Apparatus and method for modifying a loudspeaker signal for preventing diaphragm over-deflection

Abstract

Apparatus configured to predict a diaphragm deflection signal, block-by-block, in overlapping time blocks based on the loudspeaker signal to obtain one diaphragm deflection signal block per time block. The apparatus is configured to determine a temporal position of a maximum deflection of a current diaphragm deflection signal block of a current time block within an overlap area with a subsequent time block and to calculate a level up to which the current diaphragm deflection signal block can be controlled without diaphragm over-deflection for the current time block by considering a comparison of the current diaphragm deflection signal block with a subsequent diaphragm deflection signal block or an estimation of the subsequent diaphragm deflection signal block from the current diaphragm deflection signal block at the temporal position. The apparatus is configured to attenuate the current diaphragm deflection signal block and to synthesize a modified loudspeaker signal.

Claims

1. Apparatus for modifying a loudspeaker signal for preventing diaphragm over-deflection configured to predict a diaphragm deflection signal, block-by-block, in overlapping time blocks based on the loudspeaker signal to acquire one diaphragm deflection signal block per time block; determine a temporal position of a maximum deflection of a current diaphragm deflection signal block of a current time block within an overlap area with a subsequent time block; calculate a level up to which the current diaphragm deflection signal block can be controlled without diaphragm over-deflection for the current time block, by considering a comparison of the current diaphragm deflection signal block with a subsequent diaphragm deflection signal block or an estimation of the subsequent diaphragm deflection signal block from the current diaphragm deflection signal block at the temporal position; attenuate the current diaphragm deflection signal block based on the level to acquire a modified current diaphragm deflection signal block; and synthesize a modified loudspeaker signal based on the modified current diaphragm deflection signal block.

2. Apparatus according to claim 1, wherein, for calculating the level, the apparatus is configured to determine a quotient between the current diaphragm deflection signal block at the temporal position on the one hand and a sum of the subsequent diaphragm deflection signal block or the estimation of the subsequent diaphragm deflection signal block at the temporal position and the current diaphragm deflection signal block at the temporal position on the other hand.

3. Apparatus according to claim 1, wherein the apparatus is configured to scale the quotient with the safety factor.

4. Apparatus according to claim 1, wherein the apparatus is configured to decompose the current diaphragm deflection signal block into at least one frequency band signal time block.

5. Apparatus according to claim 4, wherein the apparatus is configured to determine, for each of the at least one frequency band signal time block, at least one further temporal position where a predetermined signal combination exceeds a maximum deflection limiting value in the first overlap area of the current time block with the preceding time block as well as in the residual current time block.

6. Apparatus according to claim 5, wherein the signal combination comprises: an additive combination of the current diaphragm deflection signal block with a modified preceding diaphragm deflection signal block; and an additive combination of the respective frequency band signal time block with the modified preceding diaphragm deflection signal block; and an additive combination of an amount of the current diaphragm deflection signal block with an amount of the modified preceding diaphragm deflection signal block.

7. Apparatus according to claim 5, wherein the signal combination comprises at least one of the following additive combinations: an additive combination of the current diaphragm deflection signal block with a modified preceding diaphragm deflection signal block; and an additive combination of the respective frequency band signal time block with the modified preceding diaphragm deflection signal block; and an additive combination of an amount of the current diaphragm deflection signal block with an amount of the modified preceding diaphragm deflection signal block.

8. Apparatus according to claim 5, wherein the apparatus is configured to determine an attenuation factor for each of the at least one frequency band signal time block based on the level and the further temporal position to attenuate the current diaphragm deflection signal block.

9. Apparatus according to claim 8, wherein the apparatus is configured to compare the attenuation factor per frequency band signal time block for the current time block with a version of the attenuation factor per frequency band signal time block for the preceding time block, reduced in attenuation strength by a predetermined fading-away function and to use a selected attenuation factor that is associated with a higher attenuation for the current time block of the same.

10. Apparatus according to claim 9, wherein the apparatus is configured to split the current diaphragm deflection signal block into a percussive signal portion and a harmonic signal portion and to determine the predetermined fading-away function in dependence on the percussive signal portion and/or the harmonic signal portion.

11. Apparatus according to claim 5, wherein the apparatus is configured to calculate, for each of the at least one further temporal position, a first maximum deflection portion for the current time block by considering the maximum deflection limiting value and the modified preceding diaphragm deflection signal block.

12. Apparatus according to claim 11, wherein the apparatus is configured to calculate the first maximum deflection portion from a quotient between a difference of the maximum deflection limiting value, wherein a sign of the maximum deflection limiting value depends on the current diaphragm deflection signal block at the further temporal position and the modified preceding diaphragm deflection signal block on the one hand and the maximum deflection limiting value on the other hand.

13. Apparatus according to claim 1, wherein the apparatus is configured to calculate a modified current loudspeaker signal block from the modified current diaphragm deflection signal block and to subject a first part of the modified current loudspeaker signal block to an overlap-add with a modified preceding loudspeaker signal block to synthesize a modified loudspeaker signal.

14. Apparatus according to claim 13, wherein the apparatus is configured to provide a current loudspeaker signal instead of the modified current loudspeaker signal block when a maximum deflection of the current diaphragm deflection signal block of the current time block does not exceed a maximum deflection limiting value.

15. Apparatus according to claim 1, wherein the apparatus is configured to attenuate the current diaphragm deflection signal block to calculate one or several attenuation factors based on the level and to compute the one or several attenuation factors with the current diaphragm deflection signal block.

16. Apparatus according to claim 15, wherein the apparatus is configured to compare the one or several attenuation factors for the current time block with a version of the one or several attenuation factors for the preceding time block, reduced in attenuation strength by a predetermined fading-away function and to use a selected attenuation factor that is associated with a higher attenuation for the current time block of the same.

17. Apparatus according to claim 16, wherein the apparatus is configured to split the current diaphragm deflection signal block into a percussive signal portion and a harmonic signal portion and to determine the predetermined fading-away function in dependence on the percussive signal portion and/or the harmonic signal portion.

18. Apparatus according to claim 17, wherein the predetermined fading-away function comprises shorter time constants for the percussive signal portion than for the harmonic signal portion.

19. Apparatus for modifying a loudspeaker signal for preventing diaphragm over-deflection configured to predict a diaphragm deflection signal, block-by-block, in overlapping time blocks based on the loudspeaker signal to acquire one diaphragm deflection signal block per time block; determine a first maximum deflection portion for a current diaphragm deflection signal block of a current time block in a first overlap area of the current time block with a preceding time block as well as in the residual area of the current diaphragm deflection signal block; determine a second maximum deflection portion for the current diaphragm deflection signal block of the current time block in a second overlap area of the current time block with a subsequent time block; calculate a level based on the first maximum deflection portion when the first maximum deflection portion is smaller than the second maximum deflection portion or calculate the level based on the second maximum deflection portion when the second maximum deflection portion is smaller than the first maximum deflection portion; attenuate the current diaphragm deflection signal block based on the level to acquire a modified current diaphragm deflection signal block; and synthesize a modified loudspeaker signal based on the modified current diaphragm deflection signal block.

20. Apparatus according to claim 19, wherein the apparatus is configured to calculate the level based on the first maximum deflection portion by means of a product of a maximum deflection limiting value and the first maximum deflection portion or to calculate the level based on the second maximum deflection portion by means of a product of the maximum deflection limiting value and the second maximum deflection portion.

21. Apparatus according to claim 19, wherein the apparatus is configured to determine a temporal position of a maximum deflection of the current diaphragm deflection signal block of the current time block within the second overlap area with the subsequent time block; and determine, for the second maximum deflection portion, a quotient between the current diaphragm deflection signal block at the temporal position on the one hand and a sum of a subsequent diaphragm deflection signal block or an estimation of the subsequent diaphragm deflection signal block at the temporal position, based on the current diaphragm deflection signal block of the current time block, and the current diaphragm deflection signal block at the temporal position on the other hand.

22. Apparatus according to claim 21, wherein the apparatus is configured to scale the quotient with a safety factor.

23. Apparatus according to claim 19, wherein the apparatus is configured to estimate the second maximum deflection portion based on the current diaphragm deflection signal block.

24. Apparatus according to claim 19, wherein the apparatus is configured to estimate the second maximum deflection portion based on the current diaphragm deflection signal block by means of a neural network.

25. Method for modifying a loudspeaker signal for preventing diaphragm over-deflection, comprising: block-by-block prediction of a diaphragm deflection signal in overlapping time blocks based on the loudspeaker signal to acquire one diaphragm deflection signal block per time block; determination of a temporal position of a maximum deflection of a current diaphragm deflection signal block of a current time block within an overlap area with a subsequent time block; calculation of a level up to which the current diaphragm deflection signal block can be controlled without diaphragm over-deflection for the current time block by considering a comparison of the current diaphragm deflection signal block with a subsequent diaphragm deflection signal block or an estimation of the subsequent diaphragm deflection signal block from the current diaphragm deflection signal block at the temporal position; attenuation of the current diaphragm deflection signal block based on the level to acquire a modified current diaphragm deflection signal block; and synthesization of a modified loudspeaker signal based on the modified current diaphragm deflection signal block.

26. Method for modifying a loudspeaker signal for preventing diaphragm over-deflection, comprising: block-by-block prediction of a diaphragm deflection signal in overlapping time blocks based on the loudspeaker signal to acquire one diaphragm deflection signal block per time block; determination of a first maximum deflection portion for a current diaphragm deflection signal block of a current time block in a first overlap area of the current time block with a preceding time block as well as in the residual area of the current diaphragm deflection signal block; determination of a second maximum deflection portion for the current diaphragm deflection signal block of the current time block in a second overlap area of the current time block with a subsequent time block; calculation of a level based on the first maximum deflection portion when the first maximum deflection portion is smaller than the second maximum deflection portion or calculation of the level based on the second maximum deflection portion when the second maximum deflection portion is smaller than the first maximum deflection portion; attenuation of the current diaphragm deflection signal block based on the level to acquire a modified current diaphragm deflection signal block; and synthesization of a modified loudspeaker signal based on the modified current diaphragm deflection signal block.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the present invention will be detailed subsequently referring to the appended drawings, in which:

(2) FIG. 1 is a schematic illustration of an apparatus according to an embodiment of the present invention;

(3) FIG. 2 is a schematic illustration of an apparatus for calculating a first and a second maximum deflection portion according to an embodiment of the present invention;

(4) FIG. 3 is a block diagram of an overall signal flow in the apparatus according to an embodiment of the present invention;

(5) FIG. 4 is a block diagram of a signal analysis of the apparatus according to an embodiment of the present invention;

(6) FIG. 5a is a block diagram of a calculation of a second maximum deflection portion with the apparatus according to an embodiment of the present invention;

(7) FIG. 5b is a block diagram of a calculation of a second maximum deflection portion with an estimation of a subsequent diaphragm deflection signal block by means of the apparatus according to an embodiment of the present invention;

(8) FIG. 5c is a block diagram of an estimation of a second maximum deflection portion with the apparatus according to an embodiment of the present invention;

(9) FIG. 6 is a block diagram of an adaptation of attenuation factors by means of the apparatus according to an embodiment of the present invention;

(10) FIG. 7 is a block diagram of a method for modifying a loudspeaker signal for preventing diaphragm over-deflection according to an embodiment of the present invention; and

(11) FIG. 8 is a block diagram of a method for modifying a loudspeaker signal for preventing diaphragm over-deflection with a determination of a first and a second maximum deflection portion according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

(12) Embodiments according to the present invention will be discussed in more detail below with reference to the accompanying drawings. Regarding the illustrated schematic figures, it should be noted that the illustrated functional blocks can be considered both as elements or features of the inventive apparatus and as respective method steps of the inventive method and respective method steps of the inventive method can also be derived therefrom. Before embodiments of the present invention will be discussed in more detail below with reference to the drawings, it should be noted that identical, functionally equal or equal elements, objects and/or structures in the different figures are provided with the same or similar reference numbers in the different figures such that the description of these elements illustrated in different embodiments is inter-exchangeable or inter-applicable.

(13) FIG. 1 shows a schematic illustration of an apparatus 100 for modifying a loudspeaker signal 110 for preventing diaphragm over-deflection. The apparatus 100 can be configured to predict a diaphragm deflection signal 130 block-by-block in overlapping time blocks 120.sub.1 to 120.sub.3 based on the loudspeaker signal 110 to obtain one diaphragm deflection signal block 130.sub.1 to 130.sub.3 per time block 120.sub.1 to 120.sub.3. According to FIG. 1, for example, the apparatus 100 can obtain three diaphragm deflection signal blocks 130.sub.1 to 130.sub.3 for three time blocks 120.sub.1 to 120.sub.3. This is only an embodiment and it is possible that the apparatus 100 can predict the diaphragm deflection signal 130 with more overlapping time blocks 120.sub.1 to 120.sub.3 from the loudspeaker signal 110. In that way, for example, the apparatus 100 can predetermine that the time blocks 120.sub.1 to 120.sub.3 have a block size of 10 ms (block sizes of ≥5 ms, ≥8 ms order ≥10 ms are advantageous), whereby the apparatus 100 can obtain so many diaphragm deflection signal blocks 130.sub.1 to 130.sub.3 that the complete loudspeaker signal 110 (for example the complete time period of the loudspeaker signal 110) can be completely reflected by the time blocks 120.sub.1 to 120.sub.3 with a block length predetermined by the apparatus 100. In that way, for example, the diaphragm deflection signal 130 can define how much the loudspeaker signal 110 would deflect a diaphragm at a loudspeaker at any time.

(14) Further, the apparatus 100 can be configured to determine a temporal position 140 of a maximum deflection of a current diaphragm deflection signal block 130.sub.2 of a current time block 120.sub.2 within an overlap area 124 with a subsequent time block 120.sub.3. In that way, the apparatus 100 can, for example, sample the current diaphragm deflection signal block 130.sub.2 for a maximum deflection within the overlap area 124 of the current time block 120.sub.2. This maximum deflection can result, for example, in diaphragm over-deflection when the apparatus 100 does not modify the diaphragm deflection signal 130.

(15) The apparatus 100 can be configured to calculate (for example with the processing means 150), for the current time block 120.sub.2 by considering a comparison of the current diaphragm deflection signal block 130.sub.2 with a subsequent diaphragm deflection signal block 130.sub.3 or an estimation of the subsequent diaphragm deflection signal block 130.sub.3 from the current diaphragm deflection signal block 132.sub.2 at the temporal position, a level 150a up to which the current diaphragm deflection signal block 130.sub.2 can be controlled without diaphragm over-deflection. By the level 150a, the apparatus 100 can determine how much the current diaphragm deflection signal block 130.sub.2 should, for example, be attenuated to attenuate the current diaphragm deflection signal block 130.sub.2 in a ratio matching the subsequent diaphragm deflection signal block 130.sub.3, such that high sound quality can be maintained by the apparatus 100.

(16) Further, the apparatus 100 can be configured to attenuate the current diaphragm deflection signal block 130.sub.2 based on the level 150a to obtain a modified current diaphragm deflection signal block 160.sub.2. The apparatus can perform this, for example, for each diaphragm deflection signal block 130.sub.1 to 130.sub.3 to obtain, for example, three modified diaphragm deflection signal blocks 160.sub.1 to 160.sub.3 according to FIG. 1. The modified diaphragm deflection signal blocks 160.sub.1 to 160.sub.3 can define a modified diaphragm deflection signal 160.

(17) Further, the apparatus 100 can be configured to synthesize a modified loudspeaker signal.

(18) Further, the apparatus 100 can be configured to synthesize a modified loudspeaker signal 170 based on the modified current diaphragm deflection signal block 160.sub.2. For this, the apparatus 100 can, for example, synthesize a modified loudspeaker signal block 170.sub.1, 170.sub.2 from each modified diaphragm deflection signal block 160.sub.1, 160.sub.2 and join them to the modified loudspeaker signal 170 by an overlap-add method. Alternatively, the apparatus 100 can be configured to first join the modified diaphragm deflection signal blocks 160.sub.1, 160.sub.2 to the modified diaphragm deflection signal 160 by the overlap-add method and to synthesize the complete modified loudspeaker signal 170 in one step from the modified diaphragm deflection signal 160. Optionally, the apparatus 100 can provide the loudspeaker signal 170 modified in that manner and transmit the same, for example, to a loudspeaker.

(19) According to an embodiment, for calculating the level 150a, the apparatus 100 can be configured to determine a quotient between the current diaphragm deflection signal block 130.sub.2 at the temporal position 140 on the one hand and a sum of the subsequent diaphragm deflection signal block 130.sub.3 or the estimation of the subsequent diaphragm deflection signal block 130.sub.3 at the temporal position 140 and the current diaphragm deflection signal block 130.sub.2 at the temporal position 140 on the other hand. This step can be performed, for example, with the processing means 150. The quotient can define, for example, the relationship of the current diaphragm deflection signal block 130.sub.2 to the subsequent diaphragm deflection signal block 130.sub.3 at the temporal position 140. Accordingly, the level 150a calculated in that manner can define how much the current diaphragm deflection signal block 103.sub.2 is to be attenuated at least so that a high sound quality can be maintained while preventing diaphragm over-deflection.

(20) According to an embodiment, the apparatus 100 can be configured to scale the quotient with a safety factor. This can, for example, cause strong attenuation of the diaphragm deflection signal 130, whereby mechanical protection of a diaphragm of a loudspeaker from diaphragm over-deflection can be increased. This step can be performed, for example, with the processing means 150.

(21) According to an embodiment, the apparatus 100 can be configured to decompose the current diaphragm deflection signal block 130.sub.2 in at least one frequency band signal time block. This can take place, for example, with the processing means 150. In that way, the apparatus 100 can decompose the current diaphragm deflection signal block 130.sub.2 into at least one frequency band, wherein the at least one frequency band signal time block can represent a frequency band of the current diaphragm deflection signal block 130.sub.2 for the complete current time block 120.sub.2. Thus, the apparatus 100 can decompose the current diaphragm deflection signal block 130.sub.2 into several frequency band signal time blocks, which can each represent the complete current time block 120.sub.2 and a differing frequency band of the current diaphragm deflection signal block 130.sub.2. This feature enables the apparatus 100 to attenuate or modify the current diaphragm deflection signal block 130.sub.2 such that no or only few nonlinearities result in the modified current diaphragm deflection signal block 160.sub.2. Thereby, the apparatus 100 can ensure high sound quality. The feature described herein can represent, for example, a frequency band decomposition 151 by the processing means 150.

(22) According to an embodiment, the apparatus 100 can be configured to determine at least one further temporal position 142 for each of the at least one frequency band signal time block, where a predetermined signal combination (e.g., of the current diaphragm deflection signal block 130.sub.2 with the preceding modified diaphragm deflection signal block 160.sub.1) exceeds a maximum deflection limiting value in the first overlap area 122 of the current time block 120.sub.2 with the preceding time block 120.sub.1 as well as in the residual time block 120.sub.2 (e.g., also in the overlap area 124). The predetermined signal combination can represent, for example, an overlap-add of the current diaphragm deflection signal block 130.sub.2 with the modified preceding diaphragm deflection signal block 160.sub.1 in the first overlap area 122. The maximum deflection limiting value can define, for example, from when on, e.g., an amplitude of the current diaphragm deflection signal block 130.sub.2 would result in diaphragm over-deflection when the apparatus 100 would not attenuate or modify the current diaphragm deflection signal block 130.sub.2. Thus, at least one further temporal position 142, which can also be referred to as critical temporal position, can be determined per frequency band signal time block, whereby the apparatus 100 can determine how the at least one frequency band signal time block should be modified or attenuated so that diaphragm over-deflection can be prevented, and hence good mechanical protection for loudspeakers can be ensured by the apparatus 100. This can take place, for example, by means of position determination 152 of the processing means 150.

(23) According to an embodiment, the signal combination can include an additive combination of the current diaphragm deflection signal block 130.sub.2 with a modified preceding diaphragm deflection signal block 160.sub.1 and an additive combination of the respective frequency band signal time block with the modified preceding diaphragm deflection signal block 160.sub.1, and an additive combination of an amount of the current diaphragm deflection signal block 130.sub.2 with an amount of the modified preceding diaphragm deflection signal block 160.sub.1. Thus, e.g., at least one of the three additive combinations should exceed the maximum deflection limiting value at the further temporal position 142. The three additive combinations are based on the finding that deflection peaks (for example, maximum deflections) of a frequency band signal block cancel each other out in a superposition due to a phase position and can be invisible in the current diaphragm deflection signal block 130.sub.2. Further, with the three additive combinations, it can be considered that deflection peaks in high-frequency frequency band signal time blocks should be attenuated to a sufficient extent and deflection peaks in the current diaphragm deflection signal block 130.sub.2 may be smeared by close extremes in the individual frequency band signal time blocks. Thus, at least one further temporal position 142 where diaphragm over-deflection can occur and should hence be modified by the apparatus 100 can be determined with the apparatus 100 by the three additive combinations of the signal combination.

(24) According to an embodiment, the apparatus 100 can be configured to determine an attenuation factor for each of the at least one frequency band signal time block based on the level 150a and the further temporal position 142 to attenuate the current diaphragm deflection signal block 130.sub.2. This can take place by attenuation factor determination 153. Thus, for example, the attenuation factor determination 153 can obtain the at least one further temporal position 142 per frequency band signal time block from the position determination 152 and can obtain the level 150a from the processing means 150. This can enable that, when optimizing the attenuation factors, usage of arbitrary start values can be prevented and instead deflection values can be used at the further temporal position 142, which can accelerate optimization. Thus, the apparatus 100 is configured to modify or attenuate the diaphragm deflection signal 130 very efficiently, e.g., while ensuring high sound quality, such that diaphragm over-deflection caused by the loudspeaker signal 110 can be prevented.

(25) According to the embodiment, the apparatus 100 can be configured to compare the attenuation factor per frequency band signal time block for the current time block 120.sub.2 with a version of the attenuation factor per frequency band signal time block for the preceding time block 120.sub.1, reduced in attenuation strength by a predetermined fading-away function, and use a selected attenuation factor that can be associated with a higher attenuation for the current time block 120.sub.2 of the same. Thus, it can, for example, be considered that an attenuation of the preceding diaphragm deflection signal block 130.sub.1 can comprise a fading-away function that can also influence an attenuation of the current diaphragm deflection signal block 130.sub.2 by the apparatus 100, whereby, for example, the current diaphragm deflection signal block 130.sub.2, for example, reduced in attenuation strength with the attenuation factors by the predetermined fading-away function, can already be sufficiently attenuated to prevent diaphragm over-deflection. Thus, with this feature, the attenuation factors can be selected such that the current diaphragm deflection signal block 130.sub.2 is sufficiently attenuated by the apparatus 100 to prevent diaphragm over-deflection and hence to provide good mechanical protection for a loudspeaker by the apparatus 100. This can take place with the attenuation factor determination 153.

(26) The apparatus 100 can, for example, attenuate the diaphragm deflection signal 130 with attack, hold and release functions. By short attack times, the current diaphragm deflection signal block 130.sub.2 can be attenuated very quickly, the attenuation can be held for a short time by means of the hold function and can be released again with the release function. In that way, for example, an attack time of 0 seconds can be used so that the modified loudspeaker signal 107 that can be provided by the apparatus 100 does not comprise any non-attenuated diaphragm deflection amplitudes that could result in diaphragm over-deflection. The release time can depend on the determined attenuation factors. In that way, for example, long sounds may need long time constants to cause no modulation and percussion-like sounds may need rather short time constants. The release time can monitor the attenuation factors of the individual frequency band signal time blocks by comparing the attenuation factors of the current time block 120.sub.2 with the version of the attenuation factors of the preceding time block 120.sub.1, reduced in attenuation strength by a predetermined fading-away function. If, for example, the diaphragm deflection of the current diaphragm deflection signal block 130.sub.2 increases, the attenuation factors can be adapted such that diaphragm over-deflection can be prevented. This can take place with the attenuation factor determination 153.

(27) According to an embodiment, the apparatus 100 can be configured to split the current diaphragm deflection signal block 130.sub.2 into a percussive signal portion and a harmonic signal portion, and to determine the predetermined fading-away function in dependence on the percussive signal portion and/or the harmonic signal portion. This can take place, for example, with the diaphragm deflection signal splitting 154 prior to the frequency band decomposition 151. This feature can improve the sound quality, since, for example, shorter time constants of the predetermined fading-away function are better suited for percussive signals and longer time constants, e.g., for harmonic signals. The split into the percussive signal portion and the harmonic signal portion can take place, for example, for the complete current diaphragm deflection signal block 130.sub.2 or for each frequency band signal time block of the current diaphragm deflection signal block 130.sub.2.

(28) According to an embodiment, the apparatus 100 can be configured to calculate, for each of the at least one further temporal position 142, a first maximum deflection portion for the current time block 120.sub.2 by considering the maximum deflection limiting value and the modified preceding diaphragm deflection signal block 160.sub.1. This maximum deflection portion can determine for the current diaphragm deflection signal block 130.sub.2, how much the current diaphragm deflection signal block 130.sub.2 should still be attenuated by the apparatus 100, so that no diaphragm over-deflection is caused in an overlap-add of the overlap area 122, and hence the modified preceding diaphragm deflection signal block 130.sub.1 can also be considered. This can take place with the processing means 150.

(29) According to an embodiment, the apparatus 100 can be configured to calculate the first maximum deflection portion from a quotient between a difference of the maximum deflection limiting value, wherein a sign of the maximum deflection limiting value can depend on the current diaphragm deflection signal block 130.sub.2 at the further temporal position 142 and the modified preceding diaphragm deflection signal block 160.sub.1 on the one hand, and the maximum deflection limiting value on the other hand. This can take place with the processing means 150.

(30) According to an embodiment, the apparatus 100 can be configured to calculate a modified current loudspeaker signal block 170.sub.2 from the modified current diaphragm deflection signal block 160.sub.2 and to subject the modified current loudspeaker signal block 170.sub.2 to an overlap-add with the modified preceding loudspeaker signal block 170.sub.1 to synthesize a modified loudspeaker signal 170. When calculating the current loudspeaker signal block 170.sub.2, for example, no subsequent modified loudspeaker signal block 170.sub.3 exists. Thus, an output to a loudspeaker can be, for example, the overlap of the preceding modified loudspeaker signal block 170.sub.1 with a first part of the current modified loudspeaker signal block 170.sub.2, wherein the first part can be, for example, an overlap area 122 of the current time block 120.sub.2 with the preceding time block 120.sub.1.

(31) According to an embodiment, the apparatus 100 can be configured to provide a current loudspeaker signal 110 instead of the modified current loudspeaker signal block 170.sub.2 when a maximum deflection of the current diaphragm deflection signal block 130.sub.2 of the current time block 120.sub.2 does not exceed a maximum deflection limiting value. Thus, the apparatus 100 can very efficiently protect a loudspeaker from diaphragm over-deflection by attenuating or modifying the current loudspeaker signal 110 only when the apparatus 100 detects that diaphragm over-deflection can be caused at the loudspeaker by the current loudspeaker signal 110.

(32) According to an embodiment, for attenuating the current diaphragm deflection signal block 130.sub.2, the apparatus 100 can be configured to calculate one or several attenuation factors based on the level 150a (e.g., with the attenuation factor determination 153) and to compute the one or several attenuation factors with the current diaphragm deflection signal block 130.sub.2. With a computing means 155, the one or several attenuation factors can be computed with the current diaphragm deflection signal block 130.sub.2 such that a modified current diaphragm deflection signal block 160.sub.2 can be determined, which should not result in diaphragm over-deflection.

(33) FIG. 2 shows a schematic illustration of an apparatus 100 for modifying a loudspeaker signal 110 for preventing diaphragm over-deflection, which can comprise features and functionalities according to the apparatus 100 of FIG. 1. Further, the apparatus 100 can be configured to determine a first maximum deflection portion 180a for a current diaphragm deflection signal block 130.sub.2 of a current time block 120.sub.2 in a first overlap area 142 of a current time block 120.sub.2 (e.g., the current diaphragm deflection signal block 130.sub.2) with a preceding time block 120.sub.1 as well as in the residual area of the current diaphragm deflection signal block 130.sub.2. Thus, it can be determined for the entire current time block 120.sub.2 how much the apparatus 100 should at least attenuate the current diaphragm deflection signal block 130.sub.2, so that, for example, for an overlap-add of the current modified diaphragm deflection signal block 160.sub.2 with the preceding modified diaphragm deflection signal block 160.sub.1, no diaphragm over-deflection is caused at a loudspeaker. Thus, the maximum deflection portion can indicate how large a portion of the current diaphragm deflection signal block 130.sub.2 may be of a maximum deflection limiting value so that the maximum deflection limiting value is not exceeded and hence, diaphragm over-deflection may be caused. Here, the maximum deflection limiting value can define how large a deflection after an overlap-add of the current modified diaphragm deflection signal block 160.sub.2 and the preceding modified diaphragm deflection signal block 160.sub.1 may be at most so that no diaphragm over-deflection results. A ratio of the first maximum deflection portion 180a to a residual deflection portion (e.g., a deflection portion of the preceding modified diaphragm deflection signal block 160.sub.1 of the maximum deflection limiting value) can, for example, be selected such that high sound quality can be obtained.

(34) Additionally, the apparatus 100 can be configured to determine a second maximum deflection portion 180b for the current diaphragm deflection signal block 130.sub.2 of the current time block 120.sub.2 in a second overlap area 124 of the current time block 120.sub.2 with a subsequent time block 120.sub.3. With this feature, the apparatus 100 is able to attenuate or modify the current diaphragm deflection signal block 130.sub.2 already in a look-ahead manner. Thus, the apparatus 100 can determine, for example, the relationship of the current diaphragm deflection signal block 130.sub.2 to the subsequent diaphragm deflection signal block 130.sub.3 and, accordingly, attenuate the current diaphragm deflection signal block 130.sub.2, whereby high sound quality can be obtained. The second maximum deflection portion 180b can have this relationship.

(35) Further, the apparatus 100 can be configured to calculate a level 150a based on the first maximum deflection portion 180a when the first maximum deflection portion 180a is smaller than the second maximum deflection portion 180b, or to calculate the level 150a based on the second maximum deflection portion 180b when the second maximum deflection portion 180b is smaller than the first maximum deflection portion 180a. Thus, the apparatus 100 is configured, for example, to determine a greater attenuation based on the first maximum deflection portion 180a and the second maximum deflection portion 180b and to select the same to attenuate the diaphragm deflection signal 130 to ensure that a modified diaphragm deflection signal 160 cannot result in diaphragm over-deflection. This takes place, e.g., by a processing means 150.

(36) According to an embodiment, the apparatus 100 can be configured to calculate the level 150a based on the first maximum deflection portion 180a by means of a product of the maximum deflection limiting value and the first maximum deflection portion 180a, or to calculate the level 150a based on the second maximum deflection portion 180b by means of a product of the maximum deflection limiting value and the second maximum deflection portion 180b. Thus, the level 150a can define a maximum amplitude, which the current diaphragm deflection signal block 130.sub.2 may have at maximum. Thus, the current diaphragm deflection signal block 130.sub.2 may not exceed the level 150a. Level calculation takes place, for example, via the processing means 150. With the level 150a, the apparatus 100 can check whether the current diaphragm deflection signal block 130.sub.2 should be modified or attenuated or can be provided without any further processing.

(37) According to an embodiment, the apparatus 100 can be configured to determine a temporal position 140 of a maximum deflection of the current diaphragm deflection signal block 130.sub.2 of the current time block 120.sub.2 within the second overlap area 124 with the subsequent time block 120.sub.3; and to determine, for the second maximum deflection portion 180b, a quotient between the current diaphragm deflection signal block 130.sub.2 at the temporal position 140 on the one hand and a sum of a subsequent diaphragm deflection signal block 130.sub.3 or an estimation of the subsequent diaphragm deflection signal block 130.sub.3 at the temporal position 140, based on the current diaphragm deflection signal block 130.sub.2 of the current time block 120.sub.2, and the current diaphragm deflection signal block 130.sub.2 at the temporal position 140 on the other hand. According to an embodiment, the apparatus 100 can be configured to scale the quotient with a safety factor. The temporal position can be determined, for example, via the processing means 150.

(38) According to an embodiment, the apparatus 100 can be configured to estimate the second maximum deflection portion 180b based on the current diaphragm deflection signal block 130.sub.2. Thus, for example, to determine the second maximum deflection portion 180b, not the subsequent diaphragm deflection signal block 130.sub.3 but only the current diaphragm deflection signal block 130.sub.2 is needed. This improves the efficiency of the apparatus 100 for modifying the diaphragm deflection signal 130 to prevent diaphragm over-deflection.

(39) According to an embodiment, the apparatus 100 can be configured to estimate the second maximum deflection portion 180b based on the current diaphragm deflection signal block 130.sub.2 by means of a neural network. By using a neural network, the apparatus 100 can be trained and thereby not only very efficiently perform modification of the diaphragm deflection signal 130, but additionally realize very exact modification that can result in efficiency increase/a faster algorithm.

(40) FIG. 3 shows an embodiment of the apparatus 100, in other words, e.g., an overall signal flow in a deflection limiter (e.g., the apparatus 100).

(41) In the following, the apparatus 100 can also be referred to as deflection limiter. According to an embodiment, the apparatus 100 can be configured to modify a loudspeaker signal U, 110 to prevent diaphragm over-deflection, e.g., X>X.sub.max, to predict 135 a diaphragm deflection signal X block-by-block in overlapping time blocks based on the loudspeaker signal U, 110 to obtain one diaphragm deflection signal block X.sub.i per time block. Here, e.g., the complete loudspeaker signal U, 110 can be detected by the apparatus 100 or block-by-block in overlapping time blocks. In that way, a current loudspeaker signal of a current time block can be referred to, for example, as U.sub.i and X.sub.i can be a current diaphragm deflection signal block of a same time block. Here, i can be a positive integer. The diaphragm deflection signal X can have the same features and functionalities as the diaphragm deflection signal 130 of FIG. 1 or FIG. 2, and the current diaphragm deflection signal block X.sub.i can have the same features and functionalities as the diaphragm deflection signal block 130.sub.2 of FIG. 1 or FIG. 2. Predicting 135 the diaphragm deflection signal X can take place, for example by means of a non-linear deflection model.

(42) The apparatus 100 can be configured to determine a temporal position k.sub.0 of a maximum deflection X.sub.i[k.sub.0] of a current diaphragm deflection signal block X.sub.i of a current time block i within an overlap area with a subsequent time block i+1. This takes place, e.g., in the signal analysis 150.

(43) The apparatus 100 can be configured to calculate a level (e.g., h.sub.lX.sub.max) up to which the current diaphragm deflection signal block X.sub.i can be deflected without diaphragm over-deflection for the current time block i by considering a comparison of the current diaphragm deflection signal block (e.g., X.sub.i[k.sub.0]) with a subsequent diaphragm deflection signal block (e.g., X.sub.i+1[k.sub.0-M.sub.step]) or an estimation of the subsequent diaphragm deflection signal block (e.g., X.sub.i+1[k.sub.0-M.sub.step]) from the current diaphragm deflection signal block X.sub.i at the temporal position (e.g., k.sub.0) in the current time block i and k.sub.0-M.sub.step in the subsequent time block i+1, wherein k.sub.0 and k.sub.0-M.sub.step represent the corresponding point in time for the respective diaphragm deflection signal block in the overlap area of the two time blocks. This can take place in the signal analysis 150.

(44) Further, the apparatus 100 can be configured to attenuate the current diaphragm deflection signal block X.sub.i based on the level (h.sub.lX.sub.max) to obtain a modified current diaphragm deflection signal block {tilde over (X)}.sub.i. This can take place with the computing unit 155.

(45) Further, the apparatus 100 can be configured to synthesize 175 a modified loudspeaker signal Ũ based on the modified current diaphragm deflection signal block {tilde over (X)}.sub.i. Here, for example, the complete modified loudspeaker signal Ũ can be synthesized 175 based on a completely modified diaphragm deflection signal Ũ or block-by-block in overlapping time blocks, such that a current modified loudspeaker signal Ũ.sub.i is synthesized 175, for example, based on the current modified diaphragm deflection signal block {tilde over (X)}.sub.i of a same time block. Here, i can be a positive integer and can comprise the complete modified diaphragm deflection signal {tilde over (X)} of the current diaphragm deflection signal block {tilde over (X)}.sub.i. The modified diaphragm deflection signal {tilde over (X)} can have the same features and functionalities as the modified diaphragm deflection signal 160 of FIG. 1 or FIG. 2, the current modified diaphragm deflection signal block {tilde over (X)}.sub.i can have the same features and functionalities as the current modified diaphragm deflection signal block 160.sub.2 of FIG. 1 or FIG. 2, the modified loudspeaker signal Ũ can have the same features and functionalities as the modified loudspeaker signal 170 of FIG. 1 or FIG. 2, and the current modified loudspeaker signal Ũ.sub.i can have the same features and functionalities as the current modified loudspeaker signal 170.sub.2 of FIG. 1 or FIG. 2. Synthesizing 175 the modified loudspeaker signal Ũ can take place, e.g., by means of an inverse nonlinear deflection model. Optionally, the apparatus 100 can transmit the modified loudspeaker signal Ũ or the current modified loudspeaker signal Ũ.sub.i to a loudspeaker 200.

(46) Optionally, the apparatus 100 can be configured to a current loudspeaker signal U.sub.i provide instead of the modified current loudspeaker signal block Ũ.sub.i and to transmit the same, e.g., to the loudspeaker 200 when a maximum deflection of the current diaphragm deflection signal block X.sub.i of the current time block i does not exceed a maximum deflection limiting value X.sub.max [e.g. max.sub.k|X.sub.i[k]|<X.sub.max], which the apparatus 100 can perform by a query 115.

(47) Further, the apparatus 100 can be configured to decompose the current diaphragm deflection signal block (X.sub.i) into at least one frequency band signal time block (X.sub.i,B0 to X.sub.i,BN, with N frequency band signal time blocks, wherein N is a positive integer) with a frequency band decomposition 151.

(48) According to an embodiment, the apparatus can be configured to determine, for each of the at least one frequency band signal time blocks X.sub.i,B0 to X.sub.i,BN, at least one further temporal position where a predetermined signal combination exceeds a maximum deflection limiting value X.sub.max in the first overlap area of the current time block i with the preceding time block i−1 as well as in the reciprocal time block i. This takes place, e.g., by means of the signal analysis 150.

(49) According to an embodiment, the apparatus 100 can be configured to determine an attenuation factor g.sub.0 to g.sub.N for each of the at least one frequency band signal time block X.sub.i,B0 to X.sub.i,BN, based on the level h.sub.lX.sub.max and the further temporal position to attenuate the current diaphragm deflection signal block X.sub.i. This takes place, e.g., by means of the attenuation factor determination 153a.

(50) According to an embodiment, the apparatus 100 can be configured to compare an attenuation factor g.sub.0 to g.sub.N per frequency band signal time block X.sub.i,B0 to X.sub.i,BN for the current time block i with a version of the attenuation factor g.sub.0 to g.sub.N per frequency band signal time block X.sub.i,B0 to X.sub.i,BN for the preceding time block i−1, reduced in attenuation strength by a predetermined fading-away function, and to use a selected attenuation factor g.sub.0 to g.sub.N that is associated with a higher attenuation for the current time block of the same. This takes place, e.g., via the attack, hold and release functions 153b. If it applies according to an embodiment that a current g.sub.N is smaller than an old g.sub.N, the attack, hold and release functions 153b should select the current g.sub.N, otherwise the release function should still be followed, wherein, for example, a higher attenuation factor g.sub.N can cause less attenuation. According to an embodiment, an attenuation factor g=1 can define an attenuation of 0 dB and an attenuation factor g=0.1 can define an attenuation of 20 dB.

(51) According to an embodiment, the attenuation factor determination 153a and the attack, hold and release functions 153b can have the same features and functionalities as the attenuation factor determination 153 of FIG. 1 or FIG. 2.

(52) Optionally, the attenuation factor determination 153a can use a psychoacoustic model 190 to optimize determination of the attenuation factors g.sub.0 to g.sub.N and to thereby improve the sound quality of the modified loudspeaker signal Ũ or the current loudspeaker signal block Ũ.sub.i.

(53) Thereby, the embodiment illustrated in FIG. 3 can represent an intelligent signal analysis 150, both of an overall deflection (e.g., of the current diaphragm deflection signal block (X.sub.i) as well as the deflection in the respective frequency bands (e.g., in the individual frequency band signal time blocks X.sub.i,B0 to X.sub.i,BN), which can provide for the fact that the deflection limiting value X.sub.max can be approximated without unnecessary headroom. Headroom means an unused portion of a maximum deflection of a diaphragm of the loudspeaker 200. Thus, when a headroom exists, the diaphragm is hardly or never completely deflected. If the headroom is too large, no high performance can be obtained at the loudspeaker. The apparatus causes, e.g., only a very small or no headroom, whereby a very high performance of the loudspeaker 200 can be obtained. Further, FIG. 3 can represent a system that can combine the best partial approaches to an optimum overall system and can comprise a separate attack, hold and release regulation 153b for percussive and harmonic signal portions, which result in an improvement of the sound quality.

(54) Preprocessing

(55) In other words, FIG. 3 shows an embodiment of a basic signal flow in the inventive deflection limiter 100. An input voltage U (e.g., the loudspeaker signal U, 110) is input, e.g., block-by-block (e.g., in time blocks i) into the system (e.g., the apparatus 100). The time blocks i can have a common temporal block size that defines, e.g., the length of a time block i. Block sizes ≥10 ms have proven to be particularly good. Block sizes of ≥50 μs, ≥1 ms, ≥5 ms, ≥12 ms, ≥15 ms, ≥20 ms oder ≥50 ms are also possible. A voltage curve in the i-th block can be referred to as U.sub.i. The voltage curve in the i-th block can be an example for the current loudspeaker signal block U.sub.i in the current time block i.

(56) For increasing a velocity of subsequent processing steps, low-pass filtering and undersampling may be performed, as high frequencies are hardly relevant for diaphragm deflection.

(57) The voltage U.sub.i is used, e.g., as input signal of a non-linear deflection model for prediction 135 of the deflection X.sub.i (the current diaphragm deflection signal block X.sub.i). The deflection model allows control without feedback path. Possible embodiments of the deflection models are a simple filter, a physical structural model (e.g., implemented in the state space) or a machine-learning model (e.g., neural network). The deflection model is not limited to a specific actuator principle—models for electrodynamic loudspeakers, piezoelectric loudspeakers, and electrostatic loudspeakers can be used in a modular manner. Additionally, the deflection model can be implemented in an adaptive i.e. time-variable manner to adapt the parameters continuously to the loudspeaker in an optimum way. In this case, the apparatus can comprise a feedback path. All listings of the deflection models and actuator principles stated herein are to be considered as exemplary and not limiting.

(58) If the predicted deflection (e.g., an amplitude of the current diaphragm deflection signal block X.sub.i) does not exceed the limiting value X.sub.max, the input signal is passed on directly to the loudspeaker. Otherwise, the predicted deflection signal is, e.g., processed further and decomposed into N≥1 adjacent frequency bands (e.g., the frequency band signal blocks X.sub.i,B0 to X.sub.i,BN). This corresponds, for example, to the mode of operation of a filterbank. Possible implementations are, for example, adjacent band pass filters or also perfect reconstruction filterbanks like MDCT (modified discrete cosine transformation) or PQMF (pseudo quadrature mirror filters). It should be considered that this filterbank (e.g., the frequency band decomposition 151) and the psychoacoustic model 190 should consider the same frequency bands. All listings of the methods for decomposing the predicted deflection signal X.sub.i into adjacent frequency bands (B0 to BN, wherein N is, for example, a natural number between 1 and 100) stated herein are to be considered as exemplary and not limiting.

(59) Signal Analysis and Calculation of the Attenuation Factors

(60) First, e.g., a signal analysis 150 is performed by using all time signals of the individual frequency bands X.sub.i,Bn (wherein N is a natural number between 0 and N) as well as the overall deflection signals {tilde over (X)}.sub.i−1, X.sub.i and possibly X.sub.i+1 (wherein {tilde over (X)}.sub.i−1 defines, e.g., a modified overall deflection signal of the time block i−1 preceding the current time block i). A block diagram for a possible embodiment is illustrated in FIG. 4.

(61) FIG. 4 shows a block diagram illustrating a signal flow, e.g., of the apparatus in the signal analysis that can comprise the same features and functionalities as the processing means 150 of FIG. 1 or FIG. 2 and as the signal analysis 150 of FIG. 3.

(62) First, e.g., based on an overlap-add processing, it is considered how much the overall current block i should at least be attenuated so that X.sub.max is not exceeded in the overlap area with the subsequent block i+1. Here, it should be noted that the subsequent block i+1 is not attenuated too much. The latter measure is not needed to guarantee the protection of the transducer but to maximize the sound quality. Therefore, the look-ahead headroom h.sub.a (e.g., the second maximum deflection portion 180b, h.sub.a) is calculated, e.g., by

(63) $k_o=\arg \max .Math.X_i\left[M_{step}\mathrm{.Math.}{M}_{block}\right].Math.$$h_a=\frac{X_i\left[k_o\right]}{X_i\left[k_o\right]+X_{i+1}\left[k_o-M_{step}\right]}s|s\in \left[0;1\right]$
wherein X.sub.i can represent the time signal of the overall deflection in the i-th signal block (e.g., the current diaphragm deflection signal block), M.sub.block the block length and M.sub.step the step width of the overlap-add signal processing and s a parametric safety factor. In the following, the second maximum deflection portion 180b, h.sub.a can also be referred to as look-ahead headroom. By the look-ahead, e.g., the signal delay increases by M.sub.step samples. A respective prediction model for the future deflection curve (e.g., the subsequent diaphragm deflection signal block X.sub.i+1) based on the current signal curve (e.g., the current diaphragm deflection signal block X.sub.i+1), e.g., by means of a neural network or a statistic model for direct prediction of h.sub.a, based on the current signal block (e.g., the current diaphragm deflection signal block X.sub.i), could eliminate the consideration of the second signal block (e.g., the subsequent diaphragm deflection signal block X.sub.i+1). FIG. 5a, FIG. 5b and FIG. 5c illustrate possible embodiments of the look-ahead headroom calculation 180b.

(64) FIG. 5a, FIG. 5b and FIG. 5c show embodiments of the look-ahead headroom calculation. By calculating the future signal curve (see FIG. 5a) with the estimation of the future signal curve based on the current signal block (see FIG. 5b) with direct estimation of the look-ahead headroom h.sub.a based on the current signal curve (see FIG. 5c).

(65) Thus, FIG. 5a shows, e.g., a calculation of the second maximum deflection portion 180b, h.sub.a based on a prediction 135 of the current diaphragm deflection signal block X.sub.i based on the current loudspeaker signal U.sub.i and the subsequent diaphragm deflection signal block X.sub.i+1 based on the subsequent loudspeaker signal wherein the prediction 135 takes place, e.g., by means of a non-linear deflection model. Thus, the apparatus 100 can determine exactly how much the current diaphragm deflection signal block X.sub.i should be attenuated so that a high mechanical protection of a transducer (e.g., diaphragm, suspension) of the loudspeaker 200 can be ensured by the apparatus.

(66) FIG. 5b shows, e.g., a calculation of the second maximum deflection portion 180b, h.sub.a, wherein the subsequent diaphragm deflection signal block X.sub.i+1 is estimated 136 based on the current diaphragm deflection signal block X.sub.i, whereby the second maximum deflection portion 180b, h.sub.a can be calculated faster and, hence, more efficiently. The second maximum deflection portion 180b, h.sub.a can be determined by means of the apparatus from the current diaphragm deflection signal block X.sub.i and the estimated subsequent diaphragm deflection signal block X.sub.i+1. The estimation 136 of the subsequent diaphragm deflection signal block X.sub.i+1 can take place, e.g., via a neural network.

(67) FIG. 5c shows, e.g., a calculation of the second maximum deflection portion 180b, h.sub.a, wherein the second maximum deflection portion 180b, h.sub.a is estimated 182 based on the current diaphragm deflection signal block X.sub.i. Thus, the second maximum deflection portion 180b, h.sub.a can be calculated even faster and more efficiently, as the subsequent diaphragm deflection signal block X.sub.i+1 is not needed for the calculation. The estimation 182 can take place, e.g., via a neural network.

(68) The calculation of the look-ahead headroom h.sub.a is explicitly not based on the subsequent signal analysis steps, i.e. the same can take place separately with the respective input signals.

(69) According to FIG. 4, a next step in the signal analysis 150 according to an embodiment can include the identification of critical deflection values 152 in the predicted signals (e.g., the diaphragm deflection signal X, such as X.sub.i and X.sub.i+1 or the frequency band signal blocks X.sub.i,B0 to X.sub.i,BN), which can later serve to calculate the attenuation factors for the individual frequency bands.

(70) For searching the critical deflection values 152 (or further temporal positions), e.g., three different signal combinations (e.g., additive combinations) can be used:

(71) 1. X.sub.alt+X.sub.i,

(72) 2. X.sub.alt+X.sub.i,B0 . . . X.sub.alt+X.sub.i,BN,

(73) 3. |X.sub.alt|+|X.sub.i|,

(74) wherein

(75) $X_{\mathrm{alt}}\left[k\right]=\{\begin{array}{c}{\tilde{X}}_{i-1}\left[k+M_{\mathrm{step}}\right],k<M_{block}-M_{\mathrm{step}}\\ 0,k\ge M_{block}-M_{\mathrm{step}}\end{array},$

(76) In the three signal combinations, e.g., all extremes exceeding the deflection limit X.sub.max are identified as critical deflection values. In that way, e.g., L critical deflection values and their indices k.sub.c,l with 0≤l≤L are found. The critical deflection values serve to prevent optimization of the attenuation factors with random start values, which accelerates the algorithm.

(77) In the following, a look-back headroom 180a can be a first maximum deflection portion. According to an embodiment, a processing loop having L iterations starts. Here, first the look-back headroom is calculated, e.g., for the l-th critical deflection value having the index k.sub.c,l, i.e., how large the overall attenuation of the current block i should at least be so that the overlap with the preceding already processed block i−1 does not result in exceeding the limiting value X.sub.max. The look-back headroom for the l-th critical deflection value h.sub.b,l is calculated, e.g., by:

(78) $h_{b,l}=\frac{\mathrm{sign}\left(X_i\left[k_{c,l}\right]\right)X_{\max }-X_{\mathrm{alt}}\left[k_{c,l}\right]}{X_{\max }}$

(79) A final headroom h.sub.l, 184 for the l-th critical deflection value is calculated, e.g., as
h.sub.l=sign(X.sub.i[k.sub.c,l])min(|h.sub.b,l|,|h.sub.a|).

(80) Thus, according to an embodiment, the apparatus can be configured to calculate a level (h.sub.lX.sub.max) based on the first maximum deflection portion (h.sub.b,l) when the first maximum deflection portion (h.sub.a) is smaller than the second maximum deflection portion (h.sub.a), or to calculate the level (h.sub.lX.sub.max) based on the second maximum deflection portion (h.sub.a) when the second maximum deflection portion (h.sub.a) is smaller than the first maximum deflection portion (h.sub.b,l).

(81) FIG. 4 shows, e.g., the signal analysis 150 of FIG. 3 in detail.

(82) According to FIG. 3, for each of the N frequency bands, e.g., a linear attenuation factor g.sub.0 to g.sub.N can be calculated by means of the attenuation factor determination 153a (here, linear means, e.g., that no additional nonlinearities are introduced in the respective signal block which would be the case in a compressor). Here, according to an embodiment, the deflection in every frequency band is weighted with a psychoacoustic model 190, whereby the sound quality and loudness are to stay as high as possible. The psychoacoustic model 190 consists of a simple variation, e.g., of the A evaluation according to DIN EN 61672-1 2003-10, but can also represent a nonlinear implementation of the curves having the same loudness or a more complex model (e.g., masking in the time and frequency domain). All listings of psychoacoustic models stated herein are to be considered as exemplary and not limiting.

(83) In that way, e.g., all L critical deflection values are processed until, e.g., an optimum reduction of the maximum overall deflection is obtained. If the overall deflection cannot be lowered below the limiting value h.sub.lX.sub.max based on the critical deflection values, possibly, LMS optimization with the best attenuation factors up to then can be used as start values to find adequate attenuation factors.

(84) If the attenuation factors for the current diaphragm deflection signal block X.sub.i are known, the same are regulated, e.g., via attack, hold and release functions (AHR) 153b, which can again be independent of one another within the individual frequency bands. Here, additionally, different AHR functions (e.g., a percussive AHR 153b, and a harmonic AHR 153b.sub.2) can be applied to the percussive signal portion and the harmonic signal portion, which can improve the sound quality further (shorter time constants are, e.g., suitable for percussive signals, longer ones, e.g., for harmonic signals). For this, the signal X.sub.i should previously be divided into the respective portions by harmonic-percussive decomposition 151b and then be processed by the filterbank (frequency band decomposition 151). FIG. 6 represents this process schematically.

(85) FIG. 6 shows an embodiment for different attack, hold and release functions 153b.sub.1, 153b.sub.2 for percussive and harmonic signal portions that can be a detail of the attack, hold and release functions 153b of FIG. 3. According to an embodiment, the attenuation factor determination 153 of FIG. 1 or FIG. 2 can have the same features and functionalities as the different attack, hold and release functions 153b.sub.1, 153b.sub.2.

(86) The attenuation factors g.sub.0 to g.sub.N can obtain and adapt both the percussive AHR 153b.sub.1 as well as the harmonic AHR 153b.sub.2 from the attenuation factor determination 153a (illustrated in FIG. 3).

(87) According to an embodiment, the signal analysis 150 (headroom calculation 180a, 180b, search for critical deflection values 152) and/or the calculation of the attenuation factors 153, 153a, 153b can possibly take place in another way, wherein the final result can be very similar or identical.

(88) According to an embodiment, the apparatus 100, the method 300 or the method 400 can be used in audio signal processing, the protection of loudspeakers, microspeakers in mobile devices (smartphones, tablets, notebooks, . . . ), large PA loudspeakers close to the mechanical performance limit and small transducers in Bluetooth loudspeakers.

(89) According to an embodiment, the attack, hold and release functions (AHR) 153b can comprise both the percussive AHR 153b.sub.1 as well as the harmonic AHR 153b.sub.2, as well as a synthesis of a percussive 155a and a harmonic 155b modified diaphragm deflection signal of the current time block, but it is also possible that the attack, hold and release functions (AHR) 153b only comprise the percussive AHR 153b.sub.1 and the harmonic AHR 153b.sub.2 and the computing means 155 of FIG. 1, FIG. 2 or FIG. 3 comprises the synthesis of a percussive 155a and harmonic 155b modified diaphragm deflection signal of the current time block, as well as summing up 155 the percussive 155a and harmonic 155b modified diaphragm deflection signal of the current time block of FIG. 6. In other words, in 155a and 155b, the frequency bands can already be summed up separately for a percussive 155a and harmonic 155b diaphragm deflection signal portion. Summing up 155 in FIG. 6 then sums up the two diaphragm deflection signal portions.

(90) Signal Synthesis

(91) The percussive 155a and the harmonic 155b modified diaphragm deflection signal of the current time block is then, e.g., summed up 155, and from the limited deflection g, (e.g., the modified current diaphragm deflection signal), the voltage that has to be supplied to the loudspeaker 200 to cause the desired deflection is calculated, e.g., with the help of an inverse nonlinear deflection model 175. This inverse model can also be configured as simple filter, physical structural model or machine-learning model. All listings for inverse models stated herein are to be considered as exemplary and not as limiting. The summing up 155 as well as the percussive 155a and harmonic 155b modified diaphragm deflection signals of the current time block of FIG. 6 can comprise features and functionalities of the computing means 155 of FIG. 1, FIG. 2 and FIG. 3, and/or the computing means 155 of FIG. 1, FIG. 2 and FIG. 3 can comprise features and functionalities of the summing up 155 as well as the percussive 155a and harmonic 155b modified diaphragm deflection signal of the current time block of FIG. 6.

(92) Effects and Advantages of the Technical Features of the Apparatus:

(93) The signal analysis 150 can be configured to select attenuation coefficients, e.g., such that the deflection is optimally used without unnecessary headroom and attenuation coefficients are, e.g., not optimized with random start values. This is advantageous as the deflection can be maximized and the attenuation factors can be calculated faster than with a pure optimization.

(94) The nonlinear look-ahead and look-back models can be configured to allow an exact prediction of the transducer behavior (e.g., the diaphragm deflection signal) with respective parameters. This is advantageous as the performance of the transducer (e.g., the loudspeaker 200) can be used better and no feedback path is needed.

(95) The multiband approach (e.g., the frequency band decomposition 151) in combination with block-by-block constant attenuation factors can be configured to generate no new nonlinearities within the signal blocks (e.g., the diaphragm deflection signal blocks as, e.g., X.sub.i, X.sub.i+1). This is advantageous since, compared to a compressor approach, less sound artefacts can occur by nonlinearities.

(96) The psychoacoustic model 190, e.g., an A evaluation, curves of the same loudness, masking in the time and frequency domain can be configured to select attenuation factors with regard to human hearing in an optimum manner. This is advantageous as sound quality and loudness can be maximized at the highest possible deflection.

(97) Thus, the invention described herein can be advantageous as, according to an embodiment of the present invention, an output voltage can be calculated by an inverse nonlinear loudspeaker model (X.fwdarw.U), which can predict the diaphragm deflection signal more accurately than a model according to US20180014121A1. Additionally, for example, a psychoacoustic model is used for calculating attenuation factors (in the simplest configuration, e.g., the A evaluation) to increase the sound quality.

(98) In A MODEL BASED EXCURSION PROTECTION ALGORITHM FOR LOUDSPEAKERS there is no multiband approach but all frequencies are attenuated with a compressor, which again results in nonlinearities (THD, IMD) in the signal by the nonlinear characteristic curve of the compressor. According to an embodiment of the present invention, there is a multiband approach and a psychoacoustic model for calculating the attenuation factors (in the simplest configuration, e.g., the A evaluation), whereby mechanical protection, e.g., of a diaphragm of a loudspeaker can be increased and the sound quality can also be increased.

(99) According to an embodiment of the present invention, nonlinear deflection models (forward and inverse) are used, however, in U.S. Pat. No. 8,855,322B2 the attenuation factors are applied, e.g., directly to the voltage signal; predicting the actual transducer behavior is thereby more accurate in the present invention than in U.S. Pat. No. 8,855,322B2. The optimization for calculating the subband attenuation factors uses, according to U.S. Pat. No. 8,855,322B2, an edge condition (namely multiplication of the amounts of band energy and transfer function by disregarding the phase position and possible cancellation), which has the effect that, according to the algorithm, the overall deflection frequently stays below the limiting value and the performance of the transducer is not completely used. According to an embodiment of the present invention, for example, the signal analysis has the effect that the phase position is considered and no unnecessary headroom remains.

(100) According to an embodiment of the present invention, nonlinear models are used, or an accurate signal analysis is performed in the frequency bands, in contrary to U.S. Pat. No. 9,807,502BA.

(101) FIG. 7 shows a block diagram of a method 300 for modifying a loudspeaker signal for preventing diaphragm over-deflection with block-by-block prediction 310 in overlapping time blocks based on the loudspeaker signal of a diaphragm deflection signal to obtain one diaphragm deflection signal block per time block. Further, the method 300 can comprise determination 320 of a temporal position of a maximum deflection of a current diaphragm deflection signal block of a current time block with an overlap area with a subsequent time block. Further, the method 300 can comprise calculation 330 of a level up to which the current diaphragm deflection signal block can be controlled without diaphragm over-deflection for the current time block, while considering a comparison of the current diaphragm deflection signal block with a subsequent diaphragm deflection signal block or an estimation of the subsequent diaphragm deflection signal block from the current diaphragm deflection signal block at the temporal position. Further, the method 300 can comprise an attenuation 340 of the current diaphragm deflection signal block based on the level to obtain a modified current diaphragm deflection signal block and synthesization 350 of a modified loudspeaker signal based on the modified current diaphragm deflection signal block.

(102) FIG. 8 shows a block diagram of a method 400 for modifying a loudspeaker signal for preventing diaphragm over-deflection with block-by-block prediction 310 in overlapping time blocks based on the loudspeaker signal of a diaphragm deflection signal to obtain one diaphragm deflection signal block per time block. Further, the method 400 can comprise determination 410 of a first maximum deflection portion for a current diaphragm deflection signal block of a current time block in a first overlap area of the current time block with a preceding time block as well as in a residual area of the current diaphragm deflection signal block. Further, the method 400 can comprise determination 420 of a second maximum deflection portion for the current diaphragm deflection signal block of the current time block in a second overlap area of the current time block with a subsequent time block. Further, the method 400 can comprise calculation 330 of a level based on the first maximum deflection portion when the first maximum deflection portion is smaller than the second maximum deflection portion or calculation of the level based on the second maximum deflection portion when the second maximum deflection portion is smaller than the first maximum deflection portion. Further, the method 400 can comprise attenuation 340 of the current diaphragm deflection signal block based on the level to obtain a modified current diaphragm deflection signal block and synthesization 350 a modified loudspeaker signal based on a modified current diaphragm deflection signal block.

(103) While this invention has been described in terms of several advantageous embodiments, there are alterations, permutations, and equivalents which fall within the scope of this invention. It should also be noted that there are many alternative ways of implementing the methods and compositions of the present invention. It is therefore intended that the following appended claims be interpreted as including all such alterations, permutations, and equivalents as fall within the true spirit and scope of the present invention.