JP6544276B2 - Sound signal transfer method - Google Patents

Description

  The present invention relates to a technique for combining and transferring an audio signal and additional information related to the audio signal.

  Conventionally, among audio devices that simultaneously transmit a plurality of acoustic signals in one transmission path, for example, downmixing a multi-channel signal used in a movie or the like, specifically a 5.1 ch signal, etc. 2. Description of the Related Art There is one that transfers as a 2.1 ch signal (for example, Patent Document 1 etc.). The AV amplifier disclosed in Patent Document 1 is connected to each of the source device, the TV, and the speaker. When the audio of the source device is simultaneously output to the TV and the speaker, the AV amplifier notifies, for example, the number of reproducible channels of the source device to the source device, and inputs an audio signal corresponding to the number of channels from the source device. The AV amplifier outputs a downmixed sound signal to a TV having a small number of reproducible channels. In addition, the AV amplifier outputs the sound signal without changing the number of channels to the speaker having a large number of reproducible channels.

Patent No. 5531486 gazette

  However, in the above-described audio apparatus, advanced processing (such as data compression processing) is required at the time of downmix encoding and decoding, which may increase the processing load of signal processing. Also, the audio device on the receiving side may not be able to start the decoding process until it has received the downmixed signal by a certain amount of data and stored it in a buffer or the like. As a result, there may be a delay in the time to start the regeneration.

  This application is proposed in view of the above-mentioned subject, and an object of the present invention is to provide the transfer method of the acoustic signal which can combine and transmit a plurality of information to a limited acoustic channel band signal.

The method of transferring an acoustic signal according to the present application includes an AM modulation step of AM-modulating a carrier signal of a frequency in an inaudible band or an inaudible band in an audible band using additional information; an adding step of adding the modulated signal into an acoustic signal, have a, a transfer step of transferring a single data together with the audio signal and the additional information by transferring an addition signal, acoustic signal, first The signal includes the signal and the second signal, and in the addition step, the signal AM-modulated to the first signal is added, and the negative phase component of the signal AM-modulated to the second signal is added. characterized in that it further have a difference calculation step of calculating a signal and the difference of the second signal.

In the transfer method, the additional information is AM-modulated and added to the acoustic signal for transfer. In the AM modulation step, modulation is performed using a carrier signal of a frequency that is difficult or inaudible to the human ear. As a result, even if the acoustic signal added at the transfer destination is reproduced as it is, it can be heard with a sound without discomfort. For example, even if the type and performance of the audio device on the network are unknown, it is possible to reproduce the sound with a sense of incongruity even in the audio device which can not execute the demodulation processing by using the transfer method. Therefore, it becomes possible to combine and transfer a plurality of information in a limited acoustic channel band signal. Further, in the AM modulation method, the processing load is lighter than the encoding processing related to downmixing that has been performed in the conventional transfer processing, and the time and amount of accumulation of signals before processing in the audio device of the transfer destination are conventionally reduced. The processing load can be reduced in terms of memory usage as compared to the encoding processing of.
Further, in the transfer method, it is possible to remove the in-phase components of the first and second signals by calculating the difference between the first signal and the second signal at the transfer destination. In addition, the signal of AM modulation which is added to the first signal in phase and added to the second signal in reverse phase can be extracted as a signal of twice the amplitude of the original signal by calculating the difference. It becomes possible to increase the noise ratio (S / N ratio) to suppress the influence of noise.

  Further, in the acoustic signal transfer method according to the present application, the additional information may be a low band signal of the low band channel, and may further include a downsampling step of downsampling the low band signal.

  The low band signal is composed of only the low band component, so that it is possible to reproduce the sound with no sense of incongruity even if the sampling frequency is lowered. Therefore, by performing AM modulation using sample values obtained by downsampling a low band signal, it becomes possible to combine and transmit a plurality of acoustic signals in a limited acoustic channel band signal.

  Further, the method of transferring an acoustic signal according to the present invention may further include a moving average value calculating step of calculating a moving average value for the additional information extracted in the difference calculating step.

  In the transfer method, by calculating a moving average value for the additional information extracted by the difference operation, it is possible to cancel out the component with a small change for each sample among the acoustic signals included in the additional information. .

  According to the method of transferring an acoustic signal according to the present application, it is possible to combine and transmit a plurality of information in a limited acoustic channel band signal.

It is a figure showing the network composition of the AV system concerning an embodiment. It is a block diagram showing composition of AV amplifier of living. It is a block diagram which shows the connection relation of AV amplifier of a living, and AV amplifier of study. (A) is a figure which shows the amplitude value for 8 samples which sampled by 48 kHz 3 periods of the sine wave (carrier signal) of 18 kHz from which the magnitude | size of an amplitude becomes "1". (B) is a figure which shows the waveform of a carrier signal. It is a figure which shows the amplitude value before and after multiplying the carrier wave which made the amplitude value "1" a sine wave of 18 kHz. It is a figure which shows the amplitude value before and after multiplying the carrier wave which made the amplitude value "-0.3" by the sine wave of 18 kHz. (A) is a figure which shows the value before carrying out the moving average of each sample value of the carrier signal shown in FIG. 4 (a), and the value after computing a moving average. (B) is a figure which shows the waveform of the carrier signal after calculating a moving average. It is a case where calculation of a moving average is performed, and it is a figure showing the amplitude value after [before and] multiplying the 18 kHz sine wave by the carrier signal which set amplitude value to "1". It is a case where the calculation of a moving average is implemented, Comprising: It is a figure which shows the amplitude value after [before and] multiplying the 18 kHz sine wave by the carrier signal which made the amplitude value "-0.3". It is a figure which shows the sum total value of the amplitude value at the time of shifting the position which multiplies a carrier signal, and its amplitude value. It is a data structure of a packet, Comprising: An example of the data structure after expanding bit is shown. (A), (b) is a figure which expanded the acoustic signal of L, R channel up-sampled from 16 bits to 24 bits, and shows the state which set the acoustic signal of the other channel in the expansion area | region. (A), (b) is a figure which expanded the acoustic signal of L, R channel up-sampled from 16 bits to 32 bits, and shows the state which set the acoustic signal of the other channel in the expansion area | region. It is a figure which shows the data of each sample in the sound signal of L channel before raising sampling frequency (48 kHz) and after raising it (96 kHz). It is a table | surface which shows the relationship between each mode name and the gain value of each channel. It is a flowchart which shows the process for the transfer system by a control part to select.

  Hereinafter, an AV (Audio Visual) system 10 shown in FIG. 1 will be described as an embodiment of the present invention. FIG. 1 shows an example of the network configuration of the AV system 10 of this embodiment. In the AV system 10, a smartphone 11, a plurality of AV amplifiers 13 and 14, and a TV 17 are connected to a network 19. The network 19 is, for example, an in-home LAN (local area network) that mutually connects the AV amplifiers 13 and 14 and the TV 17 installed in a plurality of rooms (living 21, kitchen 22 and study 23) in one house. ). The AV amplifier 13 or the like executes, for example, communication conforming to a predetermined network protocol, and transmits / receives a packet P in which header information or the like is added to an audio signal via the network 19.

  For example, a dedicated application for controlling the AV amplifier 13 is installed in the smartphone 11. The user U in the living room 21 controls the AV amplifier 13 while operating the smartphone 11. The smartphone 11 stores various contents such as music data, and functions as a source device of the AV system 10 of the present embodiment. The source device is not limited to the smartphone 11, and may be, for example, a CD player or a personal computer, or a network storage such as NAS (Network Attached Storage). In addition, the source device may be a music wiring server on the Internet. Also, the file format of the music data may be, for example, MP3, WAV, SoundVQ (registered trademark), WMA (registered trademark), AAC or the like.

  Further, the smartphone 11 can be connected to, for example, the AV amplifier 13 installed in the living room 21 via wireless communication. The user U operates the smartphone 11 to transmit specified content, for example, music data D1 of 2.1 ch to the AV amplifier 13. For example, Bluetooth (registered trademark) can be adopted as a standard of wireless communication used by the smartphone 11. Also, the smartphone 11 may mutually communicate with the AV amplifier 13 via a router or the like connected to the network 19 by, for example, a wireless LAN of Wi-Fi (registered trademark) standard.

  The AV amplifier 13 of the living room 21 has, for example, a 2.1ch speaker connection terminal. The analog connection cable 31 connected to this terminal is connected to the 2.1 channel speaker 33 installed in the living room 21. The AV amplifier 13 reproduces the music data D1 received from the smartphone 11 from the speaker 33. The terminal for speaker connection provided in the AV amplifier 13 is not limited to the terminal for 2.1ch, and may be, for example, a terminal for 5.1ch or 7.1ch.

  Also, the AV amplifier 13 reproduces the same music data D1 received from the smartphone 11 from the TV 17 or the like. The AV amplifier 13 performs signal processing to convert the 2.1 ch music data D1 received from the smartphone 11 into music data D2 (for L channel) and D3 (for R channel) (see FIG. 3). The AV amplifier 13 transfers the packet P in which the converted music data D2 and D3 are set to the AV amplifier 14 and the like. The converted music data D2 and D3 are data having the same channel number (2.1 ch) as the music data D1. The details will be described later.

  The TV 17 installed in the kitchen 22 receives the packet P including the music data D 2 and D 3 from the AV amplifier 13 via the network 19. The TV 17 has built-in speakers 35 of L (left) and R (right) stereo 2ch. The TV 17 reproduces the music data D2 and D3 from the speaker 35.

  The AV amplifier 14 of the study 23 has a 2.1ch speaker connection terminal, for example. The analog connection cable 37 connected to this terminal is connected to the 2.1 channel speaker 39 installed in the study 23. The AV amplifier 14 receives the packet P including the music data D2 and D3 from the AV amplifier 13 via the network 19. The AV amplifier 14 reproduces the music data D2 and D3 from the speaker 39.

  The music data D2 and D3 described above are obtained by converting the music data D1. In the present embodiment, for example, in the living room 21, the music data D1 is output from the speaker 33 of 2.1 ch. In addition, for example, in the kitchen 22, the music data D2 and D3 are output as stereo music from the 2ch speaker 35 of the TV 17 as it is. In addition, for example, in the study 23, the music data D2 and D3 are demodulated and output from the speaker 39 of 2.1 ch.

  FIG. 2 is a block diagram showing the configuration of the AV amplifier 13 of the living room 21 and shows only portions particularly related to the present invention. As shown in FIG. 2, the AV amplifier 13 includes a wireless communication unit 41, an AM modulation unit 43, a Bit expansion unit 44, a frequency expansion unit 45, an interface unit 47, and a control unit 48.

  The wireless communication unit 41 extracts the music data D1 from the data received from the smartphone 11 via wireless communication. The music data D1 of the present embodiment includes, as an example, an audio signal of 2.1 ch in which a low frequency dedicated (LFE) channel is added to a stereo L (left) channel and an R (right) channel. In the case where the content does not include a low frequency dedicated channel, the low frequency component created based on the signal obtained by extracting the low frequency component from the stereo LR signal may be LFE. Also, as one example, the AV amplifier 13 of the present embodiment transfers an acoustic signal (an example of additional information) of the LFE channel in a 2-ch acoustic signal.

  The AM modulation unit 43, the Bit expansion unit 44, and the frequency expansion unit 45 execute signal processing to include the acoustic signal of the LFE channel in the acoustic signals of two channels of L and R according to each transfer method. The music data D2 and D3 subjected to signal processing by the AM modulation unit 43 or the like are transmitted as packets P from the interface unit 47 to the network 19. The control unit 48 is a device that performs overall control of the AV amplifier 13, and which processing unit of the processing units such as the AM modulation unit 43 is used based on a judgment criterion described later, that is, which transfer method Select whether to convert D1. The AM modulation unit 43, the Bit expansion unit 44, and the frequency expansion unit 45 can be realized, for example, by executing a predetermined program by a DSP (Digital Signal Processor) for sound processing. Further, the AM modulation unit 43 or the like may be realized by, for example, an analog circuit, or may be realized by executing a program on a CPU.

<About AM modulation method>
First, an AM modulation method by the AM modulation unit 43 will be described. FIG. 3 is a block diagram showing a connection relationship between the AV amplifier 13 of the living room 21 and the AV amplifier 14 of the study 23, and shows only a portion related to the AM modulation unit 43. As shown in FIG. As shown in FIG. 3, the AM modulation unit 43 includes two adders 51 and 52, a modulation processing unit 55, and a carrier generation unit 56. The adder 51 corresponds to the L channel, and inputs the acoustic signal of the L channel among the acoustic signals extracted from the music data D1 by the wireless communication unit 41. The adder 52 corresponds to the R channel, and inputs the acoustic signal of the R channel among the acoustic signals extracted from the music data D1 by the wireless communication unit 41. Further, the modulation processing unit 55 receives an acoustic signal of the LFE channel from the wireless communication unit 41. The acoustic signals of the L channel, R channel and LFE channel are, for example, acoustic signals sampled at 48 kHz.

  Here, since the acoustic signal of the LFE channel is a signal composed of only low frequency components, it becomes possible to reproduce with a sound without discomfort even if the sampling frequency is lowered. Therefore, the modulation processing unit 55 downsamples the acoustic signal of the LFE channel. The carrier generation unit 56 outputs the carrier signal CS sampled at 48 kHz to the modulation processing unit 55. The modulation processing unit 55 performs AM modulation on the carrier signal CS input from the carrier generation unit 56 using the down-sampled sample value, and outputs the modulated signal to the adders 51 and 52.

  More specifically, the carrier generation unit 56 outputs, as the carrier signal CS, a signal in a frequency band that is normally hard to hear by the human ear. As a result, a 2-channel audio device (for example, the TV 17) that does not support multi-channel (2.1 ch) playback does not have a sense of discomfort as 2-channel music data even if the received music data D2 and D3 are played back as they are It becomes possible to reproduce stereo sound by sound.

  As an example, the case where the audio signal of the LFE channel sampled at the sampling frequency of 48 kHz is downsampled by 1/8 will be described. When 1/8 down-sampling is performed on the original signal, one piece of data may be present every 8 samples as data used for AM modulation as compared to the original data. Therefore, a signal having an 8-sample period at 48 kHz, in other words, a signal of an integral multiple of 6 kHz (= 48 kHz / 8), a signal in a band hardly audible to the human ear is used as the carrier signal CS.

In the case of one cycle with eight samples: 48 kHz / 8 samples = 6 kHz
In the case of 2 cycles with 8 samples: (48 kHz / 8 samples) * 2 = 12 kHz
In the case of 3 cycles with 8 samples: (48 kHz / 8 samples) * 3 = 18 kHz
The 6 kHz and 12 kHz are in the audible band and are likely to be noise during reproduction. Therefore, among the signals having an 8-sample period of 48 kHz, for example, an 18-kHz sine wave including signals of three cycles every eight samples can be used as the carrier signal CS as a frequency band which is difficult to hear during reproduction.

  FIG. 4A shows amplitude values for eight samples obtained by sampling at 48 kHz three periods of an 18 kHz sine wave (carrier signal CS) whose amplitude magnitude is “1”. FIG. 4B shows the waveform of the carrier signal CS. The carrier generation unit 56 outputs the carrier signal CS of the sine wave shown in FIG. 4B to the modulation processing unit 55. The modulation processing unit 55 performs AM modulation of the amplitude (volume level) of the sample value obtained by downsampling the audio signal of the LFE channel input from the wireless communication unit 41 by 1/8 with the carrier signal CS input from the carrier generation unit 56. And output to the adders 51 and 52. Since this signal is an acoustic signal of 18 kHz, even if the signal is reproduced as it is on the reproducing side, it becomes an audio that is extremely hard to hear by the human ear.

  As shown in FIG. 3, the adder 51 adds the down-sampled and AM-modulated sound signal of the LFE channel to the sound signal of the L channel sampled at 48 kHz, and interfaces as an L channel signal (music data D2) Output to the part 47. Similarly, the adder 52 adds the down-sampled and AM-modulated sound signal of the LFE channel to the R-channel sound signal sampled at 48 kHz, and outputs it as an R-channel signal (music data D3) to the interface unit 47 Do. The interface unit 47 packetizes the music data D2 and D3 corresponding to each of the L and R channels input from the adders 51 and 52, and transfers it as a packet P to the AV amplifier 14 via the network 19.

  The interface unit 61 of the AV amplifier 14 receives the packet P from the interface unit 47 of the AV amplifier 13. The interface unit 61 extracts music data D2 corresponding to the L channel and music data D3 corresponding to the R channel from the received packet P. The interface unit 61 outputs the music data D2 corresponding to the L channel to the BEF (band elimination filter) 63. The BEF 63 is a filter that passes signals other than a signal of a predetermined frequency band according to the L channel. The BEF 63 outputs an acoustic signal from which an 18 kHz AM modulation component unnecessary as an L channel is removed to a speaker 39 corresponding to the L channel.

  Similarly, the interface unit 61 outputs the music data D3 corresponding to the R channel to the BEF 64. The BEF 64 is a filter that passes signals other than a signal of a predetermined frequency band according to the R channel. The BEF 64 outputs an acoustic signal from which an 18 kHz AM modulation component unnecessary as an R channel is removed to a speaker 39 corresponding to the R channel.

  Further, the interface unit 61 outputs the music data D2 and D3 corresponding to the L and R channels to the demodulation processing unit 67. For example, the demodulation processing unit 67 multiplies the 1/8 down-sampled signal of the audio signal contained in the input music data D2 and D3 by an 18 kHz sine wave to multiply the signal into an AM-modulated signal having a period of 18 kHz. Execute processing to extract the amplitude value.

  As an example, FIG. 5 shows that an amplitude value of 1/8 down-sampled is “1.0” with the signal transferred on the carrier signal CS before and after being transferred to the 18 kHz sine wave after transfer. The amplitude value is shown. As an example, FIG. 6 shows the amplitude after 1/8 down-sampling amplitude value is transferred to carrier signal CS as “−0.3” and transferred before and after being multiplied by the 18 kHz sine wave after transfer. It shows the value. As shown in FIG. 5, the sum “4” of the amplitude values after multiplication for each sample is four times the amplitude value “1” before multiplication. Similarly, as shown in FIG. 6, the sum "-1.2" of the amplitude values after multiplication for each sample is four times the amplitude value "-0.3" before multiplication. There is. Therefore, in this case, the amplitude value obtained by multiplying the sine waves is four times as large as the original amplitude value. Therefore, the original amplitude value can be extracted by multiplying the amplitude value by 1/4. Therefore, the demodulation processing unit 67 upsamples the sum of the amplitude values calculated by multiplying the input music data D2 and D3 by a sine wave by 1/4 and generates an LFE channel by eight times.

  Here, the following two problems can be considered in the above-described AM modulation method. First, there is a possibility that the 18 kHz band signal originally contained in the L and R channel acoustic signals may affect the AM-modulated signal as a noise component. Therefore, the demodulation processing unit 67 needs to take out only the AM-modulated signal so as not to be affected by the original L and R channel acoustic signals as much as possible. Second, it is difficult to detect the start position of the period of the AM-modulated signal by superimposing the AM-modulated signal on the waveform of the L and R channel acoustic signals which are the original signals. As a result, in the demodulation processing unit 67, it becomes difficult to detect the position at which the sine wave is multiplied to the music data D2 and D3, and there is a possibility that accurate demodulation can not be performed.

<Removal of common mode component>
Therefore, the AM modulation unit 43 of the AV amplifier 13 that is the transfer source adds the AM-modulated signal to the L and R channel acoustic signals according to the following rule. In a general music signal, there is a high possibility that many in-phase components such as vocal components are included as signal components of L and R channels. The in-phase component can be removed, for example, by subtracting the R channel acoustic signal from the L channel acoustic signal (Lch-Rch). Therefore, for example, the adder 51 adds the AM-modulated signal as the in-phase component to the L channel acoustic signal. Further, the adder 52 adds the AM-modulated signal to the R channel acoustic signal as an anti-phase component. When the in-phase component contained in L and R channels is “C” and the AM-modulated signal is “D”, the acoustic signals of L and R channels after addition are represented by the following equations.
Lch = C + D
Rch = C−D

The demodulation processing unit 67 of the AV amplifier 14 which is the transfer destination subtracts the sound signal of the R channel from the sound signal of the L channel (Lch-Rch) as represented by the following equation (1).
Lch-Rch = (C + D)-(C-D) = 2D (1)
As a result, the demodulation processing unit 67 can remove the in-phase component C and can take out only the AM-modulated signal D. Also, since the signal D taken out has an amplitude twice that of the original signal, the noise ratio (S / N ratio) is increased to suppress the influence of noise.

<About calculation of average value>
In addition, a general music signal may include many low frequency components and band components of human voice (for example, 1 kHz). The low-pass component and the like have a small variation in waveform for each sample. Therefore, in the demodulation processing unit 67 of the transfer destination, for example, as shown by the following equation, weighting is performed so that the preceding and following samples cancel each other, and the moving average value is calculated to obtain the original L, R channels. Remove the signal component of
Number of samples: Value before conversion → Value after conversion 1st sample: X → X * 0.5− (X + 1) + (X + 2) * 0.5
Second sample: X + 1 → (X + 1) * 0.5-(X + 2) + (X + 3) * 0.5
Third sample: X + 2 → (X + 2) * 0.5-(X + 3) + (X + 4) * 0.5
Fourth sample: X + 3 → (X + 3) * 0.5-(X + 4) + (X + 5) * 0.5
・
・
・

  The demodulation processing unit 67 converts, for example, each sample value of the monaural signal D extracted by the equation (1) according to the above-described weighting conversion equation. FIG. 7A shows a value before moving average of each sample value of the carrier signal CS shown in FIG. 4A and a value after calculating the moving average. Further, FIG. 7 (b) shows the waveform of the carrier signal CS after the calculation of the moving average. For example, the demodulation processing unit 67 calculates the moving average of the value of each sample of the extracted signal D using the above-described weighting conversion equation, and performs 1/8 down-sampling on the amplitude value of the above moving average. It is possible to extract the amplitude value of the 1/8 down-sampled original signal by transferring the transferred data on the carrier signal CS after calculation and multiplying it by the 18 kHz sine wave after moving average. Become.

  FIG. 8 shows the case where the moving average calculation is performed as an example, and after transferring the data transferred on the carrier signal CS after moving average with the amplitude value being “1.0”, 18 kHz after moving average is transferred. The amplitude values before and after the sine waves are shown. FIG. 9 shows an example in which moving average calculation is performed, and an amplitude value of “−0.3” is transferred to carrier signal CS after moving average and transferred after transfer data is transferred to 18 kHz after moving average. The amplitude value before and after multiplying the sine waves of is shown. As shown in FIG. 8, the sum value “11.65685425” of the amplitude values after multiplication for each sample is approximately 11.6 times the amplitude value “1.0” before conversion. . Similarly, as shown in FIG. 9, the sum value “−3.497056275” of the amplitude values after multiplication for each sample is about 11.6 times the amplitude value “−0.3” before multiplication. It has become. Therefore, when using a moving average value, the demodulation processing unit 67 uses a value obtained by multiplying the sum of the amplitude values calculated by multiplying the 18 kHz sine waves after the moving average by “1 / 11.65685425 times” and increases eight times. Sample and generate as LFE channel. In this manner, the demodulation processing unit 67 removes the components of the sound signals of the L and R channels to reduce the influence of the original signal as a noise component, thereby solving the first problem described above. .

<About position detection of AM modulated signal>
As the second problem described above, it is important to detect the start position of the period of the AM-modulated signal, but the detection and setting of the start position are performed by comparing the averaged values. The waveform of the carrier signal CS always has the same shape every eight samples. Therefore, the demodulation processing unit 67 calculates the sum of the amplitude values every eight samples from an arbitrary position, and sets the sample position at which the absolute value of the sum is the largest as the start position.

  FIG. 10 shows the sum of the amplitude value and the amplitude value when the position where the carrier signal CS is multiplied is shifted. Further, FIG. 10 shows an amplitude value in the case where data of which the amplitude value is "1.0" is transferred using the carrier signal CS, and a moving average operation is performed after the transfer. As shown in FIG. 10, when sample positions 0 to 7, that is, when the start position of the arithmetic operation of addition is coincident with the start position of the waveform, the absolute value of the sum of the amplitude values is the maximum value (11.65685425 ). On the other hand, when the sample start position at which the sum value is calculated shifts by one sample to the sample positions 1 to 8, the absolute value of the sum of the amplitude values is a smaller value (8.242640687) than the maximum value. Thereby, the demodulation processing unit 67 sets the position at which the absolute value of the sum of the amplitude values is maximum as the start position of the period of the AM-modulated signal, whereby the music data D2, D3 or the signal converted into monaural It becomes possible to appropriately set the position where the sine wave is multiplied to D.

  In addition, as shown in FIG. 10, also in the case of the sample positions 4-11 which shifted 4 samples from the sample positions 0-7, the absolute value of the sum total of an amplitude value becomes a maximum value (11.65685425). In this embodiment, since the LFE signal to be subjected to AM modulation is a low-pass component and the difference for each sample is small, the sine wave is multiplied regardless of which of sample position 0 and sample position 4 is set as the start position. The error as a signal after the signal is small. Also, for example, if the original signal before AM modulation is set to a positive value in advance, the maximum value of the positive value can be detected as the start position. More specifically, for example, the modulation processing unit 55 of the transfer source is the carrier signal CS with an amplitude of “−1.0 to +1.0”, and the entire waveform is “amplitude * 0.5 + 0.5” and positive. It will be a value. The demodulation processing unit 67 of the transfer destination sets a position at which the maximum value of the sum of the amplitudes is positive as a start position, and calculates "(amplitude-0.5) * 2.0" for the signal after demodulation. It is possible to extract the signal of the LFE channel by performing inverse conversion.

<About upsampling>
Although the above description has described the case where the modulation processing unit 55 performs AM modulation on the carrier signal CS of 18 kHz, the present invention is not limited to this. For example, the modulation processing unit 55 may AM-modulate the acoustic signal of the LFE channel using the carrier signal CS in a frequency band higher than the audible band and add it to the acoustic signals of the L and R channels.

  If it is possible to upsample the acoustic signal of the L and R channels sampled at 48 kHz to four times to 192 kHz, the acoustic signal of the 1/8 down-sampled LFE channel is 192 kHz at the same time as described above. Of the signals (signals of integral multiples of 24 kHz (= 192 kHz / 8)) having an 8-sample period, a signal (for example, 72 kHz = 24 kHz * 3) higher than the audible band may be AM-modulated using the carrier signal CS. it can. In this case, when the original signal does not include a high frequency component such as 192 kHz, the original signal is not affected as noise. In addition, it is possible to separate the channels only by using the high pass filter or the low pass filter without performing the subtraction processing (Lch-Rch) and the calculation of the moving average value described above. Also, for example, if a plurality of adjacent frequencies can be used as the carrier signal CS in a high frequency band, a multichannel acoustic signal such as 5.1 ch is AM-modulated using each carrier signal CS and the high band Can be included and transferred.

<About Bit extension method>
Next, the Bit expansion method by the Bit expansion unit 44 (see FIG. 2) will be described. The Bit expansion unit 44 mixes and transfers a plurality of channel signals using an empty area of the quantization bit of the acoustic signal. For example, music content of a CD (Compact Disc) is usually quantized by 16 bits. Also, in general, when the sound signals of L and R channels quantized by 16 bits are expanded to 24 bits and transferred, a value of “0” is set to the minimum 8 bits. Therefore, when expanding each of the 16-bit-quantized L and R channel audio signals to 24 bits, the Bit expansion unit 44 transfers the audio signals of other channels using a minimum of 8 bits. The minimum 8 bits are relatively small as the sound volume (sound pressure level). Therefore, even if audio signals of other channels are set and reproduced as 24 bits, the volume area is hard to hear by the human ear, and it is possible to reproduce a sound with less discomfort at the transfer destination.

  FIG. 11 shows the data structure of the packet P transferred on the network 19, and shows an example of the data structure after bit extension. The bit expansion unit 44 can transfer the acoustic signals of the L and R channels quantized by 16 bits in 24 bits among the acoustic signals extracted from the music data D1 by the wireless communication unit 41 (see FIG. 2). Perform extended processing. In addition, the Bit expansion unit 44 adds an acoustic signal of, for example, the LFE channel to the minimum 8-bit data area that is increased by the expansion from 16 bits to 24 bits and transfers it. When the sound signal of the LFE channel is quantized by 16 bits, the Bit expansion unit 44 sets upper 8 bits in the expansion area of the sound signal of L channel as shown in FIG. 11, and interfaces as music data D2 Output to the part 47. In addition, the Bit expansion unit 44 sets lower 8 bits in the expansion area of the R channel audio signal, and outputs the lower 8 bits as the music data D2 to the interface unit 47. The interface unit 47 packetizes and transfers the music data D2 and D3 in the same packet P, for example.

  The audio device of the transfer destination performs processing according to the number of available channels. In the TV 17 in which the 2ch speaker 35 is built in, for example, the bit value of the extension area of the sound signal of the L and R channels extracted from the packet P is cleared to zero and output to the speaker 35. Alternatively, the TV 17 sets a dither signal (uncorrelated noise) to the bit value of the extension area and outputs the same to the speaker 35. Thus, the speaker 35 can reproduce the sound of the L and R channels of the music data D2 and D3 (music data D1). Also, even if the TV 17 is a device that does not support the invalidation processing of the extension area, as described above, the minimum 24 bits of 24 bits are reproduced as they are because the volume area is hard to hear by the human ear. However, the effect of noise is considered to be extremely small.

  Also, in the AV amplifier 14 to which the speaker 39 of 2.1 ch is connected, upper 8 bits and lower 8 bits of the audio signal of the LFE channel are extracted from the packet P, for example, as processing for reproducing the audio signal of the LFE channel. Also, the AV amplifier 14 combines the upper 8 bits and the lower 8 bits of the extracted LFE channel audio signal, and generates a lower audio signal quantized by 16 bits. The AV amplifier 14 outputs the generated acoustic signal of the LFE channel to the speaker 39. Further, as a process for reproducing the acoustic signals of the L and R channels, the AV amplifier 14 clears the extension areas of the acoustic signals of the L and R channels taken out of the packet P and outputs the same to the speaker 39 as in the TV 17. . In this Bit extension method, acoustic signals of a plurality of channels can be included in the same packet P, and since the number of samples can be aligned and included in the same packet P and transferred, the sound output timing of each channel can be aligned. Becomes easy.

<On the application of Bit extension method>
Next, the case where upsampling is performed in the above-described Bit expansion method will be described. The Bit expanding section 44 enlarges the above-mentioned expanded area (vacant area) by raising the sampling frequency, and simultaneously transmits audio signals and the like of more channels by mixing other signals in the expanded area. It is possible. For example, the case of upsampling to 192 kHz the acoustic signals of L and R channels sampled at 48 kHz will be described.

  FIG. 12A shows a state in which the up-sampled L channel sound signal is expanded from 16 bits to 24 bits, and the sound signals of other channels are set in the expanded area. FIG. 12B shows a state in which the up-sampled R channel sound signal is expanded from 16 bits to 24 bits, and the sound signals of other channels are set in the expansion area. As shown in FIGS. 12A and 12B, the data amount of the signal up-sampled to 192 kHz is four times as large as that of the original 48-kHz signal. Therefore, the data area of the expanded quantization bit is also quadrupled. If, for example, sound signals of other channels sampled at 48 kHz and quantized by 16 bits are set in this four-fold expansion area, sound signals of other channels can be arranged every four samples. . In other words, it becomes possible to set four types of signals of other channels quantized by 16 bits in the extension area.

  In the example shown in FIGS. 12 (a) and 12 (b), in the extension area of the first (first sample) L and R channels from the top, the upper and lower portions of the acoustic signal of the other channel (ch1 in the figure) 8 bit is set. Similarly, upper and lower 8 bits of the audio signals of ch2, ch3 and ch4 are set in the second (second sample) and subsequent expansion areas. In this case, it is possible to transfer a total of 6 channels obtained by adding 4 channels of the extension area to the original L and R channels (2 channels). In addition, as processing of the transfer destination, processing to make the sampling frequency uniform is required. For example, the AV amplifier 14 of the transfer destination upsamples the sound signals of CH1 to CH4 channels in the extended area from 48 kHz to 192 kHz, or downsamples the sound signals of L and R channels from 192 kHz to 48 kHz to obtain sampling frequencies Need to be aligned.

  Further, for example, when the signal quantized by 16 bits can be expanded to 24 bits or more, for example, 32 bits, it becomes possible to mix and transfer sound signals of more channels. FIGS. 13A and 13B show the data structure of the packet P when expanded to 32 bits. In this case, it is possible to secure a 16-bit data area (16th to 32nd bits) in one extension area corresponding to the 16-bit data of the L and R channels. As shown in FIGS. 13 (a) and 13 (b), in the extension area of the first L channel from the top, both high-order and low-order (16 bits) of the sound signal of ch1 other than L and R channels are set It is done. Similarly, in the extension area of the first R channel from the top, both high-order and low-order (16 bits) of the acoustic signal of ch 2 are set. In this case, it is possible to transfer a total of 10 channels obtained by adding 8 channels of the extension area to the original L and R channels (2 channels). Thus, the Bit expansion unit 44 can expand the number of Bits and increase the number of channels that can be set in the expansion area.

<About sampling frequency extension method>
Next, the sampling frequency expansion method by the frequency expansion unit 45 (see FIG. 2) will be described. The frequency expansion unit 45 increases the sampling frequency to secure a vacant area between data, and mixes and transfers a plurality of channel signals using the reserved vacant area. For example, when the sampling frequency of the sound signal of L and R channels is 48 kHz, the frequency expanding unit 45 doubles the sampling frequency to 96 kHz. In the case of normal upsampling, the increased sample is set with a sample value obtained by newly sampling the original signal. However, the frequency expansion unit 45 of the present embodiment maintains 48 kHz data as it is without resampling, and sets data different from the original acoustic signal in the increased sample portion. This makes it possible to mix the signals of the other channels with the acoustic signals of the L and R channels.

  FIG. 14 shows data of each sample in the L channel acoustic signal before (48 kHz) and after (96 kHz) sampling frequency increase. As shown in FIG. 14, the frequency expanding unit 45 raises the sampling frequency from 48 kHz to 96 kHz which is doubled, and secures “empty samples 1 to 4” between the samples. The frequency expansion unit 45 can transfer data of twice the number of channels as signal data by inserting data of another channel (LFE channel or the like) into the vacant samples 1 to 4. Although only the data of the L channel is illustrated in FIG. 14, the same processing can be performed for the sound signal of the R channel, so that it is possible to transfer twice the number of channels.

  In this case, in each of the L and R channels, a vacant area for one channel can be secured as a data area for setting an acoustic signal sampled at 48 kHz. For this reason, the frequency expansion unit 45 can transfer data for four channels in total of the L and R channels (two channels) combined with two additional channels. In the transfer destination AV amplifier 14, for example, each channel can be individually acquired by extracting data of a different channel every other sample from the packet P.

  In the above sampling frequency expansion method, the sampling frequency is increased only during transfer. Also, the AV amplifier 14 can restore the original 2.1 ch music data D1 without changing the sampling frequency from 96 kHz to the original 48 kHz without resampling processing for the acquired data. . Also, in the sampling frequency extension method, data of a plurality of channels are transferred in a sample-by-sample manner, unlike the normal up-sampling processing. For this reason, since acoustic signals of a plurality of channels are divided and transferred for each sample, it is possible to secure a high transfer rate and sound quality as compared with the above-described AM modulation method and Bit extension method.

<About sending metadata>
In the above example, in the three transfer methods such as the AM modulation method, the sound signal of the LFE channel is mixed with the sound signal of the L and R channels and transferred, but the mixed data is not limited to the sound signal, but metadata (text Data, control data, etc.) may be used. For example, the AV amplifier 13 may transfer control data for changing the gain as the mixed control data. Here, in general, in processing of an acoustic signal, processing for securing a head margin is required as pre-processing for executing processing of a digital area in a DSP or the like, and processing for returning head margin as pre-processing for reproducing in an analog area is It will be necessary. The AV amplifier 13 executes preprocessing for securing a head margin of -10 dB, for example, in order to prevent a clip from being generated in the digital domain with respect to an acoustic signal of the LFE channel of 0 dB full scale. The AV amplifier 13 transmits the amount of head margin (−10 dB) previously attenuated in the digital domain as control data to the audio device of the transfer destination (for example, a subwoofer that reproduces only the LFE channel). In the transfer destination subwoofer, it is possible to amplify the acoustic signal of the LFE channel by +10 dB in the processing of the analog domain based on the control data, and align and reproduce the signal level with the other L and R channels. This makes it possible to avoid the generation of clips in processing in the digital domain and to transfer with higher sound quality. As described above, in the audio device of the present embodiment, it is possible to transmit metadata such as control data in addition to or instead of the acoustic signals of a plurality of channels.

  Further, the AV amplifier 13 may mix and transfer control data relating to gain adjustment of a specific channel according to the request of the user U, and change the reproduction state of the transfer destination. FIG. 15 is an example of a table showing the relationship between each mode included in the AV amplifier 13 and the gain value of each channel. For example, the AV amplifier 13 sets gain values corresponding to the respective modes shown in FIG. 15 as control data, and mixes the sound signals with 5.1-channel multi-channel acoustic signals according to the above-described transfer methods. The audio device of the transfer destination, for example, downmixes the received 5.1 ch sound signal to 2 ch and reproduces it. The audio device of the transfer destination realizes the reproduction according to each mode by increasing or decreasing the signal level of each channel based on the gain value set in the control data.

  As shown in FIG. 15, gain values for each channel are set in the control data. Channel names L, C, R, SL, SR, and LFE in FIG. 15 respectively indicate left (left), center, right (right), surround left, surround light, and low-range channels. Further, the gain value “1.0 times (attenuation amount 0 dB)” is a signal level for reproducing normal music.

  In the karaoke mode, the center ch (Cch) containing a large amount of vocal components is muted “0 times (attenuation amount − − dB)” and downmixed to suppress vocal sound and reproduce karaoke-like sound. Yes (see the bold portion in FIG. 15). As shown in FIG. 15, in the surround channels SL and SR, the gain value is “0.7 times (attenuation amount −3 dB)”. This is because the surround channels SL and SR need to be multiplied by 0.7 (attenuation amount −3 dB) for level adjustment, for example, when downmixing 5.1 ch to 2 ch.

  Also, in the front priority mix mode, the front side (Lch, Cch, Rch) is downmixed as usual by “1.0 times (0 dB attenuation)”, but the surround side (SLch, SRch) is reduced “0.5 "(Attenuation -6 dB)". As a result, the sound reproduced from the speaker at the transfer destination suppresses the surround sound including many voices of the audience, and emphasizes the components such as the singing voice of the vocals and the performance sound of the performer to make the front side easy to hear. It becomes a sound.

  In the night listening mix mode, the signal levels of Lch, Rch, and LFEch containing many loud signals and low frequency components are lowered, and the signal level of Cch containing many vocal singing components is raised. For example, the signal levels of Lch and Rch are multiplied by 0.7, and the signal level of LFEch is multiplied by 0.3. Also, for example, the signal level of Cch is multiplied by 1.4. Thereby, in the night audition mix mode, for example, even if the volume level is reduced at night to reproduce music, raising the signal level of Cch makes it easy to hear human voice and suppresses low-frequency components. It becomes possible to suppress that the vibration etc. accompanying reproduction of music become a nuisance of the neighborhood.

  As described above, by adjusting the signal level of the channel using control data (metadata), it is possible to match the sound of the transfer destination to the preference of the user U. The change and setting of each mode may be changed, for example, by the user U operating the remote controller of the AV amplifier 13 or the operation button of the housing. Further, the control unit 48 (see FIG. 2) of the AV amplifier 13 includes, for example, a data table in which the gain values in the table shown in FIG. 15 are preset, in a memory or the like, and referring to the data table The signal level according to may be set as control data.

  Also, the AV amplifier 13 may set a time stamp indicating the reproduction time of the music data D1 as metadata, and mix it with the sound signal of the L, R channels and the like. As a result, it is possible to align the timing of sound output between the transfer source and the transfer destination.

<About transfer of downmixed sound signal>
Further, in each of the above-described transfer methods, not only normal 2ch sound signals but also signals obtained by downmixing a conventionally used multichannel to 2ch can be similarly transferred. For example, the AV amplifier 13 can also mix and transfer 5.1 ch signals according to each of the above transfer methods to the L and R channel sound signals downmixed to 2 ch. In this case, when the audio device of the transfer destination is a stereo speaker, it is possible to reproduce the downmixed 2ch signal. When the transfer destination is a speaker corresponding to a multichannel, the downmixed signal can be discarded, and the multichannel signal (5.1 ch) included in the received signal can be separated and reproduced.

<On selection of transfer method>
Next, processing for selecting the above-described three transfer methods will be described. The control unit 48 (see FIG. 2) of the AV amplifier 13 is, for example, a “priority item” at the time of transferring the music data D1 to each audio device (AV amplifier 14 etc.) and “processing performance related to the music data D1 (AV amplifier The appropriate transfer method can be determined based on the processing performance of the network equipment etc. The control unit 48 may determine the transfer method based on one of the priority items and the processing performance. In the following example, although the case where the same transfer method is applied to all the audio devices connected to the network 19 will be described, the transfer method may be determined for each audio device.

  For example, when the transfer of music data D1 is started, control unit 48 weights the transfer method according to the flowchart shown in FIG. 16 (see S11 to S13 in FIG. 16), and selects the transfer method based on the result. (See S14 in FIG. 16). First, in step (hereinafter simply referred to as “S”) 11, the control unit 48 weights the transfer method according to the performance of the transfer destination audio device. In S11, the control unit 48 determines the performance of each audio device connected to the network 19. Regarding the determination of the performance, the control unit 48 may make a determination based on, for example, the result of inquiring each audio device via the network 19, or even based on the information input from the user U. Good. In addition, the control unit 48 may not directly inquire about the processing performance related to the music data D1. For example, the control unit 48 may acquire only the performance information of the CPU of each audio device, and estimate the processing performance related to the acoustic signal based on the information.

  For example, when there is a device (such as a single speaker device) with low processing performance in the network 19, the separation processing of the channel by the above-described demodulation processing unit 67 (see FIG. 3) can not be performed by the transfer destination device. is assumed. If the separation processing can not be performed, an AM modulation method or a Bit extension method which can reproduce with a sound without discomfort even if it is reproduced without separation processing is effective. Therefore, when the control unit 48 determines that a device with low processing performance is present on the network 19, the control unit 48 raises the priorities of the AM modulation scheme and the Bit extension scheme.

  On the other hand, if the processing performance of all the audio devices connected to the network 19 is high, advanced processing can be performed at the transfer destination. If advanced processing can be performed, a sampling frequency extension method that can maintain high quality sound quality with the least loss of data in transfer processing is effective. Therefore, if the control unit 48 determines in S11 that the processing performance of all the audio devices connected to the network 19 is high, the control unit 48 raises the priority of the sampling frequency expansion method. Even when all the audio devices have high performance, transfer by the AM modulation method or the Bit extension method can be performed.

  Next, at S12, the control unit 48 weights the transfer method in accordance with the number of channels of the music data D1 to be transferred and the contents (priority items). With regard to detection of the number of channels and the like, for example, the control unit 48 can directly detect the number of channels of the music data D1 to be transferred, or detect based on input information of the user U or the like. In S12, for example, if the music data D1 is music content in which a band-limited LFE channel such as 2.1 ch is added to the basic front side 2 ch, or if the music data D1 is a basic 2 ch to the user In the case of music content added with a relatively unnoticeable sound quality such as an announcement signal (mail arrival notification), high quality (sampling frequency) is not required, so that the control unit 48 prioritizes AM modulation, for example. Raise the degree.

  When the music data D1 is 3ch or 4ch music content in which 1ch to 2ch of full band is added to 2ch of basic data, the control unit 48, for example, raises the priority of the Bit expansion method. Also, when the music data D1 is multi-channel 5.1ch or 7.1ch music content in which 3 channels or more of the full band is added to the basic 2ch, the control unit 48 performs, for example, sampling capable of high quality transfer Raise the priority of the frequency expansion method. Thus, the control unit 48 can select the transfer method according to the number of channels of the music data D1 and the content of the signal (such as sound quality). The setting of the priority described above is an example, and for example, the sampling frequency expansion method may be used even if it is 2.1 ch.

  Next, in S13, the control unit 48 weights the transfer method in accordance with the operation content (priority matter) of the user U with respect to the remote controller of the AV amplifier 13 and the operation button of the housing. For example, the user U operates the remote control or the like to select one of the three items of “reduction of power consumption of transfer destination”, “reduction of delay between multiple channels”, and “priority of high resolution sound quality”. You can choose

  Here, in the AM modulation method and the Bit extension method, since it is possible to reproduce the acoustic signal of the L and R channels as it is, if it is desired to reduce the power consumption, the separation processing is stopped at the transfer destination and the reproduction is performed as it is. By doing this, the power consumption required for the separation process can be reduced. Therefore, when the user U selects “reduction of power consumption of transfer destination”, the AM modulation method and the Bit extension method which can select the presence / absence of separation processing according to the power consumption become effective. Therefore, when “reduction of power consumption of transfer destination” is selected, the control unit 48 raises the priorities of the AM modulation method and the Bit extension method.

  Further, in order to reduce the delay between a plurality of channels, more specifically, to output sounds simultaneously from nearby speakers, the Bit expansion method which makes it easy to align the output timing of each channel is effective. Therefore, when the user U selects “reduction of delay between multiple channels”, the control unit 48 raises the priority of the Bit extension method.

  In addition, when the user U wants to prioritize sound quality, a sampling frequency expansion method capable of higher quality transfer is effective. Therefore, the control unit 48 raises the priority of the sampling frequency expansion method when “priority of high resolution sound quality” is selected by the user U.

  Next, in S14, the control unit 48 selects a transfer method based on the result of the weighting performed in S11 to S13. Thereby, the control unit 48 can transfer the music data D1 by an appropriate method by selecting the transfer method from among the three transfer methods according to the priority matters and the processing performance. In addition, weighting may be applied to each of the above-described processes of S11 to S13. For example, the control unit 48 lowers the priority in the order of S11 (processing performance), S12 (number of channels, etc.), S13 (user selection) (S11> S12> S13), and the setting of more emphasizing the determination result of S11 Good.

  Incidentally, the music data D1, D3 and D3 are examples of acoustic signals. The acoustic signal of the L channel is an example of the first signal. The acoustic signal of the R channel is an example of the second signal. The acoustic signal and metadata of the LFE channel are an example of the additional information.

As mentioned above, according to the above-mentioned embodiment, the following effects are produced.
In the AM modulation method or Bit extension method, it is assumed that the audio device of the transfer destination (for example, the TV 17) is a device not compatible with the transfer method and that the L and R channel audio signals mixed with the LFE channel audio signals are reproduced as they are It becomes possible to reproduce with a sound without a sense of incongruity. In the network 19 to which the AV system 10 is applied, there is an audio device provided with a rich DSP such as the AV amplifier 14 while on the other hand a speaker device alone merely reproduces the received music data There is also. In such a case, the above-described transfer method can reproduce the original 2ch music only by simple processing without requiring high processing performance in the transfer destination audio device. Therefore, it is possible to appropriately transfer data in which a plurality of signals are mixed within a limited voice band between audio devices of different generations, performances, purposes, solutions, and the like.

  In addition, since the above three transfer methods are relatively easy in processing content as compared to the encoding processing related to downmixing that has been performed in the conventional transfer processing, even audio devices that do not support the transfer method It is possible to cope with simple firmware update and the like.

The present invention is not limited to the above embodiment, and it goes without saying that various improvements and changes can be made without departing from the spirit of the present invention.
For example, in the above embodiment, the acoustic signal of the L and R channels is mixed with the acoustic signal of the L and R channels, but the mixed acoustic signal is not limited to the acoustic signal of the LFE channel. . Further, the signals to be mixed are not limited to the sound signals of L and R channels, but may be sound signals of surround left (SL) channel and center (C) channel.
Further, in the above embodiment, the AV amplifier 13 may change the transfer method for each transfer destination audio device. For example, the AV amplifier 13 may perform transfer with the AV amplifier 14 and the Bit extension method while transferring with the TV 17 using the AM modulation method.

  DESCRIPTION OF SYMBOLS 10 AV system, 13 AV amplifier, 33, 35, 39 Speaker, 47 interface part, 43 AM modulation part, 51, 52 adder, 55 modulation processing part, 56 carrier production | generation part, CS carrier signal, D1, D2, D3 music data.

Claims (3)

An AM modulation step of AM-modulating a carrier signal of a frequency in an inaudible band or in an inaudible band within the audible band, using additional information;
An addition step of adding the AM modulated signal to the acoustic signal;
Have a, a transfer step of transferring as the acoustic signal and the additional information and are collectively one data by transferring the sum signal,
The acoustic signal includes a first signal and a second signal,
In the addition step, the AM-modulated signal is added to the first signal, and an anti-phase component of the AM-modulated signal is added to the second signal,
In the transfer destination, the transfer method of an audio signal, wherein a difference calculation step of calculating a difference between said first signal and said second signal, and further perforated. The additional information is a low band signal of a low band channel,
The method of claim 1, further comprising a downsampling step of downsampling the low band signal. The method according to claim 1 or 2 , further comprising a moving average value calculating step of calculating a moving average value for the additional information extracted in the difference calculating step.

Images