Method, apparatus and computer program for calculating and adjusting the perceived loudness of an audio signal

• US 20070092089A1
• Filed: 05/27/2004
• Published: 04/26/2007
• Est. Priority Date: 05/28/2003
• Status: Active Grant

• Current Assignee: Dolby Laboratories Licensing Corporation (Dolby Laboratories Incorporated)
• Original Assignee: Dolby Laboratories Licensing Corporation (Dolby Laboratories Incorporated)
• Inventors: Seefeldt, Alan, Smithers, Michael, Crockett, Brett
• Granted Patent: US 8,437,482 B2
• Time in Patent Office: Days
• US Class Current: 381/104
• CPC Class Codes: G10L 25/48 specially adapted for particular use
• CPC Class Codes: H03G 5/00 Tone control or bandwidth control in amplifiers
• CPC Class Codes: H03G 5/005 of digital signals
• CPC Class Codes: H03G 9/005 of digital or coded signals
• CPC Class Codes: H03G 9/025 frequency-dependent volume compression or expansion, e.g. multiple-band systems H03G9/10, H03G9/18 take precedence

Abstract

One or a combination of two or more specific loudness model functions selected from a group of two or more of such functions are employed in calculating the perceptual loudness of an audio signal. The function or functions may be selected, for example, by a measure of the degree to which the audio signal is narrowband or wideband. Alternatively or with such a selection from a group of functions, a gain value G[t] is calculated, which gain, when applied to the audio signal, results in a perceived loudness substantially the same as a reference loudness. The gain calculating employs an iterative processing loop that includes the perceptual loudness calculation.

Claims

1. A method for processing an audio signal, comprising

calculating the perceptual loudness of the audio signal in response to the audio signal, wherein said calculating selects, from a group of two or more specific loudness model functions, one or a combination of two or more of the specific loudness model functions, the selection of which is controlled by a measure of characteristics of the input audio signal.

2. A method according to claim 1 wherein the measure of characteristics of the audio signal is a measure of the degree to which the input signal is narrowband or wideband.

3. A method according to claim 2 further comprising calculating the degree to which the input signal is narrowband or wideband by calculating the spectral flatness of the input signal.

4. A method according to claim 1 wherein said calculating selects from or combines two specific loudness model functions, a first loudness model function being selected by a measure of characteristics resulting from a narrowband input signal, a second loudness model function being selected by a measure of characteristics resulting from a wideband input signal, and a combination of the first and second loudness model functions being selected by a measure of characteristics resulting from a partly narrowband, partly wideband input signal.

5. A method according to claim 4 wherein said calculating includes producing an excitation signal in response to said input signal and wherein both the first and second loudness model functions increase monotonically above a threshold in quiet with increasing excitation according to a power law, the first loudness model function increasing faster than the second loudness model function.

6. A method according to claim 1 wherein said calculating includes producing an excitation signal in response to said input signal and wherein said calculating selects from a group of two or more specific loudness models, one or a combination of two or more of said specific loudness models in each of respective frequency bands of the excitation signal.

7. A method according to claim 1 wherein said calculating includes producing an excitation signal in response to said input signal and wherein said calculating selects from a group of two or more specific loudness models, one or a combination of two or more of said specific loudness models in a group of respective frequency bands of the excitation signal.

8. A method according to claim 7 wherein the group of respective frequency bands are all of the frequency bands of the excitation signal.

9. A method according to claim 1 wherein said calculating includes producing an excitation signal in response to said input signal and wherein the measure of characteristics of the audio signal is derived from the excitation signal.

10. A method according to claim 1 wherein said calculating includes producing an excitation signal in response to said input signal and wherein the calculating includes calculating a specific loudness in each of respective frequency bands of the excitation signal.

11. A method according to claim 10 wherein the calculating further comprises selecting the specific loudness of a frequency band to provide the perceptual loudness or combining the specific loudness of a group of frequency bands to provide the perceptual loudness.

36. Apparatus adapted to perform the methods of any one of claims 1, 12 and 35.

37. A computer program, stored on a computer-readable medium for causing a computer to perform the methods of any one of claims 1, 12 and 35.

12. A method for processing an audio signal, comprising

producing, in response to the audio signal, an excitation signal, and

calculating, in response at least to the excitation signal, a gain value G[t], which, if applied to the audio signal, would result in a perceived loudness substantially the same as a reference loudness, the calculating including an iterative processing loop that includes at least one non-linear process.

13. The method of claim 12 wherein the iterative processing loop includes calculating a perceptual loudness.

14. The method of claim 12 wherein said calculating is also in response to a measure of characteristics of the audio signal.

15. The method of claim 14 wherein said at least one non-linear process includes a specific loudness calculation that selects, from a group of two or more specific loudness model functions, one or a combination of two or more of said specific loudness model functions, the selection of which is controlled by the measure of characteristics of the input audio signal.

16. The method of claim 12 wherein the excitation signal is time smoothed and/or the method further comprises time smoothing the gain value G[t].

17. The method of claim 16 wherein the excitation signal is linearly time smoothed.

18. The method of claim 16 wherein the method further comprises smoothing the gain value G[t], said smoothing employing a histogram technique.

19. The method of claim 12 wherein the iterative processing loop includes time smoothing.

20. A method according to claim 12 or claim 15, wherein the iterative processing loop includes

adjusting the magnitude of the excitation signal in response to a function of an iteration gain value G_i such that the adjusted magnitude of the excitation signal increases with increasing values of G_i and decreases with decreasing values of G_i,

calculating a perceptual loudness in response to the magnitude-adjusted excitation signal,

comparing the calculated perceptual loudness of the audio signal to a reference perceptual loudness to generate a difference, and

adjusting the gain value G_i in response to the difference so as to reduce the difference between the calculated perceptual loudness and the reference perceptual loudness.

21. A method according to claim 20, wherein the iterative processing loop, in accordance with a minimization algorithm, repetitively adjusts the magnitude of the excitation signal, calculates a perceptual loudness, compares the calculated perceptual loudness to a reference perceptual loudness, and adjusts the gain value G_i to a final value G[t].

22. A method according to claim 21, wherein the minimization algorithm is in accordance with the gradient descent method of minimization.

23. A method according to claim 12, further comprising

controlling the amplitude of the input audio signal with the gain G[t] so that the resulting perceived loudness of the input audio signal is substantially the same as the reference loudness.

24. A method according to claim 23 wherein the reference loudness is set by a user.

25. A method according to claim 12 wherein the reference loudness is a perceptual loudness calculated by a process according to claim 13.

26. A method according to claim 12 wherein producing, in response to the audio signal, an excitation signal, comprises

linearly filtering the audio signal by a function or functions that simulate the characteristics of the outer and middle human ear to produce a linearly-filtered audio signal, and

dividing the linearly-filtered audio signal into frequency bands that simulate the excitation pattern generated along the basilar membrane of the inner ear to produce the excitation signal.

27. A method according to claim 12 wherein said at least one non-linear process includes calculating the specific loudness in each frequency band of the excitation signal.

28. A method according to claim 27 wherein said calculating the specific loudness in each frequency band of the excitation signal selects from a group of two or more specific loudness model functions, one or a combination of two or more of said specific loudness model functions, the selection of which is controlled by the measure of characteristics of the input audio signal.

29. A method according to claim 20 wherein calculating a perceptual loudness in response to the magnitude-adjusted excitation signal includes calculating the specific loudness in respective frequency bands of the excitation signal.

30. A method according to claim 29 wherein said calculating the specific loudness in each frequency band of the excitation signal selects, from a group of two or more specific loudness model functions, one or a combination of two or more of the specific loudness model functions, the selection of which is controlled by the measure of characteristics of the input audio signal.

31. A method according to claim 30 wherein calculating a perceptual loudness in response to the magnitude-adjusted excitation signal further comprises

combining the specific loudness for each frequency band into a measure of perceptual loudness.

32. A method according to claim 20 wherein the reference perceptual loudness is derived from a measure of the calculated perceptual loudness.

33. A method according to claim 32 wherein the reference perceptual loudness is a scaled version of the calculated perceptual loudness.

34. A method according to claim 32 wherein the reference perceptual loudness is greater than the calculated perceptual loudness when the calculated perceptual loudness is below a threshold and less than the calculated perceptual loudness when the calculated perceptual loudness is above a threshold.

35. A method for processing a plurality of audio signals, comprising

a plurality of processes, each receiving a respective one of the audio signals, wherein each process

produces, in response to the respective audio signal, an excitation signal,

calculates, in response at least to the excitation signal, a gain value G[t], which, if applied to the audio signal, would result in a perceived loudness substantially the same as a reference loudness, the calculating including an iterative processing loop that includes at least one non-linear process, and

controls the amplitude of the respective audio signal with the gain G[t] so that the resulting perceived loudness of the respective audio signal is substantially the same as the reference loudness, and

applying the same reference loudness to each of the plurality of processes.