Method for automatically controlling the loudness of an audio signal and apparatus for automatically controlling the loudness of an audio signal

Abstract

If the measured loudness is equal to or greater than the first target loudness, the gain factor is zero. Otherwise, if the noise level indication signal indicates a noise level equal to or greater than the intermediate noise level, the gain factor is the difference between the measured loudness and the current target loudness, limited to a maximum gain. The maximum gain is the absolute maximum gain, and the current target loudness is linearly interpolated according to the noise level indication signal between the second target loudness for a noise level equal to the intermediate noise level and the first target loudness for a noise level equal to the maximum noise level. Otherwise, if the noise level indication signal indicates a noise level less than the intermediate noise level, the gain is linearly interpolated according to the noise level between the gain defined for the intermediate noise level and zero. Further, the gain factor is applied to the reproduced audio signal.

Description

[Technical Field]

[0001] The present invention relates to a method for automatically controlling the loudness of an audio signal. The present invention also relates to a device for automatically controlling the loudness of an audio signal. In particular, the present invention relates to controlling the loudness of an audio signal played in a vehicle such as an automobile. [Background technology]

[0002] Vehicles have recently become a common environment for listening to music or other audio, such as audiobooks. However, moving vehicles generate speed-dependent noise, which can interfere with the listener's enjoyment. Therefore, many cars use very simple speed-dependent volume scaling for audio, meaning that the louder the sound, the faster the speed. This is because speed is used as a substitute for the noise level inside the vehicle. Various more sophisticated solutions are known, such as circuits for controlling the loudness of audio played inside a moving vehicle. Typically, such circuits adapt the gain of the music being played according to the ambient noise level detected by, for example, a microphone inside the vehicle.

[0003] When listening to music while driving, road noise can partially mask the music, potentially degrading the listening experience in the car. Whether or not music is masked by surrounding driving noise is, to some extent, determined by the loudness ratio.

[0004] Loudness ratio = Music loudness / Noise loudness [Overview of the project] [Problems that the invention aims to solve]

[0005] Conventional solutions attempt to maintain a relatively constant loudness ratio by adjusting the loudness of the music, thereby accounting for changes in the loudness of the noise. However, if the loudness of the music itself changes, the loudness ratio may also change. For example, quiet sections of music may be completely overwhelmed by noise, even if the larger sections are not. This problem is particularly affected in classical music due to its wide dynamic range. Current solutions to the problem of in-car noise do not take this challenge into account. [Means for solving the problem]

[0006] This invention is based on the recognition that, in a moving vehicle, controlling the loudness of reproduced sound should take into account not only speed or ambient noise levels, but also the dynamics of the sound itself. In particular, this invention provides a reduction in the speed dependence of the dynamic range of music by applying a greater compensation gain to quieter sections of music compared to louder sections. This reduces the masking of music by ambient noise, thereby allowing quieter sections of music to be heard even in the presence of road noise.

[0007] In principle, in one embodiment of the present invention, the present invention relates to a method for automatically controlling the loudness of an audio signal played in a vehicle, wherein the loudness of the incoming audio signal is measured, an ambient noise level indicator signal is received, such as a speed indicator signal or signals from one or more microphones, and a gain coefficient is calculated based on both the measured loudness and the noise level indicator signal. To calculate the gain coefficient, an intermediate speed or noise level and a corresponding gain curve are defined, and two different interpolations are used. If the speed or noise level is below the intermediate level, one interpolation is used, and if the speed or noise level is above the intermediate level, the other is used. The gain coefficient is calculated according to a continuous function, and there is no discontinuity between the two interpolations, which has the advantage that as long as the speed and / or noise level change continuously, there will be no sudden changes in the gain coefficient.

[0008] In another embodiment, the present invention relates to a device for automatically controlling the loudness of an audio signal played in a vehicle. The device comprises one or more loudness meters adapted to measure the loudness of an audio signal arriving over a defined time period; an input circuit adapted to receive a noise level indicator signal, such as a speed indicator signal; and a processor adapted to calculate a gain coefficient based on the measured loudness and noise level indicator signals.

[0009] In yet another embodiment, the present invention relates to a computer-readable storage medium that stores computer-executable instructions that, when executed on a computer, cause the computer to perform the method described herein. [Brief explanation of the drawing]

[0010] [Figure 1] This is a schematic diagram of the components of an adaptive level regulator. [Figure 2] This is a block diagram of an adaptive level regulator. [Figure 3] This is an exemplary gain curve for calculating velocity-dependent gain coefficients. [Figure 4] This is an exemplary gain curve for gain coefficients applied to various levels of measured loudness, each dependent on the ambient noise level. [Figure 5] This is a block diagram of a loudness meter. [Figure 6] This is a block diagram for calculating the gain. [Modes for carrying out the invention]

[0011] The following explanation can be better understood by referring to the attached diagram.

[0012] Figure 1 shows a schematic diagram of the components of an adaptive speed-dependent level tuner according to an embodiment of the present invention. In this application, a level tuner is understood as a device that takes into account the loudness of an audio signal to be reproduced in order to control the gain applied to that signal. The speed-dependent level tuner 100 receives a signal indicating the current speed of the vehicle, and it uses this signal as one of at least two inputs for determining the gain coefficient for the input audio signal to be reproduced. Speed ​​acts as a substitute for loudness in the vehicle. The loudness of the input audio signal, also called the program material signal, is measured in at least one loudness meter 131,132 and is used as a second input for determining the gain coefficient. In principle, the static level tuner 120 calculates the gain coefficient based on the measured loudness of the program material, and it follows a gain function. The static level tuner 120 becomes a dynamic or adaptive level tuner in the sense that some parameters of the static level tuner change with some characteristics of the external world. In this example, the gain function of the static level adjuster depends on the vehicle's current speed, which typically does not change as rapidly as the dynamics of the program material. More specifically, the current gain function follows one of three predefined gain functions or is interpolated between two of the three predefined gain functions, depending on the current speed, as outlined in the three cases below.

[0013] - If the current velocity is zero, the first gain function is used; - If the current speed is at a predefined threshold speed, also called the intermediate speed, then the second gain function is used; - If the current speed is greater than or equal to the predefined maximum speed, a third gain function is used; - If the current velocity is between zero and a predefined threshold or intermediate velocity, it is interpolated between the first and second gain functions; - If the current speed is between a threshold or intermediate speed and a defined maximum speed, it is interpolated between the second and third gain functions.

[0014] More information regarding Figure 3 is provided below. Note that when the vehicle's current speed matches a threshold or intermediate speed, the gain calculation follows a second gain function. In one embodiment, a combination of two or more loudness meters 131,132 is used, as will be described in detail later. More generally, instead of speed and signals indicating the vehicle's current speed, noise level indicators and noise level signals indicating ambient noise can also be used.

[0015] Figure 2 shows a more detailed block diagram of the adaptive level adjuster. The level adjuster 200 includes a loudness meter 210, which can measure the loudness of the input audio signal 205. For example, in one embodiment, the loudness may be measured using LKFS (Loudness, K-weighted, relative to Full Scale) or LUFS (Loudness Units in Full Scale). However, the loudness may be measured using any suitable device. For example, in one embodiment, the loudness is measured using a proprietary method. Further details regarding loudness measurement are provided below with respect to Figures 5 and 6. The loudness of the input audio signal is adapted (i.e., amplified) to meet a target loudness that depends on the loudness of the ambient noise. The applied gain 225 changes over time, depending on the loudness of the input audio signal 205 and the loudness of the ambient noise. For adaptive level adjustment, a gain calculation module 220 circuit is used. The circuit may be configured in software. The gain calculation module 220 receives at least one value 215 representing the measured loudness of the audio signal 205 and at least one value 201 representing the loudness of the ambient noise. The value 201 representing the loudness of the ambient noise may be, for example, a signal indicating the current speed of the vehicle, the rotational speed of the vehicle's engine, or signals from one or more microphones. More specifically, the gain calculation module 220 may determine the current target loudness and the current maximum allowable gain, which are parameters of a static level tuner that depend on the ambient noise level inside the vehicle. This has the advantage that the audio is amplified sufficiently so that it is not masked by ambient noise, allowing listeners inside the vehicle to hear the audio better.

[0016] The gain calculation module 220 may be implemented by one or more appropriately configured processors, which provide a gain value 225 to the amplifier 230, which applies the calculated gain value to the input audio signal 205 to obtain an amplified output audio signal 235. The output audio signal 235 may be provided to an amplifier or any other conventional module 300, which may optionally perform a velocity-dependent gain on the audio signal. Module 300 may also scale the gain with respect to the noise level in the vehicle. Module 300 and / or any subsequent modules may apply a second gain coefficient independent of the first gain coefficient applied by the level adjuster 200, such that the first and second gain coefficients together determine the loudness of the audio signal being reproduced. Both gain coefficients are independent in this embodiment of a master volume, which is a later gain stage that can be manually controlled by the user. In other embodiments, however, the gain coefficients may be coupled to the master volume. For example, if the user is listening at a lower volume level, a stronger effect (i.e., larger level adjustment and a smaller dynamic range) can be applied, while if the user is listening at a higher volume level, a weaker effect (i.e., smaller level adjustment and a larger dynamic range) can be applied.

[0017] Figure 3 shows an exemplary gain curve for calculating the velocity-dependent gain coefficient in one embodiment. Which gain curve should be applied now depends on the vehicle's velocity. Thus, each gain curve may be considered a different parameter representation of the static level adjuster. The horizontal axis represents the measured loudness L of the input signal 205. M 215 is expressed in dBFS (decibels relative to full scale), while the vertical axis represents the gain G applied by the level adjuster 200. A This represents 225. Note that while all numbers have been proven to be favorable for a pleasing sonic image in music, they are illustrative.

[0018] In this example, for a very large input signal having a measured loudness above the first target loudness TLaMS (Target Loudness at Maximum Speed) which is set to -14 dBFS, no additional amplification is required, and thus the applied gain G A is set to zero dB. Independently of the measured loudness, when the vehicle is not moving, i.e., when the speed is zero, the applied gain G A is also always set to zero dB as shown by the first gain curve G0. In other cases, i.e., for an input signal having a loudness less than the first target loudness TLaMS and a non-zero vehicle speed, the applied gain G A depends on both the measured loudness and speed. Interpolation may be performed, and the type of interpolation depends on the current speed and on its relationship to a predefined or configured maximum speed and its relationship to a predefined or configured intermediate speed. The intermediate speed is also called the threshold speed. [[ID=⑧]]

[0019] [[ID=⑨]] If the current speed matches or exceeds the defined maximum speed, the applied gain follows the third gain curve G 100 As described above, if the measured loudness is greater than the first target loudness TLaMS, the third gain curve G 100 includes a zero gain section. Further, if the measured loudness is less than the first target loudness TLaMS minus the maximum allowable gain G max , i.e., TLaMS - G max , the third gain curve G 100 includes a section of the maximum allowable gain G max Note that this calculation can be performed even when TLaMS is measured in dBFS and G max [[ID=②④]]is measured in dB. Finally, if the measured loudness is greater than that value and less than the first target loudness TLaMS, the gain curve G 100This includes a linear section (on a logarithmic scale). If the measured loudness is within this range, the applied gain G is used to reach the first target loudness TLaMS. A It is amplified by this.

[0020] Here, it should be noted that the defined maximum speed for gain calculation is not typically the highest speed a vehicle can reach, but may be significantly lower, for example, 80 km / h or 100 km / h. If the speed is higher than this, no changes are made to the algorithm parameters. The defined maximum speed may be a speed at which road noise becomes considerably louder, but relaxed driving is still possible, allowing the driver and passengers to enjoy the sound. In some embodiments, this depends on the type of road, which may be determined, for example, by parameters obtained from a navigation system. That is, if the navigation system detects that the vehicle is on a highway or other high-speed lane, the parameters (or at least the defined maximum speed for gain calculation) may be different (for example, higher) than when the vehicle is on a small, winding road.

[0021] If the current speed matches the configured intermediate speed or threshold speed, the applied gain is the second gain curve G int This follows. In one embodiment, the intermediate speed is expressed (for example, as a fraction thereof) of the defined maximum speed. In the example shown in Figure 3, the intermediate speed is 50% of the maximum speed. Generally, it may be within any range of the maximum speed, for example, 40% to 60%. The intermediate speed is configurable, and in principle, different ranges may be appropriate.

[0022] Second gain curve G int The third gain curve G is shifted by the magnitude of TLaMS-TLaIS toward lower measured loudness. 100This is similar to the case where the measured loudness is greater than the second target loudness TLaIS ("Target Loudness at Intermediate Speed"), the gain curve G int It includes a zero-gain section. Furthermore, the measured loudness is equal to the maximum allowable gain G max The second target loudness TLaIS obtained by subtracting is TLaIS-G max If it is smaller, the second gain curve G int The maximum allowable gain G max This section includes the following. Finally, if the measured loudness is greater than that value and less than the second target loudness TLaIS, the gain curve G int This includes a linear section (on a logarithmic scale). If the measured loudness is within this range, the applied gain G is used to reach the first target loudness TLaIS. A It is amplified by [the following]. In this example, the second target loudness TLaIS is -20dBFS, the maximum allowable gain is 10dB, and therefore the maximum allowable gain G max The section is the second gain curve G int Accordingly, it is effective for measured loudness below -30 dBFS. Maximum allowable gain G max Between the section and the zero gain in the second target loudness TLaIS, the applied gain G A This corresponds to the difference between the currently measured loudness and the second target loudness TLaIS. Gain curve G in this range. int It is essentially linear (in decibels on a logarithmic scale). That is, if the measured loudness is within this range, the applied gain G will cause the input signal to reach a second target loudness TLaIS. A It is amplified by this.

[0023] The intermediate velocity or threshold velocity is a threshold between two regions in the parameter space. If it is lower than the intermediate velocity, the first type of interpolation Int1 is used, in which case a linear change in velocity results in a linear change in the gain coefficient. If it is higher than the intermediate velocity, the second type of interpolation Int2 is used, in which case a linear change in velocity results in a linear change in the target loudness. In both of these cases, the applied gain G A It brings about change.

[0024] When the speed is higher than the intermediate speed and lower than the maximum speed, and the measured loudness of the input audio signal is within a predetermined range, the second type of interpolation Int2 described above is used. When the second type of interpolation is used, the second gain curve G int and the third gain curve G 100 Although having the same shape, gain curves are obtained that are shifted between them (in the left-right direction in the drawing). In particular, the obtained gain curves are the second and third gain curves G int , G 100 Same maximum allowable gain G max This method uses a different target loudness TLaCS ("Target Loudness at a Current Speed") which is interpolated between the first target loudness TLaMS and the second target loudness TLaIS. The second type of interpolation Int2 for the target loudness TLaCS at the current speed may be performed, for example, according to the following formula:

[0025] TLaCS = TLaIS + (TLaMS - TLaIS) * ((S C -S int ) / (S max -S int )) (1)

[0026] Here, S C S int S max These represent the current speed, configured intermediate speed, and configured maximum speed, respectively. The measured loudness L of the input audio signal. M However, TLaCS and TLaCS-Gmax If it falls within the range of G, the gain applied is the difference between the measured loudness and the target loudness TLaCS at the current speed, i.e., G A =TLaCS-L M Here, TLaCS is the current velocity S related to the second type of interpolation Int2. C It depends on the following: Therefore, the level adjuster outputs an audio level corresponding to the target loudness TLaCS at the current speed. The measured loudness L of the input audio signal M TLaCS*G max If it is smaller, the applied gain G A is G max That is the case.

[0027] When the speed is lower than the intermediate speed, the first type of interpolation Int1 described above is used (at least when the measured loudness is less than the second target loudness TLaiS). When the first type of interpolation Int1 is used, the second gain curve G int A gain curve with a similar shape but scaled (vertically in the drawing) is obtained. That is, a smaller gain is applied to the input signal. In one embodiment, the gain scaling factor is the second gain curve G int This applies to the case where a linear change in velocity results in a linear change in the gain scaling coefficient. For illustrative purposes, Figure 3 shows three gain curves G for current velocities of 12.5%, 25%, and 37.5% of the defined maximum velocity. 12,5 , G 25 , G 37,5 These are shown below. The gain curve at any current speed lower than the intermediate speed is scaled according to the gain scaling factor (i.e., the maximum allowable gain G). max (and interpolated between zero), and therefore, the maximum gain G max The maximum allowable gain G, which is velocity-dependent and lies between zero and zero. max,S Use the following. An example is provided below. The first type of interpolation Int1 is performed according to the following equation, with intermediate velocity S int Lower current speed S C , that is, S C int ​This may be used to obtain the maximum gain, which is speed-dependent.

[0028] G max,S =G max *S C / S int (2)

[0029] For example, the maximum possible gain G at a speed of 25% (of a predetermined maximum speed) max,25 In Figure 3, this is 5 dB (or G max 50% of the speed, while G at 37.5% of the speed max,37,5 The maximum possible gain G is 7.5 dB. max,S This is a predetermined range of the measured loudness, where the maximum possible gain G max This is the second gain curve G int The same range as when applied by TLaIS-G max It applies in the range where it becomes smaller. Therefore, with respect to a given current velocity lower than the intermediate velocity, the gain is G max,S It will be between and 0.

[0030] In the example shown in Figure 3, the velocity is equal to the intermediate velocity, and the measured loudness L M If, for example, it is -30 dBFS, then the second gain curve G int According to the data, the target loudness is -20 dBFS, and therefore a gain of 10 dB is applied. However, for lower speeds, e.g., 12.5% ​​of the predefined or configured maximum speed, and the same measured loudness of -30 dBFS, the target loudness becomes lower due to the lower speed, i.e., only -27.5 dBFS, and therefore the gain coefficient becomes 0.25, resulting in a gain of only 2.5 dB (corresponding to the gain curve G 12,5 (Applies according to) the measured loudness L at this lower speed. M However, when it increases significantly above -30 dBFS, the corresponding applied gain is the measured loudness L M This becomes equal to the second target loudness TLaIS, i.e., L M=TLaIS, and it continuously decreases until it reaches zero.

[0031] Further information regarding how the current gain scaling coefficient and the current target loudness are used in one embodiment comprising two loudness meters is provided in the description of Figure 6 below.

[0032] The advantages include mid-range speed, maximum speed, first target loudness TLaMS, second target loudness TLaIS, and maximum allowable gain G. max Any or all of these can be used as parameters for adjusting the level adjuster.

[0033] Figure 4 shows the applicable gain coefficients for the example in Figure 3, but from a different perspective. Each drawing in Figure 4 shows an exemplary gain curve for the applicable gain coefficients at various noise levels NL or speeds, respectively. Here, it depends on the measured loudness, and the one shown in the drawing is used. In other words, the drawing shows how the gain is adapted for different noise levels NL (or speeds) with respect to a given measured loudness of the input signal. The noise level NL is between zero and 100% of a predefined maximum value. It may correspond to, for example, the speed of a vehicle, but it may also be influenced by other physical quantities. Again, the predefined maximum value of 100% does not necessarily correspond to a physical maximum value, and any higher value that may occur is treated as the predefined maximum value. The gain curve shown in Figure 4a) is used for an input signal with a measured loudness of -30 dBFS. In this case, the applicable gain is between 0% and a defined intermediate value NL. intThe gain curve shown in Figure 4b) is used for a slightly larger input signal with a measured loudness of -27 dBFS. In this case, the applied gain increases linearly with respect to noise levels between 0% and 75% until it reaches a defined maximum of 10 dB, and then remains at a defined maximum for higher noise levels. The points in Figure 4b) correspond to where the curve in Figure 3 is cut off by the vertical line at -27 dB. The gain curve shown in Figure 4c) is used for an even larger input signal with a measured loudness of -24 dBFS. Here, the applied gain increases linearly until the noise level reaches the midpoint NL int (Here, as long as it is less than 50%, it increases linearly at a lower rate; otherwise, it increases at a higher rate.) The points in Figure 4c) correspond to where the curve in Figure 3 is cut off by the vertical line at -24dB.

[0034] When the measured loudness of the input signal reaches the second target loudness TLaIS, which in this example is -20dB, the applied gain is the defined intermediate value NL, as shown in Figure 4d). int It remains zero for any smaller noise level. This is because, in Figure 3, the intermediate velocity G intThe following gain curves for any speed correspond to the fact that they are zero at the second target loudness TLaIS. The applied gain increases only at higher noise levels. Similarly, Figure 4e) shows the applied gain to an input signal with a measured loudness of -17 dBFS, i.e., a measured loudness between the first target loudness TLaMS and the second target loudness TLaIS. In this case, at high noise levels, in this example, the input signal is amplified only if it is above 75% of the maximum defined noise level. Similarly, for input signals greater than -17 dBFS, the applied gain coefficient remains zero until the noise level increases, and then increases linearly. When the measured loudness reaches the first target loudness TLaMS, the applied gain remains zero, independently of the noise level, as shown in Figure 4f).

[0035] The following provides further details on how loudness meters 131 and 132 can measure the loudness of the input signal. Note that the loudness meters do not modify the input audio signal in any way. In the embodiment shown in Figure 5, each channel y of the input signal L ,y R The power is determined by mean square calculation, 521,522 and then summed up, similar to known methods for measuring loudness related to the ITU_BS1770-3 or EBU R128 broadcast standards. Optionally, K-weighting using a K-filter 511,512 may be applied before the mean square calculation. To obtain the loudness value, the sum of powers is converted to logarithmic form 540.

[0036] Conventional loudness measurements according to the ITU_BS1770-3 or EBU R128 standards use gate control, which includes a first gate control stage for excluding silence and a second gate control stage for prioritizing foreground sound over background sound. Conventionally, overlapping gate control blocks with a length of 400 milliseconds are used to determine which portion of the incoming audio signal contributes to the reported loudness measurement.

[0037] The present invention differs from conventional methods in that, in one embodiment, it does not use at least these gate control blocks or stages. As a result, foreground and background sounds are given equal priority. The advantages of omitting gate control are demonstrated, for example, with respect to music that includes both direct sound and diffuse reflection. For example, a recording of a string quartet in a concert hall includes both direct sound and diffuse reflection. Diffuse reflections tend to be quieter, but they are important for creating an irresistible sense of space and immersion for listeners, for example, in a car. When using conventional methods for measuring loudness, diffuse reflections may be excluded from the measurement. Since only audio blocks with greater loudness are considered, conventional measurements may result in overly high measured loudness. As a result, any compensation gain will also be slightly lower, and these diffuse elements may be lost below the noise floor. By using the improved solution described above, these quieter elements are included in the loudness measurement, resulting in a lower measured loudness and, therefore, a higher compensation gain.

[0038] In the present invention, silence is still excluded, but it is excluded at the audio block level rather than at the gate control block level. A given audio block is allowed to contribute to loudness measurement only if it does not contain silence, i.e., if its loudness is greater than a predefined threshold called the "Loudness of Silence:LoS". This threshold may be set, for example, to -70dBFS. The audio block may contain, for example, 512 samples. Generally, the block size is not important. If the loudness of an audio block is greater than LoS, this loudness is pushed into a queue that aids in tracking the loudness over time, and the old loudness values are popped from the queue and discarded. If the loudness of this block is not greater than LoS, it is immediately discarded and the queue remains in its previous state. This means that previous loudness measurements are used until the signal becomes greater than LoS again. Thus, the loudness measurement is based on a certain number of audio blocks each having a measured loudness between LoS and 0dBFS. <​​​Each time an audio block is processed, the Long Term Measurer (LTM) 630 reports the long-term loudness (LTL) over the previous long-term measurement interval, and the Short Term Measurer (STM) 620 reports the short-term loudness (STL) over the previous short-term measurement interval. The gain curve determination module 610 determines the current gain curve according to the current speed, as described above. Audio input signal S in These are provided to different loudness measurement blocks 620 and 630, where the long-term loudness (LTL) and short-term loudness (STL) are determined. The LTL and STL, along with the determined current gain curve, are provided to the gain determination module 640.

[0040] The Long Term Requested Gain (LTRG) first finds the difference between the target loudness and the long-term loudness (LTL), and then calculates this difference as the maximum allowable gain (G). max It is calculated by clamping to [a specific value]. Similarly, the Short-Term Requested Gain (STRG) is calculated by finding the difference between the target loudness and the short-term loudness (STL) and using this difference as the maximum allowable gain (G). max It is calculated by clamping to . The same target loudness is used in both calculations, which is obtained according to the noise level or speed of the vehicle, as previously mentioned with respect to Figure 3, for example. The requested gains LTRG and STRG may then be compared against the gain calculation module 650 which calculates the applied gain. Here, a predefined (i.e., configurable) Long Term Influence Extension (LTIE) may be used, which is also given in decibels dB. The LTIE is used to calculate the applied gain G even if STRG is smaller than LTRG. AEnables the LTL to play a predetermined role when determining. This can occur, for example, when the music is generally quiet (thus having a small LTL and a large LTRG), but contains a louder sound in a short period than in the remaining period (resulting in a larger STL and a smaller STRG). The applied gain G A may be obtained as follows by the gain calculation module 650.

[0041] (When LTRG < STRG), (G A = LTRG is applied so that). (When STRG < LTRG), two cases can occur: (When STRG < LTRG - LTIE), (G A = STRG is applied so that). (When LTRG - LTIE < STRG < LTRG), interpolation is performed for G A .

[0042] The interpolation may provide a gain level between STRG and LTRG, for example, according to G A = LTRG - (LTRG - STRG) 2 / LTIE. When using this interpolation, as the difference between LTRG and STRG becomes larger, the applied gain approaches STRG (i.e., the applied gain becomes smaller), while as the difference becomes smaller, the applied gain approaches LTRG (i.e., the applied gain becomes larger). The interpolation in the case of LTRG - LTIE < STRG < LTRG has no discontinuity. When the difference LTRG - STRG approaches zero, the applied gain G A approaches LTRG, and when the difference approaches LTIE, the applied gain G AThis approaches STRG. LTIE can be used to find a good balance between reducing pumpiness and reducing overshoot. High LTIE values ​​(e.g., above 6dB) reduce the pumpiness of the music, although they risk introducing overshoot, while low or zero LTIE values ​​reduce overshoot but may result in some pumpiness.

[0043] In all cases, the G generated from the above comparison A As a final step, the noise level is scaled by a gain scaling coefficient according to the vehicle's noise level or speed, as described above with respect to Figure 3, for example.

[0044] For example, the long-term measurement interval may be in the range of 10 to 20 seconds, and the short-term measurement interval may be in the range of 1 to 4 seconds. If the loudness measured over any of the time intervals falls below a predefined silence threshold ST, the applied gain is fixed, that is, it retains the previous value (as shown in the leftmost part of Figure 3), and can only be changed once both loudness values ​​have again exceeded the silence threshold. Note that the silence threshold ST configured in the level adjuster is separate from the silence loudness LoS mentioned above. However, the silence threshold cannot be set lower than the silence loudness.

[0045] In an alternative embodiment, the algorithm may be configured to ignore short-term loudness (STL) measurements, and loudness is measured only over a configured long-term interval. In this case, short-term loudness (STL) and long-term impact extension (LTIE) are not used, and the silence threshold is compared only against long-term loudness (LTL).

[0046] Although various different embodiments have been described, it is clear that combinations of features from different embodiments are possible, even if not explicitly mentioned in this application. Such combinations are considered to be within the scope of the present invention.

Claims

1. A method for automatically controlling the loudness of an audio signal played in a vehicle, wherein the above method is - The automatic loudness meter measures the loudness of incoming audio signals, including sequences of digital audio blocks, - Receiving a noise level display signal indicating the noise level inside the above vehicle, - The processor includes calculating a gain coefficient based on the measured loudness of the incoming audio signal and the noise level indicator signal. If the measured loudness is greater than or equal to a predefined first target loudness (TLamS), the gain coefficient is zero. Otherwise, if the noise level indicator signal indicates a noise level greater than or equal to a predetermined intermediate noise level, the gain coefficient is the difference between the measured loudness and the current target loudness, and is limited to the maximum gain, the maximum gain being a defined absolute maximum gain, and the current target loudness being linearly interpolated according to the noise level indicator signal between a predetermined second target loudness (TLaiS) for a noise level equal to the intermediate noise level and a predetermined first target loudness (TLamS) for a noise level equal to the predetermined maximum noise level. Otherwise, if the above noise level indicator signal shows a noise level smaller than the above intermediate noise level, the above gain coefficient is linearly interpolated between the gain defined with respect to the above intermediate noise level and zero, according to the above noise level. The above method, - Including applying the above gain coefficient to the audio signal to be reproduced, method.

2. The information indicating the above noise level is obtained from the signal indicating the speed of the above vehicle. The method according to claim 1.

3. The noise level information mentioned above is obtained from one or more microphones inside the vehicle. The method according to claim 1.

4. The loudness measurement described above is based on a certain number of audio blocks that have a loudness greater than the silence threshold. Audio blocks with a loudness lower than the above silence threshold will not change the internal state of the above automatic loudness meter. The automatic loudness meter will retain its latest loudness measurement until the loudness of the incoming audio block exceeds the silence threshold. The method according to claim 1.

5. The calculated gain coefficient above is the first gain coefficient. The first gain coefficient is added to the second gain coefficient, and both the first and second gain coefficients determine the loudness of the audio signal being reproduced. The method according to claim 1.

6. The above gain coefficient is independent of the user-controllable volume. The method according to claim 1.

7. The above gain coefficient is linked to a user-controllable volume. The method according to claim 1.

8. A method for automatically controlling the loudness of an audio signal played in a vehicle, wherein the above method is - An automated loudness meter measures the loudness of an incoming audio signal, including a sequence of digital audio blocks. A digital audio block with a loudness level lower than the silence threshold is detected and skipped in the loudness measurement described above. The above measurement is based on a digital audio block having a loudness greater than the silence threshold mentioned above. The above method, - Receiving a speed indicator signal that shows the speed at which the above vehicle is moving, - The processor includes calculating a gain coefficient based on the measured loudness and the speed display signal. If the measured loudness is greater than or equal to a predefined first target loudness (TLamS), the gain coefficient is zero. Otherwise, if the speed indicator signal indicates a speed greater than or equal to a defined intermediate speed, the gain coefficient is the difference between the measured loudness and the current target loudness, and is limited to the maximum gain, the maximum gain being the defined absolute maximum gain, and the current target loudness being linearly interpolated according to the speed indicator signal between a predefined second target loudness (TLaiS) for a speed equal to the intermediate speed and a predefined first target loudness (TLamS) for a speed equal to the predefined maximum speed. Otherwise, the maximum gain is linearly interpolated between the absolute maximum gain and zero according to the velocity to obtain the current maximum gain, and the gain coefficient is equal to the current maximum gain, or obtained by linearly interpolating between the current maximum gain and zero according to the measured loudness, if the measured loudness is less than the second target loudness (TLaiS) obtained by subtracting the absolute maximum gain. The above method, - Including applying the above gain coefficient to the audio signal to be reproduced, method.

9. The loudness measurement described above is based on a certain number of audio blocks having a loudness greater than the silence threshold described above. The method according to claim 8.

10. The calculated gain coefficient above is the first gain coefficient. The first gain coefficient is added to the second gain coefficient, and both the first and second gain coefficients determine the loudness of the audio signal being reproduced. The method according to claim 8.

11. The above gain coefficient is independent of the user-controllable volume. The method according to claim 8.

12. The above gain coefficient is linked to a user-controllable volume. The method according to claim 8.

13. A device for automatically controlling the loudness of audio signals played in a vehicle, wherein the device is - An automated loudness meter adapted to measure the loudness of incoming audio signals, including sequences of digital audio blocks, - An input circuit adapted to receive a noise level indicator signal showing the noise level inside the above vehicle, - comprising at least one processor adapted to calculate a gain coefficient based on the measured loudness and the noise level display signal, If the measured loudness is greater than or equal to a predefined first target loudness (TLamS), the gain coefficient is zero. Otherwise, if the noise level indicator signal indicates a noise level greater than or equal to a predetermined intermediate noise level, the gain coefficient is the difference between the measured loudness and the current target loudness, and is limited to the maximum gain, the maximum gain being a defined absolute maximum gain, and the current target loudness being linearly interpolated according to the noise level indicator signal between a predetermined second target loudness (TLaiS) for a noise level equal to the intermediate noise level and a predetermined first target loudness (TLamS) for a noise level equal to the predetermined maximum noise level. Otherwise, if the above noise level indicator signal shows a noise level smaller than the above intermediate noise level, the above gain coefficient is linearly interpolated between the gain defined with respect to the above intermediate noise level and zero, according to the above noise level. The above device is - The amplifier is adapted to apply the above-mentioned gain coefficient to the above-mentioned audio signal to be reproduced. Device.

14. The information indicating the above noise level is obtained from the signal indicating the speed of the above vehicle. The apparatus according to claim 13.

15. The noise level information mentioned above is obtained from one or more microphones inside the vehicle. The apparatus according to claim 13.

16. The loudness measurement described above is based on a certain number of audio blocks that have a loudness greater than the silence threshold. Audio blocks with a loudness lower than the above silence threshold will not change the internal state of the above automatic loudness meter. The automatic loudness meter will retain its latest loudness measurement until the loudness of the incoming audio block exceeds the silence threshold. The apparatus according to claim 13.

17. The calculated gain coefficient above is the first gain coefficient. The first gain coefficient is added to the second gain coefficient, and both the first and second gain coefficients determine the loudness of the audio signal being reproduced. The apparatus according to claim 13.

18. The above gain coefficient is independent of the user-controllable volume. The apparatus according to claim 13.

19. The above gain coefficient is linked to a user-controllable volume. The apparatus according to claim 13.

20. A device for automatically controlling the loudness of audio signals played in a vehicle, wherein the device is - Equipped with an automated loudness meter adapted to measure the loudness of incoming audio signals, including sequences of digital audio blocks, A digital audio block with a loudness level lower than the silence threshold is detected and skipped in the loudness measurement described above. The above measurement is based on a digital audio block having a loudness greater than the silence threshold mentioned above. The above device is - An input circuit adapted to receive a speed indicator signal showing the speed at which the above vehicle is moving, - comprising at least one processor adapted to calculate a gain coefficient based on the measured loudness and the speed display signal, If the measured loudness is greater than or equal to a predefined first target loudness (TLamS), the gain coefficient is zero. Otherwise, if the speed indicator signal indicates a speed greater than or equal to a defined intermediate speed, the gain coefficient is the difference between the measured loudness and the current target loudness, and is limited to the maximum gain, the maximum gain being the defined absolute maximum gain, and the current target loudness being linearly interpolated according to the speed indicator signal between a predefined second target loudness (TLaiS) for a speed equal to the intermediate speed and a predefined first target loudness (TLamS) for a speed equal to the predefined maximum speed. Otherwise, if the speed indicator signal indicates a speed lower than the intermediate speed, the maximum gain is linearly interpolated between the absolute maximum gain and zero according to the speed to obtain the current maximum gain, and the gain coefficient is obtained by linearly interpolating between the current maximum gain and zero according to the measured loudness, if the measured loudness is less than the second target loudness (TLaiS) obtained by subtracting the absolute maximum gain, and the measured loudness. The above device is - The amplifier is adapted to apply the above-mentioned gain coefficient to the above-mentioned audio signal to be reproduced. Device.

21. The loudness measurement described above is based on a certain number of audio blocks having a loudness greater than the silence threshold described above. The apparatus according to claim 20.

22. The calculated gain coefficient above is the first gain coefficient. The first gain coefficient is added to the second gain coefficient, and both the first and second gain coefficients determine the loudness of the audio signal being reproduced. The apparatus according to claim 20.

23. The above gain coefficient is independent of the user-controllable volume. The apparatus according to claim 20.

24. The above gain coefficient is linked to a user-controllable volume. The apparatus according to claim 20.

25. A computer-readable, non-temporary storage medium that stores computer-executable instructions that, when executed on a computer, cause the computer to perform the method described in claim 1.

26. A computer-readable, non-temporary storage medium that stores computer-executable instructions, which, when executed on a computer, cause the computer to perform the method described in claim 8.

Images