JP2012244336A - Audio signal processing device, audio signal processing method and acoustic reproduction device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain good listening quality for a reflected sound of a sound which was output from a speaker and then thrown back to the listener upon reflection at a surrounding object.SOLUTION: A cancellation audio signal is generated based on an input audio signal Di. Then, the cancellation audio signal is added to the input audio signal to obtain an output audio signal Di". The cancellation audio signal is used to cancel, among reflected sounds reaching a listening point after being output from a speaker, a prescribed second reflected sound (sound to be a sub) which reaches the listening point after a first reflected sound (sound to be the main) has arrived. Since the output audio signal is supplied to the speaker, the reflected sounds reaching a listening point after being output from the speaker become to include a reflected sound which cancels the second reflected sound. As a result, the second reflected sound is canceled at the listening point. The speaker is installed face down on the rear side of the cabinet of a display apparatus such as a television receiver, etc.

Description

  The present technology relates to an audio signal processing device, an audio signal processing method, and an acoustic reproduction device. In particular, the present technology relates to a sound signal processing device and the like that can be applied to a sound that is received after a sound output from a speaker is reflected by a surrounding object, that is, a reflected sound.

  In recent years, display devices such as plasma and liquid crystal have been widely used, and television receivers are becoming thinner and larger. In such a television receiver, the speaker size is greatly restricted due to the reduction in thickness, and the installation of speakers at both ends of the screen has become difficult due to the enlargement and narrowing of the screen. As a result, the speaker installed in the television receiver is often small in size, moved to the rear side, and further installed downward (see, for example, Patent Document 1).

  Therefore, the sound that reaches the user is no longer a very good sound. Note that “good sound” here is not an individual's subjective sound, but a speaker with a physically ideal characteristic, that is, a flat frequency characteristic. It is assumed that the viewer listens to the music by placing them on the screen. That is, it is assumed that a “good sound” is realized when the frequency characteristics of the recorded sound to be reproduced can be received as much as possible.

  Now, it is easy to imagine that most people will not be able to hear sound well just because the speaker is facing downward. Here is an objective example of the cause. That is, since the directivity of the low frequency component of the sound is not sharp, it is easy to reach the user even if the speaker is facing downward, so that it is relatively easy to hear. However, the higher the component, the stronger the directivity, making it difficult for the user to hear. For this reason, the sound is a so-called “muffled sound” with no high range. As measures against this, efforts are made to achieve a “good sound” by eliminating the “muffled sound” by implementing signal processing that raises the high range to achieve a sufficiently high frequency reproduction by making the frequency characteristics close to flat. Has been made.

JP 2010-239249 A

  However, even if the frequency characteristics are adjusted by signal processing, it is unclear at the design stage how the television is installed in the user environment. For example, when a speaker is mounted downward on a housing of a television receiver, the designer cannot grasp what frequency characteristics are present at the user's listening point. In this case, the frequency characteristics vary greatly depending on whether the television receiver is placed on a low television stand or on a television stand having a height of about 1 m.

  After all, at the design stage, it is impossible to clearly determine the characteristics of the inverse filter for flattening the frequency characteristics, such as how much high frequency should be increased in signal processing. For this reason, the designer must “guess” the user's environment or decide the contents of the signal processing based on “assuming” assuming a specific environment. Therefore, even if the signal processing function is implemented in the television receiver, it is unclear whether the user can receive the benefit.

  Further, flat adjustment is difficult only by adjusting the frequency characteristics of two points such as a low range and a high range as a whole, and a “good sound” cannot be realized. If the frequency characteristics are adjusted roughly even in the low and high frequencies, the sound quality will improve somewhat. However, there are many cases where minor sound quality problems such as “difficult to hear sound” and “balance of each instrument” are not easily improved.

  An object of the present technology is to make it possible to satisfactorily listen to a reflected sound that reaches a user after sound output from a speaker is reflected by a surrounding object.

The concept of this technology is
Based on the input audio signal, a canceling audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound among the reflected sounds that are output from the speaker and reaches the listening point An audio signal generation unit for cancellation to be generated;
An audio signal processing apparatus comprising: an audio signal adding unit that adds the cancellation audio signal generated by the cancellation audio signal generation unit to the input audio signal to obtain an output audio signal.

  In the present technology, a cancellation audio signal is generated by the cancellation audio signal generation unit. The audio signal adding unit adds the cancel audio signal to the input audio signal to obtain an output audio signal. The cancellation audio signal is for canceling a predetermined second reflected sound that arrives after the first reflected sound is reached among the reflected sounds that are output from the speaker and reach the listening point. The speaker is attached downward on the back side of the housing of a display device such as a television receiver. In this case, the user listens to the reflected sound that is output from the speaker and reaches the listening point. This reflected sound is a sound that arrives after the sound output from the speaker is reflected by surrounding objects.

  As described above, in the present technology, the output audio signal is obtained by adding the cancellation audio signal to the input audio signal. Therefore, by supplying the output audio signal to the speaker, the reflected sound that is output from the speaker and reaches the listening point includes the reflected sound that cancels the second reflected sound, and as a result, The second reflected sound is canceled at the listening point. Therefore, the user can listen to the first reflected sound satisfactorily at the listening point without being affected by the second reflected sound.

  In the present technology, for example, the cancellation audio signal generation unit includes a delay unit that delays the input audio signal by a time corresponding to a time difference between the first reflected sound and the second reflected sound that reaches the listening point. And a gain adjusting unit that adjusts the gain of the output signal of the delay unit according to the amount of reflection to obtain a canceling audio signal.

  In the present technology, for example, a user interface for a user to input information about a distance from a speaker installation position to a wall as information for adjusting a delay amount in the delay unit, and a distance input through the user interface And a control unit that controls a delay amount in the delay unit based on the information. In this case, the user interface includes a display unit that performs GUI display for the user to input distance information, and a user input unit for the user to input distance information using the GUI display displayed on the display unit. It may be made to have. By providing the user interface in this way, the delay amount in the delay unit can be optimally adjusted according to the distance from the actual speaker installation position to the wall.

  Further, in the present technology, for example, a distance sensor that obtains information on a distance from a speaker setting position to a wall, and a control unit that controls a delay amount in the delay unit based on the distance information obtained by the distance sensor. Further, it may be provided. In this case, since the distance sensor can obtain information on the distance from the actual speaker installation position to the wall, it is possible to optimally adjust the delay amount in the delay unit without requiring the user. .

  Further, in the present technology, for example, based on a user interface for the user to input information on the reflection amount as information for adjusting the gain in the gain adjustment unit, and information on the reflection amount input on the user interface, And a controller that controls the gain in the gain adjusting unit. In this case, the user interface includes a display unit that performs GUI display for the user to input information on the amount of reflection, and a user input for the user to input information about the amount of reflection using the GUI display displayed on the display unit. It may be made to have a part. By providing the user interface in this way, it is possible to optimally adjust the gain in the gain adjustment unit.

  According to the present technology, it is possible to satisfactorily listen to the reflected sound that arrives after the sound output from the speaker is reflected on a surrounding object.

It is a figure which shows the example by which the speaker SP is arrange | positioned at the front side of the housing | casing of a television receiver (TV). It is a figure which shows the example by which the speaker SP is arrange | positioned downward on the back side of the housing | casing of a television receiver (TV). It is a figure for demonstrating the frequency characteristic of the reflected sound which reaches the listening point. It is a block diagram which shows an example of a sound reproducing device. It is a figure which shows an example in which speaker SP is arrange | positioned at the front side of the housing | casing of a television receiver (TV), and also considered reflection of sound. An example is shown in which the speaker SP is disposed downward on the rear side of the housing of the television receiver (TV), taking into account reflections relating to the main sound for listening, and also taking into account reflections other than the main sound. FIG. It is a figure which shows an example of the graph of the simulation result of a frequency characteristic when the main sound and the sub sound are mixed. It is a figure which shows the other example of the graph of the simulation result of a frequency characteristic when the main sound and the sub sound are mixed. It is a figure which shows an example of the problem-solving method by this technique. It is a block diagram which shows the structural example of the sound reproduction apparatus as 1st Embodiment. It is a figure for demonstrating the effect of the sound reproduction apparatus in this technique. It is a figure which shows the case where a television receiver (TV) is installed from the listening point P of a stand. It is a figure which shows the case where a television receiver (TV) is installed from the wall of a stand. It is a block diagram which shows the structural example of the sound reproduction apparatus as 2nd Embodiment. It is a figure which shows that the GUI display for a user to input distance information is performed on a display part. It is a figure for demonstrating installing a distance sensor in the back side of the housing | casing of a television receiver (TV). It is a block diagram which shows the structural example of the sound reproduction apparatus as 3rd Embodiment. It is a figure which shows that the GUI display for a user to input reflection amount information is performed on a display part.

Hereinafter, modes for carrying out the invention (hereinafter referred to as “embodiments”) will be described. The description will be given in the following order.
1. 1. First embodiment 2. Second embodiment 3. Third embodiment Modified example

<1. First Embodiment>
First, the principle of the present technology will be described.
[When the speaker is on the front side]
FIG. 1 shows an example in which the speaker SP is arranged on the front side of the casing of a television receiver (TV). In this case, the user mainly views and watches on the front side of the television receiver. Therefore, the sound Do output from the speaker SP reaches the listening point P directly as the sound Dp.

  This sound Dp may be subject to some attenuation or change in frequency characteristics during propagation through space. However, here, for the sake of simplification of explanation, it is assumed that attenuation and frequency characteristic change during spatial propagation are very small and can be ignored. However, it is assumed that a time delay occurs when the propagation path lengths are different.

[When the speaker is on the back side]
FIG. 2 shows an example in which the speaker SP is disposed downward on the back side of the housing of the television receiver (TV). In this case, the user mainly views and watches on the front side of the television receiver. In this case, the sound Do output from the speaker SP is first reflected at the point S on the stage on which the television receiver is mounted, and then reaches the listening point P as the sound Dp ′. The present technology focuses on this reflection, and it is assumed that there is attenuation due to the reflection.

  In general, the reflectance of sound by an object is not uniform for all frequency components. For this reason, the reflected sound reaches the point P as the sound Dp ′ after undergoing a change in frequency characteristics at the point S. At this time, for example, it may be a “muffled sound” that reaches the point P due to a large attenuation in the high frequency range. As a result, there is a problem in that the music is not gorgeous and boring, or the voice becomes less clear and difficult to hear.

[Difference between ideal frequency characteristic D and actual frequency characteristic D ']
FIG. 3A shows, as an example, a comparison between the frequency characteristic f (Dp) of the sound Dp reaching the point P in FIG. 1 and the frequency characteristic f (Dp ′) of the sound Dp ′ reaching the point P in FIG. ing. Here, for simplification of description, it is assumed that the evaluation is performed by flattening this with reference to f (Dp). f (Dp ′) is subjected to a change in frequency characteristics due to reflection at the point S, and for example, as shown in the figure, the size G is reduced as a whole, or a valley is formed.

  Although it can be intuitively understood that the size is reduced as a whole, this is a phenomenon that occurs because the reflectance of an object is not 100%. Further, the attenuation rate varies depending on the frequency, which is a phenomenon that occurs because the material of the object surface that causes reflection and the internal loss rate are not uniform depending on the frequency. As a result, f (Dp ′) becomes a so-called “muffled sound” as the high frequency is attenuated compared to f (Dp). Such a phenomenon is well known among the traders. Therefore, further explanation is omitted.

[Signal processing for bringing f (Dp ′) closer to the ideal characteristic f (Dp) and adjusting it to f (Dp ″)]
Thus, even if the same sound is output from the same speaker SP, it has been well known that a large difference in sound appears at the point P depending on the arrangement of the speakers SP. Therefore, in order to restore the glitteringness and clarity of the sound of f (Dp ′), as shown in FIG. 3B, the frequency characteristic f (Dp ′) of Dp ′ is expressed as f (Dp ″). A technique for adjusting is generally used, specifically, filtering is performed by analog signal processing or digital signal processing to adjust f (Dp ′) as f (Dp ″).

  Regardless of analog or digital, the signal processing used is to adjust the size for each frequency region. At present, digital signal processing is widely used. Such adjustment can be easily realized by using, for example, a FIR (Finite Impulse Response) filter or an IIR (Infinite Impulse Response) filter.

  However, it is difficult to completely return f (Dp ′) to the frequency characteristic of f (Dp). The reason is that even in FIR / IIR, there are limitations on the resources (computation amount, ROM / RAM capacity) in the product. For example, no matter how much the adjustment is made, a slight deviation such as f (Dp ″) will occur. The more you try to get closer to f (Dp), the larger the resources become, causing cost and power consumption to increase, causing the price of the product to rise. There is always a need to properly balance the advantages and disadvantages of the designer with the advantages and disadvantages of the designer.

  FIG. 4A shows an example of a sound reproducing device that realizes f (Dp) in FIG. The input audio signal Di is amplified by an amplifier and then sent to the speaker SP. The input signal is reproduced as a sound Do from the speaker SP and reaches the point P as a sound Dp. Here, for convenience of explanation, it is defined that the frequency characteristic of the reproduced sound Dp at the point P at this time is equal to the frequency characteristic of the input audio signal Di.

  FIG. 4B shows an example of a sound reproducing device that realizes f (Dp ′) in FIG. The characteristics of the input audio signal Di are the same as those in FIG. 4A. After the power of the input audio signal Di is amplified by the amplifier, it is sent to the speaker SP, and is reproduced from the speaker SP as sound Do. However, as described with reference to FIGS. 2 and 3A, the sound Do changes to the frequency characteristic due to reflection at the point S and reaches Dp ′ instead of Dp before reaching the point P.

  4 (c) shows an example of a sound reproducing device that realizes f (Dp ″) in FIG. 3 (b). The characteristics of the input audio signal Di are shown in FIGS. 4 (a) and 4 (b). However, before the power is amplified by the amplifier, the frequency is adjusted by the filter, changed to Di ′, and transmitted to the amplifier. Is amplified by an amplifier, reproduced from the speaker as sound Do ′, and reaches the point P as sound Dp ″.

  The details of the frequency adjustment have been described in detail with reference to FIG. As a result, it is possible to adjust the frequency characteristic f (Dp ″) of Dp ″ at the point P to a form close to the frequency characteristic f (Dp) of Dp, and to achieve a sound quality close to the frequency characteristic that should be reproduced originally. It can be realized. In order to carry out this adjustment method, the cause of the change in the frequency characteristics is not required, and the frequency adjustment of the filter is set only by looking at the result so that f (Dp) and f (Dp ″) match as much as possible. That's fine.

[When the speaker is on the front side + Reflection is considered]
FIG. 1 described above shows an example in which the speaker SP is disposed on the front side of the casing of the television receiver (TV), but the phenomenon of sound reflection is not taken into consideration. FIG. 5 shows an example in which the speaker SP is disposed on the front side of the casing of the television receiver (TV) and further reflection of sound is taken into consideration.

  In FIG. 5, as in FIG. 1, the main sound for listening to the point P is the direct sound Dp that reaches from the speaker SP. In addition to this, for example, the reflected sound from the point S1 on the ceiling It is assumed that there is Rp1 and reflected sound Rp2 from point S2 on the floor. In reality, reflections are everywhere, but for the sake of simplicity of explanation, the explanation will be continued by focusing on the above-mentioned two points that are likely to have a high reflectance. Here, Dp is referred to as “main sound”, whereas Rp1 and Rp2 are originally unnecessary, and thus referred to as “sub-sound”. In this case, since Rp1 and Rp2 other than Dp reach the point P, no matter how ideal the frequency characteristic is emitted from the speaker SP, when the sound reaches the point P, some change in the frequency characteristic occurs. Will receive.

[When the speaker is on the back side + Reflection is considered]
FIG. 2 described above shows an example in which the speaker SP is disposed downward on the back side of the casing of the television receiver (TV). In this example, the reflection related to the main sound for listening is taken into consideration, but the reflected sound other than the main sound is not taken into account as in FIG. In FIG. 6, the speaker SP is disposed on the back side of the television receiver (TV) housing in a downward direction, taking into account reflections relating to the main sound for listening, and also taking into account reflections other than the main sound. An example is shown.

  In FIG. 6, as in FIG. 2, the main sound for listening is Dp that the sound Do emitted from the speaker SP reflects at the point S <b> 1 and reaches the point P. In addition, for example, the sound Do emitted from the same speaker SP is reflected at the point S2 and further reflected at the point S3 on the wall, and then reaches the point P as Rp. Here, in practice, the reflection is everywhere, but for the sake of simplicity, the description will be continued assuming that there is only reflection from the point S2 that is symmetrical to the point S1 with respect to the speaker SP. Similarly to FIG. 5 described above, Dp is referred to as “main sound”, while Rp is referred to as “sub sound”. In this case, since Rp reaches other than Dp at point P, no matter how ideal the frequency characteristic is emitted from the speaker SP, when the point P is reached, some frequency characteristic change is received. It will be.

[Frequency characteristics when the speaker is on the front side and reflection is considered]
FIG. 7 shows a graph of the simulation result of the frequency characteristics when the main sound and the sub sound are mixed in FIG. In this simulation, white noise whose characteristics are easily evaluated is reproduced as a sound source for 10 seconds, and the overall frequency characteristics for 10 seconds are examined. As shown in FIG. 4A, the frequency characteristic of the playback device is also flat here.

First, let tf0 be a direct transfer characteristic from the speaker SP to the point P. For simplicity of explanation, it is assumed that this transfer characteristic does not cause a change in magnitude and frequency characteristic. If the time delay amount x possessed by tf0 is reflected in parentheses, Dp can be expressed by the following equation (1). The operator “*” is used below to reflect the transfer characteristics.
Dp = Do (0) * tf0 = Do (x) (1)

  Do (0) is a sound just emitted from the speaker SP. When “0” in the parentheses is a reference time, although the frequency characteristic does not change due to the transfer characteristic tf0, a time delay amount x corresponding to the distance from the speaker SP to the point P is generated, so Dp is Do ( x). The numbers in parentheses indicate the amount of time delay based on “0” of Do (0), and are not indexes for specifying samples such as digital data. FIG. 7A shows the frequency characteristic of Do (x), and a flat characteristic of white noise can be seen.

Next, the transfer characteristic from the speaker SP to the point P via the point S1 is set to tf1. For simplicity of explanation, it is assumed that this transfer characteristic does not cause a change in frequency characteristic, but the size becomes 1/10 due to the influence of reflection at the point S1. When the time delay amount of tf1 is x + y, Rp1 reaching the point P can be expressed by the following equation (2). The time delay amount of tf1 is x + y because the path of tf1 is longer than tf0 and increases by y from tf0.
Rp1 = Do (0) * tf1
= Do (x + y) * 1/10 (2)

Similarly, assuming that the transfer characteristic from the speaker SP to the point P via the point S2 is tf2, the magnitude due to reflection is 1/10, and the amount of time delay is x + z, Rp1 reaching the point P is Can be expressed by the following equation (3).
Rp2 = Do (0) * tf2
= Do (x + z) * 1/10 (3)

For simplicity of explanation, assuming that z = y, all sounds Pall reaching the point P can be expressed by the following equation (4).
Pall = Dp + Rp1 + Rp2
= Do (x) + Do (x + y) * 1/5 (4)

  That is, in addition to the original Do, unnecessary components are mixed with each other at 1/5 the size of Do. Furthermore, this unnecessary component is delayed by y as compared with Do (x). This amount of time delay varies depending on the length of the route. For example, considering the size of the room, for the sake of simplicity of explanation, if the length L of the extra path is 3.4 m, the amount of time delay is 10 msec if the speed of sound is 340 m / s. FIG. 7B shows frequency characteristics when Dp, Rp1, and Rp2 are mixed when Rp1 and Rp2 are defined in this way.

  In the frequency characteristic shown in FIG. 7B, it can be seen that peaks and valleys of about ± 2 dB are generated at a constant fine interval as compared with the frequency characteristic shown in FIG. This phenomenon is frequently caused by interference between Dp and Rp1 and Rp2. However, in reality, since the sizes of Rp1 and Rp2 are often smaller, the difference between the frequency characteristics in FIG. 7A and the frequency characteristics in FIG. 7B may not be so large. It is almost. Therefore, it has little influence on sound quality and rarely becomes a problem.

[Frequency characteristics when the speaker is on the back and reflection is considered]
FIG. 8 is a graph showing a simulation result of frequency characteristics when the main sound and the sub sound in FIG. 6 are mixed. In this simulation, as in FIG. 7, it is assumed that white noise whose characteristics are easy to evaluate is reproduced as a sound source for about 10 seconds, and the overall frequency characteristics for 10 seconds are examined.

First, let tf1 be the transfer characteristic from the speaker SP to point P via point S1. For simplification of explanation, it is assumed that this transfer characteristic does not cause a change in frequency characteristic, but its size is halved due to the influence of reflection. If the time delay amount x possessed by tf1 is reflected in parentheses, Dp can be expressed by the following equation (5).
Dp = Do (0) * tf1 = Do (x) * 1/2 (5)

  Do (0) is a sound just emitted from the speaker SP. Assuming that “0” in the parentheses is a reference time, although the frequency characteristic does not change due to the transfer characteristic tf1, a time delay amount x corresponding to the change in size and the distance from the speaker to the point P occurs. Dp is expressed as Do (x) * 1/2.

  FIG. 8A shows the frequency characteristic of Do (x), and the flat characteristic of white noise can be seen. However, in reality, the reflectance may be different for each frequency, and as shown in FIGS. 3A and 3B, the frequency characteristics may not be flat. That is, FIG. 7A and FIG. 8A may not match. However, since the present technology mainly focuses on the improvement of the interference problem at the time of reflection described below, the effects may be compared between those having no frequency characteristic change due to the transfer characteristic tf1 as described above.

Next, let tf2 be the transfer characteristic from the speaker SP to the point S2, and let tf3 be the transfer characteristic from the point S2 to the point P through the point S3. For simplification of explanation, it is assumed that this transfer characteristic does not cause a change in frequency characteristic, but the size is halved at the points S2 and S3 due to the influence of reflection. For simplification of description, if the time delay amounts of tf2 and tf3 are combined and set to x + y, Rp reaching the point P can be expressed by the following equation (6). Note that the amount of time delay is x + y because the path of tf2 and tf3 is longer than tf1 and increases by y from tf1.
Rp = Do (0) * tf2 * tf3
= Do (x + y) * 1/2 * 1/2
= Do (x + y) * 1/4 (6)

Therefore, all the sounds Pall reaching the point P can be expressed by the following equation (7).
Pall = Dp + Rp
= Do (x) * 1/2 + Do (x + y) * 1/4 (7)

  That is, in addition to the original Do, originally unnecessary components are mixed in a size that is ½ of Do. Furthermore, this originally unnecessary component is delayed by y as compared with Do (x). This amount of time delay varies depending on the length of the route. Considering the distance between the TV receiver (TV) and the wall, for simplicity of explanation, if the length L of the extra path is 34 cm, the time delay amount is 1 msec if the speed of sound is 340 m / s. It becomes. FIG. 8B shows the frequency characteristics when Dp and Rp are mixed when Rp is defined in this way.

  In the frequency characteristic shown in FIG. 8B, it can be seen that peaks and valleys of about ± 6 dB are generated at regular intervals as compared with the frequency characteristic shown in FIG. This is because both Dp and Rp are reflection components in the first place, the main sound Dp itself is small, and it is close to the magnitude of the reflection component Rp, and the time delay amount of Rp is relatively short. It is a phenomenon that occurs. Therefore, compared with FIG. 7B, the interference effect that Dp and Rp are strengthened at a specific frequency and Dp and Rp are weakened at another frequency is remarkable, which may cause a sense of discomfort. .

  As the time delay amount y between Dp and Rp is larger, the interval between the peaks and valleys due to this interference action is shorter. In FIG. 8B, the interval between the peaks and valleys is longer than that in FIG. It can be seen that is large. This phenomenon is well understood by simply considering adding two sounds. For example, if the same two signals without time delay are added, the magnitude of the signal is doubled. Next, assuming that this signal is a digital signal, if only one of the signals is delayed and added, the phase difference is small in the low range, and the magnitude is almost doubled, but the magnitude increases as the high range is reached. Is getting smaller.

  This is intuitively understood because it has a low-pass characteristic that the magnitude is 0 because signals with completely different signs are added if the signals of the Nyquist frequency limit are shifted by one sample and added. it can. When the time delay is delayed by 2 samples and 3 samples, the filter has a comb-like filter characteristic rather than a simple low-pass characteristic, and the comb teeth become finer as the time delay amount increases.

  As a cause of the above-mentioned uncomfortable feeling, for example, a mountain portion (formant) that is a feature of voice coincides with a valley generated by interference, and the mountain is crushed. As shown in FIG. 7B, even if the peaks and valleys due to interference are fine and there are multiple peaks and valleys of interference in each of the peaks and valleys of the voice feature, the influence of the interference is different for each peak or valley. Since it is averaged, the characteristics of the voice are not greatly changed. However, as shown in FIG. 8B, when the interval between the peaks and valleys due to the interference is large, a specific peak that is a feature of the voice is largely crushed, and conversely, the valley portion is unnecessarily enhanced, and the feature of the voice is greatly destroyed. End up.

  If the characteristics of the voice are disrupted in this way, sound quality degradation such as “feeling like talking” or “difficult to hear what is being spoken” occurs. This is because the so-called formant structure is destroyed. It is a well-known fact that vowels become unrecognized when formants are removed and played back. In music, only a specific musical instrument hitting a mountain can be heard loudly, and another musical instrument hitting a valley can be heard small, resulting in a sound quality deterioration such that the balance of the entire music becomes strange.

  The above-mentioned problem is not seen from a global conventional viewpoint that the frequency characteristic is simply strong in the low range or weak in the high range, but can be said to be a new way of grasping the problem focusing on the fine structure of the frequency characteristic. Even if such a fine structure is left untouched, the sound quality deterioration may not be improved even if the low and high frequencies are adjusted globally.

[Problem solving method using this technology]
FIG. 9 shows an example of a solution according to the present technology for the above problem. First, as shown in FIG. 6, it is assumed that the speaker SP is disposed downward on the back side of the casing of the television receiver (TV). Then, the main sound Dp that the sound emitted from the speaker SP is reflected at the point S1 and the subordinate sound Rp that the sound emitted from the speaker SP is reflected at the points S2 and S3 reach the point P. I will do it. Here, as described with reference to FIG. 8, it is assumed that an uncomfortable feeling in hearing occurs due to the interference between Dp and Rp.

  The main reason for the uncomfortable feeling is that Rp which is originally unnecessary has an adverse effect on Dp. Therefore, in the present technology, as shown in the figure, a cancel signal Ci for generating a sound Cp for canceling Rp is generated in advance and superimposed on the input audio signal Di. And by reproducing | regenerating as synthetic | combination sound Do "(= Do + Co) from the speaker SP, Rp in P point is canceled. When this process is expressed by numerical formula, it will become as follows.

First, the sound Do ″ generated from the speaker SP is defined as the following equation (8). In this equation (8), Co is a canceling sound newly added and generated from the speaker SP. , Y reaches the point P from the speaker SP via the points S2 and S3, and the time delay from the speaker SP to the point P via the point S1, as shown in FIGS. It is the difference of the time delay until.
Do ″ (0) = Do (0) + Co (y) (8)

Next, Dp is expressed as the following equation (9).
Dp = Do ″ (0) * tf1
= Do (x) * 1/2 + Co (x + y) * 1/2 (9)

Further, Rp is expressed as the following equation (10).
Rp = Do ″ (0) * tf2 * tf3
= Do (x + y) * 1/4 + Co (x + 2y) * 1/4 (10)

Therefore, all the sounds Pall reaching the point P can be expressed as the following equation (11).
Pall = Dp + Rp
= Do (x) * 1/2 + Co (x + y) * 1/2
+ Do (x + y) * 1/4 + Co (x + 2y) * 1/4
(11)

Here, if Co (x + y) = − Do (x + y) * 1/2, Co (x + 2y) = − Do (x + 2y) * 1/2, the two intermediate terms are canceled, and the equation (11) is The following equation (12) is obtained.
Pall = Do (x) * 1 / 2-Do (x + y) * 1/4
+ Do (x + y) * 1 / 4-Do (x + 2y) * 1/8
= Do (x) * 1 / 2-Do (x + 2y) * 1/8
(12)

  In other words, if the speaker SP produces a sound with the phase reversed after the Do is made 1/2 and the time delay amount y is given in addition to Do, the two intermediate terms are canceled. As a result, in the past, a sub sound having a magnitude of 1/2 with respect to the main sound was added with a delay of y time. Sub sound having a magnitude of 4 is subtracted with a delay of 2y hours.

In this case, the output sound Do ″ from the speaker SP is expressed by the following equation (13). If Do ″ is set in this way, Rp is reduced to ½ at the point P. Is possible.
Do ″ (0) = Do (0) + Co (y)
= Do (0)-Do (y) * 1/2 (13)

[Expression when the reflectance is generalized to α]
The above is a case where the reflectance (reflection amount) is halved as shown in the equation (5). If the reflectance is generalized to 1 / α, the equation (5) is transformed into the following equation (14).
Dp = Do (0) * tf1 = Do (x) * 1 / α (14)

Furthermore, the equation (6) is transformed into the following equation (15).
Rp = Do (0) * tf2 * tf3
= Do (x + y) * 1 / α * 1 / α
= Do (x + y) * (1 / α) ² (15)

Therefore, all the sounds Pall reaching the point P can be expressed by the following equation (16). That is, in addition to the original Do, originally unnecessary components are mixed in a size 1 / α times Do.
Pall = Dp + Rp
= Do (x) * 1 / α + Do (x + y) * (1 / α) ²
... (16)

Here, as shown in FIG. 9, the sound Do ″ generated from the speaker SP to cancel Rp is defined as the following equation (17). In this equation (17), Co is a new one. 6 is a canceling sound generated from the speaker SP, and y is the time delay from the speaker SP to the point P via the point S1 as shown in FIGS. Is the difference in time delay until the point P is reached via the points S2 and S3.
Do ″ (0) = Do (0) + Co (y) (17)

Next, Dp is expressed as the following equation (18).
Dp = Do ″ (0) * tf1
= Do (x) * 1 / α + Co (x + y) * 1 / α
... (18)

Further, Rp is expressed as the following equation (19).
Rp = Do ″ (0) * tf2 * tf3
= Do (x + y) * (1 / α) ²
+ Co (x + 2y) * (1 / α) ² (19)

Therefore, all the sounds Pall reaching the point P can be expressed as the following equation (20).
Pall = Dp + Rp
= Do (x) * 1 / α + Co (x + y) * 1 / α
+ Do (x + y) * (1 / α) ²
+ Co (x + 2y) * (1 / α) ² (20)

Here, if Co (x + y) = − Do (x + y) * 1 / α and Co (x + 2y) = − Do (x + 2y) * 1 / α, the intermediate two terms are canceled, and the equation (20) is The following equation (21) is obtained.
Pall = Do (x) * 1 / α−Do (x + y) * (1 / α) ²
+ Do (x + y) * (1 / α) ² −Do (x + 2y) * (1 / α) ³
= Do (x) * 1 / α-Do (x + 2y) * (1 / α) ³
... (21)

In other words, if the speaker SP outputs a sound in which the phase is reversed after setting Do to 1 / α times in addition to Do and then giving the time delay amount y, the two intermediate terms are canceled. As a result, in the past, a sub sound having a magnitude of 1 / α times that of the main sound was added with a delay of y hours, but the (1 / α) ² of the main sound is obtained by using this technology. A secondary sound having a double magnitude is subtracted with a delay of 2y hours.

In this case, the output sound Do ″ from the speaker SP is expressed by the following equation (22). If Do ″ is set in this way, Rp is reduced to 1 / α times the magnitude at the point P. It becomes possible. That is, if the condition of α> 1 is satisfied, the influence of reflected sound can be reduced.
Do ″ (0) = Do (0) + Co (y)
= Do (0) −Do (y) * 1 / α (22)

[Configuration example of sound reproduction device]
FIG. 10 shows a configuration example of the sound reproducing device 100 as the first embodiment. The sound reproducing device 100 includes a digital signal processor (DSP) 101, an amplifier 102, and a speaker 103. The digital signal processor 101 constitutes an audio signal processing unit. The amplifier 102 constitutes an audio signal amplification unit.

  The digital signal processor 101 processes the input audio signal Di to obtain an output audio signal Di ″. The digital signal processor 101 includes a delay unit 111, a gain adjustment unit 112, and an addition unit 113. The delay unit. 111 and the gain adjusting unit 112 constitute a canceling audio signal generating unit that generates a canceling audio signal, and an adding unit 113 adds the canceling audio signal to the input audio signal Di and outputs the output audio signal Di. ″ Get.

  The cancellation audio signal is a predetermined second reflected sound (secondary sound) that arrives after the first reflected sound (main sound) reaches among the reflected sounds that are output from the speaker 103 and reach the listening point P. Is a sound signal for canceling. That is, the delay unit 111 delays the input audio signal Di by a time corresponding to the time difference between the first reflected sound and the second reflected sound that reaches the listening point P. In addition, the gain adjustment unit 112 adjusts the gain of the output signal of the delay unit 111 according to the amount of reflection to obtain a cancellation audio signal. Here, the first reflected sound is equivalent to Dp in FIG. 9, and the second reflected sound is equivalent to Rp in FIG.

  The amplifier 102 amplifies the output audio signal Di ″ obtained by the digital signal processor 101 and supplies it to the speaker 103. The speaker 103 faces downward on the back side of the casing of an electronic device, for example, a television receiver (TV). Therefore, the user listens to the reflected sound that is output from the speaker 103 and reaches the listening point P. The amplifier 102 is often analog. When the amplifier 102 is analog, Di ″ needs to be supplied to the amplifier 102 after being converted from a digital signal to an analog signal. However, recently, digital amplifiers are also common, and for simplicity of explanation, it is assumed that digital signals can be transmitted to the amplifier 102 as they are. The speaker 103 outputs sound based on the sound signal supplied from the amplifier 102.

  The operation of the sound reproducing device 100 shown in FIG. 10 will be described. The input audio signal Di is supplied to the digital signal processor 101. That is, the audio signal Di is supplied to the adding unit 113 and also supplied to the delay unit 111. In the delay unit 111, the input audio signal Di is delayed. Here, when the input audio signal is Di (t), the delay unit 111 holds the input audio signal Di for n samples and outputs the delayed audio signal Di (t + n).

  Next, the delayed audio signal Di (t + n) is supplied to the gain adjustment unit 112. In the gain adjusting unit 112, the gain Gi is adjusted, and a gain-adjusted audio signal Di (t + n) * Gi is obtained. The audio signal Di (t + n) * Gi is supplied to the adder 113. Then, the adder 113 obtains Di (t) + Di (t + n) * Gi as the output audio signal Di ″. The output audio signal Di ″ is amplified by the amplifier 102 and then sent to the speaker 103. The speaker 103 outputs a sound Do ″.

Here, the following equation (23) is established.
Di ″ = Di (t) + Di (t + n) * Gi (23)

If t = 0, n = y, and Gi = −1 / 2, the equation (23) becomes the following equation (24). Comparing Di ″ in the equation (24) with Do ″ (see equation (13)) obtained in FIG. 9, it can be seen that this is the input signal Ci for outputting Do ″ exactly.
Di ″ = Di (0) −Di (y) * 1/2 (24)

  As described above, in the sound reproducing device 100 shown in FIG. 10, it is possible to satisfactorily listen to the reflected sound Dp that reaches the listening point P after the sound output from the speaker 103 (speaker SP) is reflected by a surrounding object. Become. In addition, the sound reproducing device 100 shown in FIG. 10 achieves effective sound quality adjustment and is a disadvantageous sound quality speaker installation position, despite being simpler and lower in cost than the conventional signal processing technology. However, you can enjoy “good sound”.

  FIG. 11 shows the effect of the sound reproducing device 100 shown in FIG. FIG. 11A shows the frequency characteristics of Pall before application of the present technology. FIG. 11B shows the frequency characteristics of Pall after application of the present technology. It can be seen that the size of the peaks and valleys caused by the interference decreases from ± 6 dB to ± 3 dB before and after application. As a result, it is possible to effectively reduce the peaks and valleys caused by the interference from destroying the voice characteristics, and the influences are easily averaged by making the intervals between the peaks and valleys close. Therefore, it is possible to effectively suppress sound quality deterioration.

<2. Second Embodiment>
[Adjust delay amount]
The effectiveness of the present technology has been described with reference to FIGS. However, for example, the relationship between the main sound Dp and the subordinate sound Rp changes depending on where the television receiver (TV) is placed on the stand. For example, FIG. 12 shows a case where a television receiver (TV) is installed from the point P of the stand. In this case, Rp arrives relatively later with respect to Dp as compared to Rp in FIG. FIG. 13 shows a case where a television receiver (TV) is installed on the base wall. In this case, Rp reaches faster relative to Rp in FIG. 9 when Dp is used as a reference.

That is, even if the effect is obtained in the state of FIG. 9, the effect may not be obtained if the installation position of the television receiver (TV) is changed. This can also be said from the following. As described above, the sound Do ″ generated from the speaker SP is defined as the following equation (25). The equation (25) is the same as the above equation (8).
Do ″ (0) = Do (0) + Co (y) (25)

In this equation (25), if y, that is, the time delay amount y of Rp with reference to Dp is not appropriate, y in the middle two terms of the following equation (26) will not match, so Dp and Rp This is because the interference cannot be suppressed. The equation (26) is included in the above equation (12).
Pall = Do (x) * 1 / 2-Do (x + y) * 1/4
+ Do (x + y) * 1 / 4-Do (x + 2y) * 1/8
... (26)

[Configuration example of sound reproduction device]
FIG. 14 shows a configuration example of a sound reproducing device 100A as the second embodiment. In FIG. 14, parts corresponding to those in FIG. 10 are denoted by the same reference numerals, and detailed description thereof is omitted. The sound reproducing device 100A includes a digital signal processor 101, an amplifier 102, a speaker 103, a control unit 104, a user input unit 105, and a display unit 106. The user input unit 105 and the display unit 106 constitute a user interface.

  The control unit 104 controls the delay amount in the delay unit 111 based on the distance information from the speaker 103 to the wall input by the user (user) from the user input unit 105. In addition to the user input unit 105, a display unit 106 is connected to the control unit 104. The user input unit 105 is, for example, an operation button, an operation knob, a remote control device, or the like disposed on a housing of a television receiver (TV). The display unit 106 includes a liquid crystal display element or the like, but can also serve as an image display unit of a television receiver (TV).

  When the user (user) inputs the above-described distance information from the user input unit 105, a GUI (Graphical User Interface) for the user to input distance information on the display unit 106 as shown in FIG. Display is performed. The user (user) inputs information on the distance from the speaker 103 to the wall using this GUI display. Thus, by inputting the distance information from the user input unit 105, the control unit 104 adjusts the delay amount in the delay unit 111 to a value corresponding to the actual distance from the speaker 103 to the wall.

  As described above, the reason why the middle two terms in the equation (26) are not canceled is that the respective y's do not match. This is a problem that can be solved by adjusting y in the above equation (25). In the sound reproducing device 100A shown in FIG. 14, the user (user) can input information on the distance from the speaker 103 to the wall by the user input unit 105 to appropriately adjust y, and the above-described problem can be solved. .

  For example, when the television receiver (TV) is relatively close to the point P, the distance to the wall becomes long, so the user (user) sets the back distance long. Conversely, when the television receiver (TV) is relatively far from the point P, the distance to the wall is shortened, so the user (user) sets the back distance short.

  Increasing the back surface distance means that the delay of Rp becomes larger than Dp. Therefore, in this case, the control unit 104 adjusts in a direction to increase y. Conversely, reducing the back distance means that the delay between Dp and Rp is reduced. Therefore, in this case, the control unit 104 performs adjustment in a direction to reduce y. As described above, the user (user) operates the user input unit 105, for example, a remote control device to change the back surface distance, so that appropriate adjustment can be made according to various installation positions of the television receiver (TV). It becomes.

<3. Third Embodiment>
[Distance sensor]
In the sound reproducing device 100 </ b> A shown in FIG. 14, a user (user) can input information on the distance from the speaker 103 to the wall by the user input unit 105. However, it is also conceivable that distance information 107 is obtained from the distance sensor 107 by installing a distance sensor 107 on the back side of the housing of the television receiver (TV) as shown in FIG. By providing the distance sensor 107 in this way, it is possible to avoid the trouble of the user (user) inputting distance information from the user input unit 105.

[Configuration example of sound reproduction device]
FIG. 17 shows a configuration example of a sound reproduction device 100B as the third embodiment. In FIG. 17, portions corresponding to those in FIG. 14 are denoted by the same reference numerals, and detailed description thereof is omitted. The sound reproducing device 100B includes a digital signal processor 101, an amplifier 102, a speaker 103, a control unit 104B, a user input unit 105, a display unit 106, and a distance sensor 107. The distance sensor 107 is, for example, an infrared distance sensor.

  The control unit 104B controls the delay amount in the delay unit 111 based on the distance information from the speaker 103 to the wall obtained by the distance sensor 107. Thereby, the delay amount in the delay unit 111 is set to a value according to the actual distance from the speaker 103 to the wall.

  The control unit 104 </ b> B controls the gain in the gain adjustment unit 112 based on the reflection amount information input by the user (user) from the user input unit 105. In addition to the user input unit 105, a display unit 106 is connected to the control unit 104B. The user input unit 105 is, for example, an operation button, an operation knob, a remote control device, or the like disposed on a housing of a television receiver (TV). The display unit 106 includes a liquid crystal display element or the like, but can also serve as an image display unit of a television receiver (TV).

  When the user (user) inputs the above-described reflection amount information from the user input unit 105, a GUI (Graphical User) for the user to input the reflection amount information to the display unit 106 as shown in FIG. Interface) display is performed. A user (user) inputs information on the amount of reflection using the GUI display. Thus, by inputting the reflection amount information from the user input unit 105, the gain in the gain adjustment unit 112 is adjusted to a value according to the input reflection amount information by the control unit 104B.

<4. Modification>
In the above-described embodiment, the case where the speaker 103 (speaker SP) is installed downward on the back side of the housing of the television receiver (TV) has been described as an example. However, the present technology can be similarly applied to other electronic devices in which the speaker is installed in the same state, such as a photo frame.

  In the above-described embodiment, only one system for generating a canceling audio signal is provided (see FIGS. 10, 14, and 17). In order to cancel a plurality of reflected sounds having different delay amounts and gains, a configuration including a plurality of systems for generating a canceling audio signal is also conceivable.

Moreover, this technique can also take the following structures.
(1) Among the reflected sounds that are output from the speaker and reach the listening point, a cancellation audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound arrives is designated as an input audio signal. An audio signal generation unit for cancellation generated based on
An audio signal processing device comprising: an audio signal adding unit that adds the cancellation audio signal generated by the cancellation audio signal generation unit to the input audio signal to obtain an output audio signal.
(2) The cancellation audio signal generation unit delays the input audio signal by a time corresponding to a time difference between the first reflected sound and the second reflected sound that reaches the listening point; ,
The audio signal processing device according to (1), further including: a gain adjusting unit that adjusts a gain of the output signal of the delay unit according to a reflection amount to obtain the canceling audio signal.
(3) a user interface for the user to input information on the distance from the speaker installation position to the wall as information for adjusting the delay amount in the delay unit;
The audio signal processing device according to (2), further including: a control unit that controls a delay amount in the delay unit based on the distance information input through the user interface.
(4) The user interface is
A display unit for displaying a GUI for the user to input the distance information;
The audio signal processing device according to (3), further including: a user input unit for a user to input the distance information using a GUI display displayed on the display unit.
(5) a distance sensor that obtains information on the distance from the set position of the speaker to the wall;
The audio signal processing device according to (2), further comprising: a control unit that controls a delay amount in the delay unit based on the distance information obtained by the distance sensor.
(6) a user interface for the user to input information on the reflection amount as information for adjusting the gain in the gain adjustment unit;
The audio signal processing device according to any one of (1) to (5), further including: a control unit that controls gain in the gain adjustment unit based on information on the reflection amount input through the user interface.
(7) The user interface is
A display unit that performs GUI display for the user to input information on the reflection amount;
The audio signal processing device according to (6), further including: a user input unit for a user to input the information on the reflection amount using a GUI display displayed on the display unit.
(8) The audio signal processing device according to any one of (1) to (7), wherein the speaker is attached downward on a back side of a housing of the display device.
(9) Among the reflected sounds that are output from the speaker and reach the listening point, a canceling audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound arrives is designated as an input audio signal. Generating based on:
An audio signal processing method comprising: adding the generated cancellation audio signal to the input audio signal to obtain an output audio signal.
(10) a speaker attached downward on the back side of the housing of the display device;
An audio signal processor that processes an input audio signal to obtain an output audio signal;
An audio signal amplifier that amplifies the output audio signal obtained by the audio signal processor and supplies the amplified audio signal to the speaker;
The audio signal processing unit is
Among the reflected sounds that are output from the speaker and reach the listening point, a cancellation audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound arrives is used as the input audio signal. An audio signal generation unit for cancellation generated based on
An audio reproduction device comprising: an audio signal adding unit that adds the cancellation audio signal generated by the cancellation audio signal generation unit to the input audio signal to obtain the output audio signal.
(11) The sound reproduction device according to (10), wherein the display device is a television receiver.

DESCRIPTION OF SYMBOLS 100,100A, 100B ... Sound reproduction apparatus 101 ... Digital signal processor 102 ... Amplifier 103 ... Speaker 104, 104B ... Control part 105 ... User input part 106 ... Display part 107 ... Distance sensor 111 ... Delay unit 112 ... Gain adjustment unit 113 ... Addition unit

Claims (11)

Based on the input audio signal, a canceling audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound among the reflected sounds that are output from the speaker and reaches the listening point An audio signal generation unit for cancellation to be generated;
An audio signal processing device comprising: an audio signal adding unit that adds the cancellation audio signal generated by the cancellation audio signal generation unit to the input audio signal to obtain an output audio signal. The cancellation audio signal generation unit delays the input audio signal by a time corresponding to the time difference between the first reflected sound and the second reflected sound that reaches the listening point;
The audio signal processing device according to claim 1, further comprising: a gain adjusting unit that adjusts a gain of the output signal of the delay unit according to a reflection amount to obtain the canceling audio signal. A user interface for the user to input information about the distance from the speaker installation position to the wall as information for adjusting the amount of delay in the delay unit;
The audio signal processing apparatus according to claim 2, further comprising: a control unit that controls a delay amount in the delay unit based on the distance information input through the user interface. The above user interface is
A display unit for displaying a GUI for the user to input the distance information;
The audio signal processing apparatus according to claim 3, further comprising: a user input unit for a user to input the distance information using a GUI display displayed on the display unit. A distance sensor for obtaining information about the distance from the set position of the speaker to the wall;
The audio signal processing device according to claim 2, further comprising: a control unit that controls a delay amount in the delay unit based on the distance information obtained by the distance sensor. A user interface for the user to input information on the amount of reflection as information for adjusting the gain in the gain adjustment unit;
The audio signal processing apparatus according to claim 2, further comprising: a control unit that controls a gain in the gain adjustment unit based on information on the reflection amount input through the user interface. The above user interface is
A display unit that performs GUI display for the user to input information on the reflection amount;
The audio signal processing device according to claim 6, further comprising: a user input unit for a user to input information on the reflection amount using a GUI display displayed on the display unit. The audio signal processing device according to claim 1, wherein the speaker is attached downward on a back side of a housing of the display device. Based on the input audio signal, a canceling audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound among the reflected sounds that are output from the speaker and reaches the listening point Generating step;
An audio signal processing method comprising: adding the generated cancellation audio signal to the input audio signal to obtain an output audio signal. A speaker mounted downward on the back side of the housing of the display device;
An audio signal processor that processes an input audio signal to obtain an output audio signal;
An audio signal amplifier that amplifies the output audio signal obtained by the audio signal processor and supplies the amplified audio signal to the speaker;
The audio signal processing unit is
Among the reflected sounds that are output from the speaker and reach the listening point, a cancellation audio signal for canceling a predetermined second reflected sound that arrives after the first reflected sound arrives is used as the input audio signal. An audio signal generation unit for cancellation generated based on
An audio reproduction device comprising: an audio signal adding unit that adds the cancellation audio signal generated by the cancellation audio signal generation unit to the input audio signal to obtain the output audio signal. The sound reproduction apparatus according to claim 10, wherein the display device is a television receiver.

Images