts for explaining the decoding operation as described above in the configuration of FIG. 2, for example.

In Fig. 3, the encoded information such as the encoded audio stream supplied to the input terminal 11 is HF generated encoded information (SBR data) and spatial encoding as described above as the encoded information multiplexed in step S101. In order to extract information (PS data), decoding processing (deformatting processing) of data encoded by the HE AAC v2 scheme is performed. Also, for the AAC core information, AAC signal processing is performed in step S102. In the next step S103, it is determined whether or not the above-described SBR processing is performed. If YES, processing proceeds to step S104, while if NO, the processing proceeds to step S114. This processing corresponds to, for example, the processing performed by the payload deformatter 12 and the AAC core decoder 13 of FIG.

In step S104, for example, the QMF analyzer 21 performs QMF band division processing. In the next step S105, it is determined whether or not the multiplexed encoded information has already been decoded. If YES, the process proceeds to step S106, and if NO, the process proceeds to step S113. In step S106, for example, the HF signal generation process is performed by the HF generation section 23 using the multiplexed HF signal generation encoding information (already decoded information). In the next step S107, it is determined whether or not the PS processing is performed.

If it is determined in step S107 that YES (PS processing is performed), the processing proceeds to step S111 after the PS processing is performed in step S120, while NO (PS processing is performed in step S107). Is not performed), the process proceeds directly to step S111. Hereinafter, the specific example of PS process in step S120 is demonstrated with reference to FIG. 4 or FIG.

In step S111, QMF band synthesis processing of Lch is performed, and in step S112, QMF band synthesis processing of Rch is performed. Thereafter, the resulting audio signal is output. Further, in step S113, QMF band synthesis processing of Lch is performed, and in step S114, a monaural signal is duplicated to generate a stereo signal as necessary. Thereafter, the resulting audio signal is output. Such processing corresponds to, for example, processing performed by the QMF synthesizing units 33 and 34 via the conversion switches 35X and 36X in FIG.

4 shows a specific example of the PS processing in step S120 described above in the embodiment of the present invention. If it is determined in step S107 of FIG. 3 described above that YES (PS processing is performed), the process proceeds to step S108, where a hybrid analysis process is performed, and in step S109, stereo based on spatial information Signal generation processing is performed. Thereafter, after the hybrid synthesis process is performed in step S110, the control proceeds to step S115. In step S115, it is determined whether or not the state variable (delay signal) for the QMF band synthesis process of Rch, for example, the state variable of the QMF synthesizing section 34 in Fig. 2 has already been updated. If YES, the processing proceeds to step S111 of FIG. 3 described above, and if NO, the processing proceeds to step S116. In step S116, the state variable for the QMF band synthesis process of Lch is copied to the state variable for the QMF band synthesis process of Rch, and then control proceeds to step S111 of FIG. Such a process corresponds to, for example, a process from the process performed by the hybrid analyzer 27 of FIG. 2 to the process performed by the QMF synthesis units 33 and 34.

In the specific example shown in Figs. 3 and 4, when a part of encoded information is reproduced from any frame of encoded audio data transmitted by multiplexing, the internal state of the playback apparatus is configured to be initialized and multiplexed. Even in the absence of the encoded information to be transmitted, the monaural audio signal is divided into at least two or more subbands, and the resultant signal is upsampled by a band synthesis filtering process in which delay occurs to output the monaural audio signal. Then, when the multiplexed encoding information is supplied and the process of first generating the stereo signal from the monaural signal is performed, the filtering state variable (delay signal) for the monaural signal is processed as the filtering state variable for the stereo signal (step S114). , S115, S116)), which is configured to prevent the occurrence of irregular sound due to the delay caused by the filtering process.

Next, FIG. 5 shows another specific example of the PS processing in step S120 of FIG. 3 described above in the embodiment of the present invention. That is, if it is determined in step S107 of FIG. 3 described above that YES (PS processing is performed), the process proceeds to step S108 of FIG. 5, and the hybrid analysis process (for example, the hybrid analysis of FIG. 2). Processing by the unit 27). The process then proceeds to step S119, where it is determined whether or not all the state variables (delay signals) for the hybrid analysis process have already been updated. If YES, the process proceeds to step S109 and if NO, the process proceeds to step S117. In step S109, a stereo signal generation process based on spatial information is performed, and in step S110, a hybrid synthesis process is performed. The process then proceeds to step S115. In step S117, since all of the state variables for the hybrid analysis process have not yet been updated, the monaural signal is duplicated to generate a stereo signal, and the generated stereo signal is used as the output of the hybrid synthesis process. Thereafter, the control proceeds to step S118, where necessary state variables are updated, and then processing proceeds to step S115.

In step S115, it is determined whether or not the state variable (for example, the state variable of the QMF synthesizing section 34 in Fig. 2) for the QMF band synthesis process of Rch has been updated. If YES, the process proceeds to step S111 of FIG. 3 described above, and if NO, the process proceeds to step S116. In step S116, the state variable for the QMF band synthesis process of Lch is copied to the state variable for the QMF band synthesis process of Rch, and then control proceeds to step S111 of FIG.

3 and 5, in addition to the configuration of the specific example described with reference to FIG. 4 above, as shown in steps S119, S117 and S118, the delay in the filtering process affects the output audio signal. In order not to give an error, the updating process of the filtering state variable and the duplication process of the output signal are performed until at least all the filtering state variables (delay signals) are updated. Then, after all the filtering state variables are completely updated, normal playback processing is performed, whereby occurrence of irregular sounds of the output audio signal due to delay in the filtering processing is prevented.

11C and 11D show examples of Lch and Rch stereo output audio signals in the embodiment of the present invention. The above description with reference to FIGS. 10A and 10B is similarly applied to the times t1 to t3. That is, there is no stereo processing information available until time t1 (for example, only the AAC-LC (Low Complexity) encoded information signal is input and only up sampling is performed in the SBR process). . At time t1, multiplexed encoded information including stereo processing information becomes valid (usable), whereby AAC processing, SBR processing, and PS processing are started. FIG. 11C shows an Lch output audio signal, and FIG. 11D shows an Rch output audio signal.

In the embodiment of the present invention, the Lch and Rch stereo output audio signals shown in Figs. 11C and 11D are between time t1 to t2, as is apparent from the comparison with the conventional output audio signals shown in Figs. 11A and 11B. Hybrid filter (Hybrid analysis unit 27) for the PS processing between the time (t2 ~ t3) and the effect of the state variable (delayed signal) of the band synthesis unit (QMF synthesis unit 34) for the SBR processing Free from influence by state variables According to the embodiment of the present invention, even when the encoded information (stereo processing information, etc.) multiplexed for the first time from the state in which the internal state variable is initialized, a good stereo audio signal free from irregular sounds or the like can be reproduced.

According to the embodiment of the present invention described above, in the decoding processing and reproduction of encoded audio data in which a part of encoded information including stereo processing information is multiplexed into a monaural audio signal and transmitted, the multiplexed encoded information that is usable is not supplied. The internal state variable (delay signal) is initialized in the non-state state, and the stereo audio signal is output using the monaural audio signal. When the multiplexed encoding information is supplied with the internal state variable initialized, updating of the internal state variable is started, and the stereo audio signal is output using the monaural audio signal until all the state variables are updated. And when the internal state variables are all updated, perform signal processing including stereo processing based on the multiplexed encoding information on the monaural audio signal to generate and output a stereo audio signal.

In other words, in decoding and reproducing encoded audio data in which a part of encoded information including stereo processing information is intermittently multiplexed into a monaural audio signal and transmitted, if stereo processing information is not supplied, stereo audio is used using a monaural audio signal. Configured to output a signal. When the stereo processing information is supplied, it is configured to start updating the internal state variable in the filter and output the stereo audio signal using the monaural audio signal until all the state variables are updated. Once all the state variables in the filter have been updated, it is configured to perform stereo processing based on the stereo processing information on the monaural audio signal to generate and output the stereo audio signal.

In another embodiment of the present invention, an apparatus for reproducing encoded audio data is provided. The reproducing apparatus includes decoding means, information acquisition means, audio signal band division means, high frequency information generation means, stereo signal generation means ( stereo information generation means, sub-band divided signal synthesis means and output audio signal generation means. The decoding means decodes the encoded audio data transmitted by multiplexing a part of the encoded information. The information acquiring means acquires information for generating an output audio signal from a part of the transmitted encoded information even when a part of the multiplexed encoded information is not transmitted. The audio signal band dividing means divides into at least two subbands to generate a band-divided signal. The high frequency information generating means generates high frequency information with respect to the generated band split signal when the band extension coded information is transmitted. The stereo signal generating means causes the sub-band divided signal generation means for requesting the delay to generate the subband split signal for the band split signal, and when the spatial encoded information is transmitted, Generate a stereo signal from the monaural signal. The subband split signal synthesizing means synthesizes the subband split signal into a band split signal. The output audio signal generating means causes audio signal synthesis means, which requires a delay, to synthesize the synthesized band split signal to generate the output audio signal. In the playback apparatus, when playing back from discrete positions (frames), subband signal generation means, state variable initialization means, playback continuing means, and monaural signals Monaural signal state variable utilization means is installed. The subband signal generating means requires a delay of the encoded audio data reproducing apparatus. The state variable initialization means initializes the state variable (delay signal) of the audio signal synthesizing means. The reproducing continuation means continues the regeneration after the initialization. When the monaural signal state variable utilization means is first transmitted after the initialization, the encoded information is decoded and the stereo signal is generated from the monaural signal. The state variable of the audio signal synthesizing means for the generated stereo signal ( Processing is performed using the state variable (delay signal) for the monaural signal as a delay signal).

In addition, pseudo-subband-divided signal generation means, replication and output means, updating means and stereo signal generation performing means Is also installed. The pseudo subband split signal generating means decodes the spatially encoded information when the multiplexed encoded information is transmitted for the first time after initialization of the delayed signal of the encoded audio data reproducing apparatus, and generates a stereo signal from the monaural signal. Subband division signal generation is performed in a pseudo manner until all state variables (delay signals) of the generating means are updated. The copying and outputting means duplicates the monaural band splitting signal supplied to the subband splitting signal generating means during the period in which the pseudo subband splitting signal generating means is pseudoly operated, and outputs the stereo band splitting signal to the audio signal synthesizing means. . The division coefficient updating means is a division which is updated by the difference of the stereo signal generating means for generating the stereo signal from the monaural signal during the period in which the pseudo subband division signal generating means is operating pseudo. Update the coefficients. The stereo signal generating means generates a stereo signal from the monaural signal based on the spatial encoding information after all delay signals of the subband split signal generating means are updated.

That is, in performing normal reproduction from an arbitrary frame by decoding processing on coded audio data in which a part of encoded information is multiplexed and transmitted, even if there is no encoded information transmitted by multiplexing the delay signal of the decoder, Splitting into at least two subbands and performing upsampling by a band synthesis filtering process requiring a delay for replicating the monaural audio signal, so that the duplicated monaural audio signal can be output as a stereo audio signal On the other hand, if the encoded information is transmitted for the first time and thus the spatial encoding process is valid, the delayed signal for the audio signal band synthesis processing for the monaural signal is treated as the delayed signal for the audio signal band synthesis processing for the stereo signal. Is composed, by it QMF sum The occurrence of irregular sound in the audio signal due to the delayed output of the filtering process can be prevented.

Further, at least all delay signals of the subband division filtering process are updated so that the delay in the subband division filtering process is updated and the output signal duplication process is performed. Then, after all of the delay signals are updated, normal playback processing is performed, whereby occurrence of irregular sounds in the output audio signal due to the delay caused by the filtering processing can be prevented.

As a result of this configuration, even in the case of encoded audio data requiring spatial decoding processing in which part of encoded information is multiplexed and transmitted, reproduction from any position can be realized without generating irregular sounds.

The present invention is not limited only to the above-described embodiment, of course, and can be modified in various ways within the scope and spirit of the present invention. For example, in the above-described embodiment of the present invention, a playback apparatus or playback method having a hardware configuration has been described. However, the above processing step may be implemented by executing the program by software, that is, by a computer using a CPU (Central Processing Unit). In addition, such a computer program may be provided recorded on a recording medium.

1 is a block diagram showing a schematic configuration of a playback apparatus according to an embodiment of the present invention.

Fig. 2 is a block diagram showing an example of the configuration of an embodiment of the present invention to be applied to a reproduction device for reproducing encoded audio data encoded by the HE AAC v2 method.

FIG. 3 is a flow chart for explaining the operation of the playback device shown in FIG.

FIG. 4 is a flowchart for explaining a specific example of the PS processing in step S120 of FIG. 3.

FIG. 5 is a flowchart for explaining another specific example of the PS processing in step S120 of FIG. 3.

6 is a block diagram showing a configuration example of a conventional stereo processing apparatus.

FIG. 7 is a diagram illustrating an example of a signal supplied to the stereo processing apparatus of FIG. 6.

8 is a diagram illustrating an example of a signal supplied to a stereo processing apparatus of an HE AAC v2 system.

9 is a block diagram showing an example of the configuration of a reproduction device for reproducing coded audio data encoded in the HE AAC v2 system.

FIG. 10 is a flowchart for explaining the operation of the playback device shown in FIG.

Fig. 11 is a waveform diagram comparing the output audio signal of the conventional playback apparatus with the output audio signal of the playback apparatus to which the embodiment of the present invention is applied.

[Description of the code]

11. Input terminal of encoded audio data 21. QMF analyzer

23. HF generator 27. Hybrid analyzer

30, 45. Stereo processing section 31, 32. Hybrid synthesis section

33, 34. QMF Synthesizer 44 Band Division

51, 52. Band Synthesis Unit

Claims (9)

The encoded stereo data required for stereo processing is transmitted intermittently multiplexed into coded information of a monaural audio signal. In a playback method for decoding and processing coded audio data, A first step of outputting a stereo audio signal using the monaural audio signal if the necessary stereo processing information is not supplied; When the necessary stereo processing information is supplied, starting updating of a state variable in the filter and outputting the stereo audio signal using the monaural audio signal until all the state variables are updated. With two stages, If all the state variables in the filter have been updated, based on stereo process information obtained by the necessary stereo processing information for the monaural audio signal to generate an output stereo audio signal; And a third step of performing stereo processing. The method of claim 1, And the stereo processing is performed on a band-extended monaural audio signal. The method of claim 1, In the first step, the monaural audio signal is divided into at least two or more subbands by a band division filtering process, and the at least two or more subbands are band synthesis filtering process. up-sampled to output a stereo audio signal using the monaural audio signal, In the second step, the state variable in the filter for the monaural audio signal is processed as a filtering state variable for the stereo audio signal. The method of claim 1, The encoded audio data is, AAC core coded information equivalent to monaural audio data based on HE AAC (High Efficiency Advanced Audio Coding) coding scheme, Encoding information for SBR (Spectral Band Replication) processing, Has encoding information for PS (Parametric Stereo) processing, The encoded information for the SBR processing includes an SBR header, which is multiplexed and intermittently transmitted information, and SBR data, which is encoded information, which is always transmitted. PS data, which is encoding information for the PS processing, is included in an extended area of the SBR data and transmitted. And the SBR header is the necessary stereo processing information required for decoding the SBR data. A playback apparatus for decoding and reproducing encoded audio data transmitted by intermittently multiplexing necessary stereo processing information required for stereo processing into encoded information of a monaural audio signal, Band division means for band-dividing the supplied monaural audio signal; Stereo processing means for stereo processing the signal from the band dividing means based on the stereo processing information included in the multiplexed encoding information; Band synthesis means for band-synthesizing left and right channel stereo signals from the stereo processing means, respectively; If the necessary stereo processing information is not supplied, the monaural audio signal is used to output a stereo audio signal, and when the required stereo processing information is supplied, when the state variable in the filter is started and all the state variables are updated. And outputting a stereo audio signal by using the monaural audio signal until the state variable in the filter has been updated to generate and output a stereo audio signal by the necessary stereo processing information for the monaural audio signal. And control means for controlling to perform stereo processing based on the stereo processing information. The method of claim 5, And the stereo processing is performed on a band-extended monaural audio signal. A program for causing a computer to execute a process of decoding and reproducing encoded audio data transmitted by intermittently multiplexing necessary stereo processing information required for stereo processing into encoded information of a monaural audio signal. A first step of outputting a stereo audio signal using the monaural audio signal if the necessary stereo processing information is not supplied; A second step of, when the required stereo processing information is supplied, starting updating of the state variables in the filter and outputting a stereo audio signal using the monaural audio signal until all the state variables are updated; A third step of performing stereo processing on the monaural audio signal based on the stereo processing information obtained by the necessary stereo processing information to generate and output a stereo audio signal when all the state variables in the filter have been updated. Program characterized in that configured by. In a recording medium on which a program for causing a computer to execute a process of decoding and reproducing encoded audio data transmitted by intermittently multiplexing necessary stereo processing information required for stereo processing to encoded information of a monaural audio signal is provided. , The recording medium, A first step of outputting a stereo audio signal using the monaural audio signal if the necessary stereo processing information is not supplied; A second step of, when the required stereo processing information is supplied, starting updating of the state variables in the filter and outputting a stereo audio signal using the monaural audio signal until all the state variables are updated; A third step of performing stereo processing on the monaural audio signal based on the stereo processing information obtained by the necessary stereo processing information to generate and output a stereo audio signal when all the state variables in the filter have been updated. The recording medium, characterized in that the recorded program is recorded. A playback apparatus for decoding and reproducing encoded audio data transmitted by intermittently multiplexing necessary stereo processing information required for stereo processing into encoded information of a monaural audio signal, A band divider for band-dividing the supplied monaural audio signal; A stereo processor which stereo-processes the signal from the band splitter based on the stereo processing information included in the multiplexed encoding information; A band synthesizer for band-synthesizing left and right channel stereo signals from the stereo processor, respectively; If the necessary stereo processing information is not supplied, the monaural audio signal is used to output a stereo audio signal, and when the required stereo processing information is supplied, when the state variable in the filter is started and all the state variables are updated. And outputting a stereo audio signal by using the monaural audio signal until the state variable in the filter has been updated to generate and output a stereo audio signal by the necessary stereo processing information for the monaural audio signal. And a controller for controlling to perform stereo processing based on the stereo processing information.

Images