KR20100050414A - Method and apparatus for processing an audio signal - Google Patents

Abstract

PURPOSE: An audio signal processing method and an apparatus thereof are provided to remarkably reduce the number of bits of the bit stream by skipping transmission of information on noise peeling. CONSTITUTION: A multiplexer(210) extracts the noise filling flag information and coding scheme information. In the noise filling flag information is a plurality of frames, the active condition of the noise filling is directed. In the coding scheme information is the frequency domain or the time domain, the coding state of the current frame is shown. The noise information decoding part(201) extracts the noise level information about the current frame.

Description

Method and apparatus for processing audio signal {METHOD AND APPARATUS FOR PROCESSING AN AUDIO SIGNAL}

The present invention relates to an audio signal processing method and apparatus capable of encoding or decoding an audio signal.

In general, a coding scheme based on audio characteristics is applied to an audio signal such as a music signal, and a coding scheme based on speech characteristics is applied to a speech signal.

When any one of the coding schemes is applied to a signal in which audio and voice characteristics are mixed, audio coding efficiency or sound quality deteriorates.

The present invention has been made to solve the above problems, and to provide an audio signal processing method and apparatus capable of applying a noise filling method in a decoder to compensate for a signal lost in the quantization process of encoding.

It is still another object of the present invention to provide an audio signal processing method and apparatus which can omit transmission of information on noise filling for a frame to which a noise filling method is not applied.

It is still another object of the present invention to provide an audio signal processing method and apparatus for encoding information on noise filling based on a characteristic in which information on noise filling (noise level or noise offset) has almost similar values every frame. To provide.

In order to achieve the above object, the present invention comprises the steps of: extracting noise filling flag information indicating whether noise filling is used in a plurality of frames; Extracting coding scheme information indicating whether a current frame including the plurality of frames is coded in a frequency domain or in a time domain; Extracting noise level information for a current frame when the noise filling flag information indicates that the noise filling is used in the plurality of frames and the coding information indicates that the current frame is coded in the frequency domain; Extracting noise offset information for a current frame when a noise level value corresponding to the noise level information satisfies a predetermined level; And when the noise offset information is extracted, performing noise filling on the current frame based on the noise level value and the noise level information.

According to the present invention, the noise filling may include: determining a loss area of the current frame using spectral data of the current frame; Generating compensated spectral data by filling a compensation signal in the loss region using the noise level value; And generating a compensated scale factor based on the noise offset information.

According to the present invention, there is provided a method comprising extracting a level pilot value representing a reference value of a noise level and an offset pilot value representing a reference value of a noise offset; Obtaining the noise level value by adding the level pilot value and the noise level information; And when the noise offset information is extracted, obtaining the noise offset value by adding the offset pilot value and the noise offset information, wherein the noise filling uses the noise level value and the noise offset value. Can be performed.

According to the present invention, using the noise level value of the previous frame, and the noise level information of the current frame, obtaining a noise level value of the current frame; When the noise offset information is extracted, the method may further include obtaining a noise offset value of the current frame by using the noise offset value of the previous frame and the noise offset information of the current frame. The noise level value and the noise offset value may be performed.

According to the present invention, the noise level information and the noise offset information may be extracted according to a variable length coding scheme.

According to another aspect of the present invention, noise filling flag information indicating whether noise filling is used in a plurality of frames and whether a current frame including the plurality of frames is coded in the frequency domain or in the time domain A multiplexer for extracting representing coding scheme information; When the noise filling flag information indicates that the noise filling is used in the plurality of frames and the coding information indicates that the current frame is coded in the frequency domain, extracting noise level information for a current frame, and extracting the noise A noise information decoding part for extracting noise offset information for the current frame when a noise level value corresponding to the level information satisfies a predetermined level; And a loss compensation part configured to perform noise peeling on the current frame based on the noise level value and the noise level information when the noise offset information is extracted.

According to the present invention, the loss compensation part is, by using the spectral data of the current frame to determine the loss area of the current frame, by using the noise level value by filling the compensation signal in the loss area, the compensation The spectral data may be generated and a compensated scale factor may be generated based on the noise offset information.

According to the present invention, a level pilot value indicating a reference value of a noise level and an offset pilot value indicating a reference value of a noise offset are extracted, and the noise level value is obtained by adding the level pilot value and the noise level information. And when the noise offset information is extracted, further comprising a data decoding part configured to obtain a noise offset value by adding the offset pilot value and the noise offset information, wherein the noise filling uses the noise level value and the noise offset value. It may be to be performed by.

According to the present invention, the noise level value of the current frame is obtained using the noise level value of the previous frame and the noise level information of the current frame, and when the noise offset information is extracted, the noise offset value of the previous frame. And a data decoding part for obtaining a noise offset value of the current frame using the noise offset information of the current frame, wherein the noise filling is performed using the noise level value and the noise offset value. It may be.

According to the present invention, the noise level information and the noise offset information may be extracted according to a variable length coding scheme.

According to another aspect of the invention, generating a noise level value and a noise offset value based on the quantized signal; Generating noise filling flag information indicating whether noise filling is used for the plurality of frames; Generating coding scheme information indicating whether a current frame included in the plurality of frames is coded in a frequency domain or a time domain; When the noise filling flag information indicates that noise filling is used for the plurality of frames and the coding scheme information indicates whether the current frame is coded in the frequency domain, the noise frame value of the current frame corresponding to the noise level value. Inserting noise level information into the bitstream; And inserting noise offset information corresponding to the noise offset value into the bitstream when the noise level value satisfies a predetermined level.

According to another aspect of the present invention, a loss compensation estimation part for generating a noise level value and a noise offset value based on a quantized signal and generating noise filling flag information indicating whether noise filling is used in a plurality of frames ; A signal classifier configured to generate coding scheme information indicating whether a current frame included in the plurality of frames is coded in a frequency domain or a time domain; When the noise filling flag information indicates that noise filling is used for the plurality of frames and the coding scheme information indicates whether the current frame is coded in the frequency domain, the noise frame value of the current frame corresponding to the noise level value. And a noise information encoding part for inserting noise level information into the bitstream and inserting noise offset information corresponding to the noise offset value into the bitstream when the noise level value satisfies a predetermined level. An apparatus is provided.

According to another aspect of the present invention, the method includes: extracting noise filling flag information indicating whether noise filling is used for a plurality of frames; Extracting coding scheme information indicating whether a current frame including the plurality of frames is coded in a frequency domain or in a time domain; Extracting noise level information for a current frame when the noise filling flag information indicates that the noise filling is used in the plurality of frames and the coding information indicates that the current frame is coded in the frequency domain; Extracting noise offset information for a current frame when a noise level value corresponding to the noise level information satisfies a predetermined level; And when the noise offset information is extracted, performing noise filling on a current frame based on the noise level value and the noise level information, when executed by the processor. A computer readable storage medium having stored therein a command is provided.

The present invention provides the following effects and advantages.

First, for a frame to which the noise filling method is not applied, the number of bits in the bitstream can be significantly reduced, which can omit the transmission of information on the noise filling.

Second, since the current frame determines whether noise filling is applied and extracts specific information about the noise filling from the bitstream, information necessary for efficient parsing without increasing complexity for the parsing process is required. Can be obtained.

Third, for information having similar values in each frame (e.g., noise level or noise offset), the number of bits can be further reduced by transmitting the difference value with the corresponding value of the previous frame without transmitting the value as it is. have.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Prior to this, terms or words used in the specification and claims should not be construed as having a conventional or dictionary meaning, and the inventors should properly explain the concept of terms in order to best explain their own invention. Based on the principle that can be defined, it should be interpreted as meaning and concept corresponding to the technical idea of the present invention. Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention and do not represent all of the technical idea of the present invention, various modifications that can be replaced at the time of the present application It should be understood that there may be equivalents and variations.

In the present invention, the following terms may be interpreted based on the following criteria, and terms not described may be interpreted according to the following meanings. Coding can be interpreted as encoding or decoding in some cases, and information is a term that encompasses values, parameters, coefficients, elements, and so on. It may be interpreted otherwise, but the present invention is not limited thereto.

Here, the audio signal is a concept that is broadly distinguished from the video signal, and refers to a signal that can be visually identified during reproduction. The narrow signal is a concept that is distinguished from a speech signal. Means a signal with little or no characteristics. The audio signal in the present invention should be interpreted broadly and can be understood as a narrow audio signal when used separately from a voice signal.

1 is a view showing the configuration of an encoder in an audio signal processing apparatus according to an embodiment of the present invention, Figure 2 is a view showing the sequence of the encoding method of the audio signal processing method according to an embodiment of the present invention.

First, referring to FIG. 1, an encoder side 100 of an audio signal processing apparatus includes a noise information encoding part 101, a data encoding part 102, and an entropy coding part ( 103, a loss compensation estimating part 110 and a multiplexer 120 may be further included. The audio signal processing apparatus according to the present invention encodes a noise offset based on the noise level.

The loss compensation estimation part 110 generates information about noise filling based on the quantized signal. The information about the noise filling may include noise filling flag information, noise level, noise offset, and the like.

In detail, the loss compensation estimation part 110 first receives the quantized signal and the coding identification information (step S110). The coding scheme information is information indicating whether the current frame is a frequency domain based scheme or a time domain based scheme, and may be information generated by a signal classifier (not shown). The loss compensation estimation part 110 may generate information about noise filling only when the signal is a frequency domain signal. This coding scheme information may be passed to the multiplexer 120, an example of syntax for encoding coding scheme information will be described later.

On the other hand, quantization is a process of obtaining scale factor and spectral data from spectral coefficients. Here, the scale factor and spectral data are quantized signals. Here, the spectral coefficient may be an MDCT coefficient obtained through a Modified Discrete Transform (MDCT) transformation, but the present invention is not limited thereto. In other words, the spectral coefficients may be similarly expressed using scale factors that are integers and spectral data that are integers as shown in Equation 1 below.

Where X is spectral coefficient, scalefactor is scale factor, and spectral data is spectral data.

3 is a diagram for explaining a quantization concept. Referring to FIG. 3, a process of representing spectral coefficients (a, b, c, etc.) as scale factors (A, B, C, etc.) and spectral data (a ', b', c ', etc.) is conceptually shown. have. Scale factors (A, B, C, etc.) are factors applied to groups (specific bands or specific intervals). As such, by using a scale factor representing a group (eg, scale factor band), the coding efficiency can be improved by collectively converting the sizes of the coefficients belonging to the group. The scale factor and spectral data determined as described above may be used as they are, but may be corrected through a masking process based on a psychoacoustic model, which will not be described in detail.

The loss compensation estimation part 110 determines a loss area in which a loss signal exists based on the spectral data. 4 is a diagram for describing the concept of a loss signal and a loss area. Referring to FIG. 4, it can be seen that at least one spectral data exists for each spectral band sfb ₁ , sfb ₂ , and sfb ₄ . Each spectral data corresponds to an integer value between 0 and 7. Of course, the spectral data may be a value from -50 to 100 instead of an integer of 0 to 7, but Figure 4 is an example for explaining the concept is not limited thereto. In some samples, bins, or regions, when the absolute value of the spectral data indicates a value that is less than or equal to a specific value (eg, 0), it may be determined that a signal is lost or a lossy region exists. If a specific value is 0, as shown in FIG. 4, a loss signal is generated in the second and third spectral bands sfb ₂ and sfb ₃ , and in the case of the _third spectral band sfb ₃ , the entire band is It can be said to correspond to a loss area.

The loss compensation estimation part 110 determines whether to use a noise filling method for a plurality of frames or a sequence to compensate for a loss signal in the loss area, and generates noise filling flag information based on the loss compensation estimation part 110. That is, the noise filling flag information is information indicating whether a noise filling method is used to compensate for a lost signal for a plurality of frames or sequences. On the other hand, it is not whether the noise filling method is used for all the frames belonging to a plurality of frames or sequences, but whether the noise filling method is likely to be used for a specific frame therein. The noise filling flag information may be included in a header corresponding to information common to a plurality of frames or sequences. The generated noise filling flag information is transmitted to the multiplexer 120. 5 is an example of syntax for encoding noise filling flag information. Referring to (L1) of FIG. 5, the noise filling flag information (noisefilling) is included in a header (USACSpecificConfig ()) that transmits information (for example, frame length, eSBR usage, etc.) commonly applied to the entire sequence. It can be seen that. If the noise filling flag information is 0, it means that the noise filling method cannot be used for the entire sequence. In contrast, when the noise filling flag information is 1, it may mean that the noise filling method is used for one or more frames included in the entire sequence.

Referring back to FIGS. 1 and 2, the loss compensation estimation part 110 generates a noise level and a noise offset for a loss area in which a loss signal exists (S130). 6 is a diagram for explaining a noise level and a noise offset. Referring to FIG. 6, the noise level may generate a compensation signal (for example, a random signal) in place of a loss signal in an area where spectral data is lost, and information for determining the level of the compensation signal may be noise. Level. The noise level and the compensation signal (random signal) can be expressed by the following equation. The noise level can be determined for each frame.

spectral_data = noise _val X random_signal

spectral_data is spectral data, noise _val is a value obtained using noise level, and random_signal is a random signal

On the other hand, the noise offset is information for correcting the scale factor. As mentioned above, the noise level is a factor for correcting spectral data in Equation 2, but the value of the noise level is limited in range. For the lossy region, it may be more effective to correct the scale factor, rather than correcting the spectral data through the noise level, to finally make the spectral coefficient large. At this time, a value for correcting the scale factor is a noise offset. The relationship between the noise offset and the scale factor may be expressed as the following equation.

sfc_d = sfc_c-noise_offset

sfc_c is the scale factor, sfc_d is the scale factor to be transmitted, and noise_offset is the noise offset.

The noise offset may be applied only when the entire spectral band is a loss region. For example, in FIG. 6, the noise offset may be applied only to the _third spectral band sfb ₃ . This is because, when only a part of one spectral band has a lossy region, when the noise offset is applied, the number of bits of spectral data corresponding to the non-lossy region may be increased.

The noise information encoding part 101 encodes the noise offset based on the noise level value and the noise offset value received from the loss compensation estimation part 110. For example, the noise offset value may be encoded only when the noise level value satisfies a predetermined condition (specific level range). For example, when the noise level value exceeds 0 ('no' in step S140), since the noise filling method is performed, the noise offset information is transmitted by passing the noise offset value to the data coding part 102. Included in the stream (step S160).

On the contrary, when the noise level value is 0 ('yes' in step S140), since the noise filling method is not performed, only the noise level value of 0 is encoded and the noise offset value is excluded from the bitstream. (Step S150).

7 is an example of syntax for encoding a noise level and a noise offset. Referring to (L1) of FIG. 7, it can be seen that the current frame corresponds to a signal of a frequency domain. Referring to (L2) and (L3), it can be seen that noise level information (noise_level) is included in the bitstream only when noise filling flag information (noiseFilling) is 1. This is because when the noise filling flag information (noiseFilling) is 0, it means that the noise filling is not applied to the entire sequence to which the current frame belongs. On the other hand, referring to (L4) and (L5), it can be seen that the noise offset information (noise_offset) is included in the bitstream only when the noise level value is greater than zero.

Referring back to FIGS. 1 and 2, the data coding part 102 data codes a noise level value (and a noise offset value) using a differential coding scheme or a pilot coding scheme. . Here, the differential coding method is a method of transmitting only a difference value between a noise level value of a previous frame and a noise level value of a current frame, and may be expressed as in the following equation.

noise_info_diff_cur = noise_info_cur-noise_info_prev

noise_info_cur is the noise level (or offset) of the current frame, noise_info_prev is the noise level (or offset) of the previous frame, and noise_info_diff_cur is the difference value.

Only the difference value obtained by subtracting the noise level (or offset) value of the previous frame from the noise level (or offset) of the current frame received from the noise information encoding part 101 is transmitted to the entropy coding part 103.

On the other hand, the pilot coding method is a reference value corresponding to a noise level (or offset) value corresponding to two or more frames (eg, an average value, a median value, a mode value, etc. of noise levels (or offsets) of N frames in total). The pilot value is determined and the difference value between the pilot value and the noise level (or offset) of the current frame is transmitted.

noise_info_diff_cur = noise_info_cur-noise_info_pilot

noise_info_cur is the noise level (or offset) of the current frame, noise_info_pilot is the pilot of the noise level (or offset), and noise_info_diff_cur is the difference value.

Here, the pilot of the noise level (or offset) may be transmitted through the header. Here, the header may be the same as the header through which the noise filling flag information is transmitted.

That is, when the differential coding scheme or the pilot coding scheme is applied, the noise level value of the current frame does not become noise level information included in the bitstream as it is, but the difference value of the noise level value (difference value of DIFF coding and pilot coding). Difference value) becomes noise level information.

In this way, noise level information is generated by performing differential coding or pilot coding (steps S170 and S180), and when a noise offset value is generated, differential coding or pilot coding is also performed on the noise offset value. By doing so, noise offset information is generated (step S180). This noise level information (and noise offset information) is transmitted to the entropy coding part 103.

Entropy coding part 103 performs entropy coding on noise level information (and noise offset information). When the noise level information (and the noise offset information) is coded according to the differential coding scheme or the pilot coding scheme through the data coding part 102, the information corresponding to the difference value is converted into the variable length coding scheme (eg, the entropy coding scheme). Huffman coding). Since the difference value may be 0 or a value similar to 0, encoding according to a variable length coding scheme rather than encoding with fixed bits may further reduce the number of bits.

The multiplexer 120 receives the coding scheme information received from the signal classifier (not shown), the noise level information (and noise offset information) received through the entropy coding part 103, and the loss compensation estimation part 110. The bit filling is generated by multiplexing the noise filling flag information and the quantized signal (spectral data and scale factor). The syntax for encoding the noise filling flag information may be the same as in FIG. 5 described above, and the syntax for encoding the noise level information (and the noise offset information) may be as in FIG. 7 described above.

8 is an example of syntax for encoding coding scheme information. Referring to (L1) of FIG. 8, it can be seen that coding scheme information (core_mode) indicating whether the current frame is applied to the frequency domain based method or the time domain based method is included. Referring to (L2) and (L3), it can be seen that when the coding scheme information indicates that the time domain based scheme is applied, the time domain based channel stream is transmitted. Referring to (L4) and (L5), it can be seen that when the coding scheme information indicates that the frequency domain based scheme is applied, the channel stream based on the frequency domain is transmitted. As mentioned above, the frequency domain based channel stream fd_channel_stream () may include information on noise filling (noise level information (and noise offset information)), as mentioned above with reference to FIG. 7. .

As described above, the apparatus and method for encoding an audio signal according to an embodiment of the present invention may include information on noise filling (especially noise) according to whether a noise filling method is actually applied to a specific frame among sequences in which the noise filling method may be used. Offset information) or encode.

FIG. 9 is a diagram illustrating a decoder of an audio signal processing apparatus according to an exemplary embodiment of the present invention, and FIG. 10 is a diagram illustrating a detailed configuration of a loss compensation part of FIG. 9. 11 is a diagram illustrating a procedure of a decoding method of an audio signal processing method according to an embodiment of the present invention.

First, referring to FIGS. 9 and 11, the decoding apparatus 200 of the audio signal processing apparatus includes a noise information decoding part 201, an entropy decoding part 202, a data decoding part 203, and a multiplexer 210. The loss compensation part 220 and the scaling part 230 may be further included.

First, the multiplexer 210 extracts noise filling flag information from a bitstream (particularly, a header) (step S210). Then, the coding scheme information and the quantized signal for the current frame are received (S220). The noise filling flag information, the coding scheme information, and the quantized signal are as described above. The noise filling flag information is information on whether a noise filling scheme is used for a plurality of frames, and the coding scheme information is information on whether a current frame is applied to a frequency domain based method or a time domain based method among the plurality of frames. Information indicating whether or not. The quantized signal may be spectral data and scale factor when a frequency domain based scheme is applied. The noise filling information may be extracted according to the syntax shown in FIG. 5, and the coding scheme information may be extracted according to the syntax shown in FIG. 8. The noise filling information and the coding scheme information extracted by the multiplexer 210 are transferred to the noise information decoding part 201.

The noise information decoding part 201 extracts information on noise filling (noise level information, noise offset information) from the bitstream based on the noise filling flag information and the coding scheme information. Specifically, the noise filling flag information indicates that the noise filling scheme may be used for a plurality of frames (YES in step S230), and when the current frame is applied in a frequency domain based manner (yes in step S240). The information decoding part 201 extracts noise level information from the bitstream (S250). Of course, step S240 may be performed before step S230. Steps S230 to S250 may be performed according to the syntax indicated in the above-described lines (L1) to (L3) of FIG. 7. As described above with reference to FIG. 6, the noise level information is information about a level of a compensation signal (eg, a random signal) inserted into a region (or a sample or a bin) in which spectral data is lost.

In step S230, when the noise filling flag information indicates that the noise filling scheme cannot be used for one frame of the plurality of frames (no in step S230), no step on noise filling is performed and the procedure ends. Can be. Even in step S240, even when the current frame is a frame to which a time domain based method is applied (no in step S240), a procedure regarding noise filling may not be performed.

The dequantization part (not shown) dequantizes the received spectral data to generate dequantized spectral data. Here, the dequantized spectral data is a 4/3 power to the received spectral data, as shown in Equation 1 above.

If the noise level information is extracted in step S250, and the noise level is greater than 0 (since the noise filling scheme is applied to the current frame) (YES in step S260), the noise information decoding part 201 is determined from the bitstream. The noise offset information is extracted (step S270). Steps S260 to S270 may be performed according to the syntax previously described in the (L4) and (L5) lines of FIG. 7. As described with reference to FIG. 6, the noise offset information is information for correcting a scale factor corresponding to a specific scale factor band. In this case, the specific scale factor band may be a scale factor band in which all spectral data is lost. If this noise offset information is obtained, the dequantized spectral data and scale factor for the current frame are passed through the loss compensation part 220. If the noise offset information is not obtained, the inverse quantized spectral for the current frame is obtained. The data and scale factor bypass the loss compensation part 220 and directly enter the scaling part 230.

The noise level information extracted in step S250 and the noise offset information extracted in step S270 are entropy decoded in the entropy decoding part 202. If the information is encoded according to a variable length coding scheme (eg, Huffman coding) among the entropy decoding schemes, the information may be entropy decoded according to the variable length decoding scheme.

The data decoding part 203 performs data decoding on entropy decoded noise level information (and noise offset information) according to a differential method or a pilot method. In the case of the DIFF coding, the noise level (or offset) of the current frame may be obtained according to the following equation.

noise_info_cur = noise_info_prev + noise_info_diff_cur

noise_info_cur is the noise level (or offset) of the current frame, noise_info_prev is the noise level (or offset) of the previous frame, and noise_info_diff_cur is the difference value.

If it corresponds to pilot coding, the noise level (or offset) of the current frame can be obtained according to the following equation.

noise_info_cur = noise_info_pilot + noise_info_diff_cur

noise_info_cur is the noise level (or offset) of the current frame, noise_info_pilot is the pilot of the noise level (or offset), and noise_info_diff_cur is the difference value.

Here, the pilot of the noise level (or offset) may be information included in the header. The noise level (and noise offset) thus obtained is transferred to the loss compensation part 230.

When both the noise level and the noise offset are acquired, the loss compensation part 220 performs noise peeling on the current frame based on this (step S280). A detailed schematic diagram of the loss compensation part 220 is shown in FIG. 10.

Referring to FIG. 10, the loss compensation part 220 includes a spectral data filling part 222 and a scale factor correction part 224. The spectral data filling part 222 determines whether there is a loss area among the spectral data belonging to the current frame, and fills the compensation signal using the noise level for the loss area. As a result of analyzing the received spectral data, if the spectral data is less than or equal to a predetermined value (for example, 0), the corresponding sample is determined as a loss region. The loss area may be as shown in FIG. 4 above. As shown in Equation 2, spectral data corresponding to a lossy region may be generated by applying a noise level value to a compensation signal (for example, a random signal). In this way, the compensated spectral data is generated by filling the compensation signal in the loss region.

The scale factor correction part 224 compensates for the received scale factor with a noise offset. The scale factor can be compensated according to the following equation.

sfc_c = sfc_d + noise_offset

sfc_c is the compensated scale factor, sfc_d is the transmitted scale factor, and noise_offset is the noise offset.

As mentioned above, the compensation of the noise offset may be performed only for the scale factor band when the entire scale factor band corresponds to the loss region. The compensated spectral data and the compensated scale factor generated in the loss compensation part 22 are input to the scaling part 230 shown in FIG. 9.

9 and 11, the scaling part 230 scales one of the received spectral data and the compensated spectral data using the scale factor or the compensated scale factor (step S290). Scaling is to obtain spectral coefficients using inverse quantized spectral data (spectral_data ^{4/3 in} the following equation) using a scale factor as shown in the following equation.

X 'is the restored spectral coefficient, spectral_data is the received spectral data or compensated spectral data, and scalefactor is the received scale factor or compensated scale factor.

The decoding apparatus of the audio signal processing apparatus according to the embodiment of the present invention performs noise filling by acquiring information on noise filling by performing the above steps.

12 is an example of an audio signal encoding apparatus to which an audio signal processing apparatus according to an embodiment of the present invention is applied, and FIG. 13 is an example of an audio signal decoding apparatus to which an audio signal processing apparatus according to an embodiment of the present invention is applied.

The audio signal processing apparatus 100 of FIG. 12 may include the noise information encoding part 101 described with reference to FIG. 1, and may further include a data coding part 102 and an entropy coding part 103. The audio signal processing apparatus 200 of FIG. 13 may include the noise information decoding part 201 described with reference to FIG. 9, and may further include an entropy decoding part 202 and a data decoding part 203.

First, referring to FIG. 12, an audio signal encoding apparatus 300 includes a multi-channel encoder 310, a band extension coding unit 320, an audio signal encoder 330, a voice signal encoder 350, and a loss compensation estimation unit 350. ), An audio signal processing apparatus 100, and a multiplexer 360.

The multi-channel encoder 310 receives a plurality of channel signals (two or more channel signals) (hereinafter, referred to as a multi-channel signal), performs downmixing to generate a mono or stereo downmix signal, and multi-channels the downmix signal. Generates spatial information needed for upmixing to a signal. The spatial information may include channel level difference information, interchannel correlation information, channel prediction coefficients, downmix gain information, and the like. If the audio signal encoding apparatus 300 receives the mono signal, the multi-channel encoder 310 may bypass the mono signal without downmixing.

The band extension encoder 320 may apply a band extension method to the downmix signal output from the multi-channel encoder 310 to generate spectral data corresponding to a low frequency band and band extension information for high frequency band extension. . That is, the spectral data of some bands (eg, high frequency bands) of the downmix signal may be excluded, and band extension information for restoring the excluded data may be generated.

The signal generated through the band extension coding unit 320 is input to the audio signal encoder 330 or the voice signal decoder 340 according to the coding scheme information generated by the signal classification unit (not shown).

The audio signal encoder 330 encodes the downmix signal according to an audio coding scheme when a specific frame or a specific segment of the downmix signal has a large audio characteristic. Here, the audio coding scheme may be based on an AAC standard or a high efficiency advanced audio coding (HE-AAC) standard, but the present invention is not limited thereto. The audio signal encoder 330 may correspond to a modified discrete transform (MDCT) encoder.

The speech signal encoder 340 encodes the downmix signal according to a speech coding scheme when a specific frame or a segment of the downmix signal has a large speech characteristic. Here, the speech coding scheme may be based on an adaptive multi-rate wide-band (AMR-WB) standard, but the present invention is not limited thereto. Meanwhile, the speech signal encoder 340 may further use a linear prediction coding (LPC) method. When the harmonic signal has high redundancy on the time axis, the harmonic signal may be modeled by linear prediction that predicts the current signal from the past signal. In this case, the linear prediction coding method may increase coding efficiency. Meanwhile, the voice signal encoder 340 may correspond to a time domain encoder.

Since the loss compensation estimation unit 350 may perform the same function as the loss compensation estimation unit 110 described above with reference to FIG. 1, a detailed description thereof will be omitted.

The audio signal processing unit 100 includes the noise information encoding part 101 described with reference to FIG. 1, thereby encoding the noise level and the noise offset generated by the loss compensation estimation unit 350.

The multiplexer 360 includes spatial information, bandwidth extension information, a signal encoded by each of the audio signal encoder 330 to the voice signal encoder 340, noise filling flag information, and noise generated by the audio signal processing unit 100. By multiplexing the level information (and noise offset information), one or more bitstreams are generated.

Referring to FIG. 13, the audio signal decoding apparatus 400 includes a demultiplexer 410, an audio signal processing apparatus 200, a loss compensation part 420, a scaling part 430, an audio signal decoder 440, and a voice signal decoder. 450, band extension decoding unit 460, and multi-channel decoder 470.

The demultiplexer 410 extracts noise filling flag information, quantized signals, coding scheme information, band extension information, spatial information, and the like from the audio signal bitstream.

The audio signal processing unit 200 includes the noise information decoding unit 201 described with reference to FIG. 9 as mentioned above, and based on the noise filling flag information and the coding scheme information, noise level information (and noise) from the bitstream. Offset information).

The inverse quantization unit (not shown) performs inverse quantization on the received spectral data to directly transfer the dequantized spectral data to the loss compensation part 420, or when the noise filling is skipped, the loss compensation part 420. Bypasses and passes to the scaling part 430.

The loss compensation part 420 is the same component as the loss compensation part 220 described with reference to FIG. 9, and when noise filling is applied to the current frame, noise filling is performed on the current frame using the noise level and the noise offset. Do this.

Scaling part 430 is the same component as scaling part 230 described with reference to FIG. 9 and obtains spectral coefficients by performing scaling on dequantized spectral data or compensated spectral data.

The audio signal decoder 440 decodes the audio signal by an audio coding method when the audio signal (eg, spectral coefficient) has a large audio characteristic. As described above, the audio coding scheme may be based on the AAC standard and the HE-AAC standard. The speech signal decoder 450 decodes the downmix signal using a speech coding scheme when the audio signal has a large speech characteristic. As described above, the speech coding scheme may conform to the AMR-WB standard, but the present invention is not limited thereto.

The band extension decoding unit 460 performs a band extension decoding method on the output signals of the audio signal decoder 440 and the voice signal decoder 450, thereby restoring a high frequency band signal based on the band extension information.

When the decoded audio signal is downmixed, the multichannel decoder 470 generates an output channel signal of a multichannel signal (including a stereo signal) using spatial information.

The audio signal processing apparatus according to the present invention can be included and used in various products. These products can be broadly divided into stand alone and portable groups, which can include TVs, monitors and set-top boxes, and portable groups include PMPs, mobile phones, and navigation. can do.

14 is a diagram illustrating a relationship between products in which an audio signal processing device according to an embodiment of the present invention is implemented. First, referring to FIG. 14, the wired / wireless communication unit 510 receives a bitstream through a wired / wireless communication method. Specifically, the wired / wireless communication unit 510 may include at least one of a wired communication unit 510A, an infrared communication unit 510B, a Bluetooth unit 510C, and a wireless LAN communication unit 510D.

The user authentication unit 520 receives user information and performs user authentication. The user authentication unit 520 receives one or more of a fingerprint recognition unit 520A, an iris recognition unit 520B, a face recognition unit 520C, and a voice recognition unit 520D. The fingerprint, iris information, facial contour information, and voice information may be input, converted into user information, and the user authentication may be performed by determining whether the user information matches the existing registered user data. .

The input unit 530 is an input device for a user to input various types of commands, and may include one or more of a keypad unit 530A, a touch pad unit 530B, and a remote controller unit 530C. It is not limited.

The signal coding unit 540 encodes or decodes an audio signal and / or a video signal received through the wired / wireless communication unit 510, and outputs an audio signal of a time domain. Audio signal processing device 545, which corresponds to the embodiment of the present invention described above (i.e., encoding side 100 and / or decoding side 200). The signal coding unit including this may be implemented by one or more processors.

The controller 550 receives input signals from the input devices, and controls all processes of the signal decoding unit 540 and the output unit 560. The output unit 560 is a component that outputs an output signal generated by the signal decoding unit 540, and may include a speaker unit 560A and a display unit 560B. When the output signal is an audio signal, the output signal is output to the speaker, and when the output signal is a video signal, the output signal is output through the display.

15 is a relationship diagram of products in which an audio signal processing apparatus according to an embodiment of the present invention is implemented. FIG. 15 illustrates a relationship between a terminal and a server corresponding to the product illustrated in FIG. 14. Referring to FIG. 15A, the first terminal 500. 1 and the second terminal 500. It can be seen that the data to the bitstream can be bidirectionally communicated through the wired / wireless communication unit. Referring to FIG. 15B, it can be seen that the server 600 and the first terminal 500.1 may also perform wired / wireless communication with each other.

The audio signal processing method according to the present invention can be stored in a computer-readable recording medium which is produced as a program for execution in a computer, and multimedia data having a data structure according to the present invention can also be stored in a computer-readable recording medium. Can be stored. The computer readable recording medium includes all kinds of storage devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disks, optical data storage devices, and the like, which are also implemented in the form of carrier waves (for example, transmission over the Internet). It also includes. In addition, the bitstream generated by the encoding method may be stored in a computer-readable recording medium or transmitted using a wired / wireless communication network.

As described above, although the present invention has been described by way of limited embodiments and drawings, the present invention is not limited thereto and is intended by those skilled in the art to which the present invention pertains. Of course, various modifications and variations are possible within the scope of equivalents of the claims to be described.

The present invention can be applied to process and output audio signals.

1 is a block diagram of an encoder in an audio signal processing apparatus according to an embodiment of the present invention.

2 is a flowchart of an encoding method of an audio signal processing method according to an embodiment of the present invention.

3 is a diagram for explaining a quantization concept.

4 is a diagram for explaining a concept of a loss signal and a loss area.

5 is an example of syntax for encoding noise filling flag information.

6 is a diagram for explaining a noise level and a noise offset.

7 is an example of syntax for encoding noise level and noise offset.

8 is an example of syntax for encoding coding scheme information.

9 is a block diagram of a decoder in an audio signal processing apparatus according to an embodiment of the present invention.

10 is a detailed configuration diagram of the loss compensation part of FIG. 9;

11 is a flowchart of a decoding method of an audio signal processing method according to an embodiment of the present invention.

12 is an example of an audio signal encoding apparatus to which an audio signal processing apparatus according to an embodiment of the present invention is applied.

13 is an example of an audio signal decoding apparatus to which an audio signal processing apparatus according to an embodiment of the present invention is applied.

14 is a schematic configuration diagram of a product on which an audio signal processing device according to an embodiment of the present invention is implemented.

15 is a relationship diagram of products in which an audio signal processing device according to an embodiment of the present invention is implemented.

Claims (13)

Extracting noise filling flag information indicating whether noise filling is used for a plurality of frames; Extracting coding scheme information indicating whether a current frame including the plurality of frames is coded in a frequency domain or in a time domain; Extracting noise level information for a current frame when the noise filling flag information indicates that the noise filling is used in the plurality of frames and the coding information indicates that the current frame is coded in the frequency domain; Extracting noise offset information for a current frame when a noise level value corresponding to the noise level information satisfies a predetermined level; And, And if the noise offset information is extracted, performing noise filling on a current frame based on the noise level value and the noise level information. The method of claim 1, The noise filling, Determining a loss area of the current frame using the spectral data of the current frame; Generating compensated spectral data by filling a compensation signal in the loss region using the noise level value; And And generating a compensated scale factor based on the noise offset information. The method of claim 1, Extracting a level pilot value representing a reference value of the noise level and an offset pilot value representing the reference value of the noise offset; Obtaining the noise level value by adding the level pilot value and the noise level information; And, Obtaining the noise offset value by adding the offset pilot value and the noise offset information when the noise offset information is extracted; The noise filling is performed using the noise level value and the noise offset value. The method of claim 1, Acquiring a noise level value of the current frame by using a noise level value of a previous frame and noise level information of the current frame; Obtaining the noise offset value of the current frame by using the noise offset value of the previous frame and the noise offset information of the current frame when the noise offset information is extracted. The noise filling is performed by using the noise level value and the noise offset value. The method of claim 1, The noise level information and the noise offset information is extracted according to a variable length coding scheme. A multiplexer for extracting noise filling flag information indicating whether noise filling is used in a plurality of frames and coding scheme information indicating whether a current frame including the plurality of frames is coded in a frequency domain or a time domain; When the noise filling flag information indicates that the noise filling is used in the plurality of frames and the coding information indicates that the current frame is coded in the frequency domain, extracting noise level information for a current frame, and extracting the noise A noise information decoding part for extracting noise offset information for the current frame when a noise level value corresponding to the level information satisfies a predetermined level; And, And a loss compensation part for performing noise filling on the current frame based on the noise level value and the noise level information when the noise offset information is extracted. The method of claim 6, The loss compensation part, Determine the loss region of the current frame by using the spectral data of the current frame, By using the noise level value to fill a compensation signal in the loss region, to generate compensated spectral data And generating a compensated scale factor based on the noise offset information. The method of claim 6, Extract a level pilot value representing a reference value of the noise level, and an offset pilot value representing the reference value of the noise offset, Obtain the noise level value by adding the level pilot value and the noise level information, When the noise offset information is extracted, further comprising a data decoding part for obtaining a noise offset value by adding the offset pilot value and the noise offset information, The noise filling is performed by using the noise level value and the noise offset value. The method of claim 6, Obtain a noise level value of the current frame by using a noise level value of a previous frame and noise level information of the current frame, If the noise offset information is extracted, further comprising a data decoding part for obtaining a noise offset value of the current frame by using the noise offset value of the previous frame and the noise offset information of the current frame, The noise filling is performed by using the noise level value and the noise offset value. The method of claim 6, And the noise level information and the noise offset information are extracted according to a variable length coding scheme. Generating a noise level value and a noise offset value based on the quantized signal; Generating noise filling flag information indicating whether noise filling is used for the plurality of frames; Generating coding scheme information indicating whether a current frame included in the plurality of frames is coded in a frequency domain or a time domain; When the noise filling flag information indicates that noise filling is used for the plurality of frames and the coding scheme information indicates whether the current frame is coded in the frequency domain, the noise frame value of the current frame corresponding to the noise level value. Inserting noise level information into the bitstream; And And inserting noise offset information corresponding to the noise offset value into the bitstream when the noise level value satisfies a predetermined level. A loss compensation estimation part for generating a noise level value and a noise offset value based on the quantized signal, and generating noise filling flag information indicating whether noise filling is used in a plurality of frames; A signal classifier configured to generate coding scheme information indicating whether a current frame included in the plurality of frames is coded in a frequency domain or a time domain; When the noise filling flag information indicates that noise filling is used for the plurality of frames and the coding scheme information indicates whether the current frame is coded in the frequency domain, the noise frame value of the current frame corresponding to the noise level value. And inserting noise level information into the bitstream and inserting noise offset information corresponding to the noise offset value into the bitstream when the noise level value satisfies a predetermined level. Audio signal processing device. Extracting noise filling flag information indicating whether noise filling is used for a plurality of frames; Extracting coding scheme information indicating whether a current frame including the plurality of frames is coded in a frequency domain or in a time domain; Extracting noise level information for a current frame when the noise filling flag information indicates that the noise filling is used in the plurality of frames and the coding information indicates that the current frame is coded in the frequency domain; Extracting noise offset information for a current frame when a noise level value corresponding to the noise level information satisfies a predetermined level; And, When the noise offset information is extracted, performing the operations including performing the noise filling on the current frame based on the noise level value and the noise level information, when executed by the processor. The computer-readable storage medium on which this is stored.

Images