Audio codec post-filter

• US 7,707,034 B2
• Filed: 05/31/2005
• Issued: 04/27/2010
• Est. Priority Date: 05/31/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Sun, Xiaoqin, Khalil, Hosam A., Chen, Wei-Ge, Koishida, Kazuhito, Wang, Tian
• Primary Examiner(s): ABEBE, DANIEL DEMELASH
• Application Number:
  US11/142,603
• Publication Number:
  US 20060271354A1
• Time in Patent Office: 1,792 Days
• Field of Search: 704/228, 704/500, 704/258, 704/262, 381/312, 381/320
• US Class Current: 704/262
• CPC Class Codes: G10L 19/26 Pre-filtering or post-filtering

Abstract

Techniques and tools are described for processing reconstructed audio signals. For example, a reconstructed audio signal is filtered in the time domain using filter coefficients that are calculated, at least in part, in the frequency domain. As another example, producing a set of filter coefficients for filtering a reconstructed audio signal includes clipping one or more peaks of a set of coefficient values. As yet another example, for a sub-band codec, in a frequency region near an intersection between two sub-bands, a reconstructed composite signal is enhanced.

Claims

1. A computer-implemented method comprising:

calculating a set of filter coefficients for application to a reconstructed audio signal, wherein the calculating the set of filter coefficients comprises;

performing a transform of a set of initial time domain values from a time domain into a frequency domain, thereby producing a set of initial frequency domain values;

performing one or more frequency domain calculations using the initial frequency domain values to produce a set of processed frequency domain values; and

performing a transform of the processed frequency domain values from the frequency domain into the time domain, thereby producing a set of processed time domain values; and

producing a filtered audio signal by filtering at least a portion of the reconstructed audio signal in a time domain using the set of filter coefficients; and

wherein performing one or more frequency domain calculations using the initial frequency domain values to produce a set of processed frequency domain values comprises clipping frequency domain values in the frequency domain such that only those frequency domain values which exceed a maximum clip value are clipped.

2. The method of claim 1, wherein the filtered audio signal represents a frequency sub-band of the reconstructed audio signal.

3. The method of claim 1, wherein calculating the set of filter coefficients further comprises:

before the transform of the initial time domain values, padding the initial time domain values up to a length for the transform of the initial time domain values; and

after the transform of the processed frequency domain values, truncating the set of processed time domain values in the time domain.

4. The method of claim 1, wherein the set of initial time domain values comprises a set of linear prediction coefficients.

5. The method of claim 4, wherein clipping the frequency domain values in the frequency domain comprises capping a spectrum derived from the set of liner prediction coefficients at a maximum value.

6. The method of claim 4, wherein performing the one or more frequency domain calculations comprises reducing a range of a spectrum derived from the set of linear prediction coefficients.

7. The method of claim 6, wherein reducing a range of a spectrum derived from the set of linear prediction coefficients comprises normalizing values in the spectrum.

8. The method of claim 7, wherein the linear prediction coefficients are for a multi-band codec and the normalizing values in the spectrum comprises normalizing values within a single band.

9. The method of claim 8, wherein the linear prediction coefficients are for a full band codec and the normalizing values in the spectrum comprises normalizing values for the full band.

10. The method of claim 6, wherein reducing a range of a spectrum derived from the set of linear prediction coefficients comprises performing nonlinear compression on values in the spectrum.

11. The method of claim 1, wherein the one or more frequency domain calculations comprises one or more calculations in a logarithmic domain.

12. The method of claim 1, wherein

the filtered audio signal comprises plural reconstructed frequency sub-band signals, the plural reconstructed frequency sub-band signals including a reconstructed first frequency sub- band signal for a first frequency band and a reconstructed second frequency sub-band signal for a second frequency band;

and

the method further comprises selectively enhancing the reconstructed composite signal at a frequency region around an intersection between the first frequency band and the second frequency band, wherein enhancing the reconstructed composite signal comprises passing the reconstructed composite signal through a band pass filter, wherein a pass band of the band pass filter corresponds to the frequency region around the intersection between the first frequency band and the second frequency band.

13. A method comprising:

producing a set of filter coefficients for application to a reconstructed audio signal, including processing a set of values in a frequency domain representing one or more peaks and one or more valleys, wherein the processing the set of values in the frequency domain comprises clipping one or more of the peaks or valleys, and wherein the clipping includes capping the set of values in the frequency domain at a maximum clip value by setting values which exceed the maximum clip value to the clip value and maintaining the values which do not exceed the maximum clip value; and

filtering at least a portion of the reconstructed audio signal using the filter coefficients.

14. The method of claim 13, wherein producing a set of filter coefficients further comprises calculating the maximum clip value as a function of an average of the set of values in the frequency domain.

15. The method of claim 13, wherein the set of values in the frequency domain is based at least in part on a set of linear prediction coefficient values.

16. The method of claim 13, wherein the clipping is performed in the frequency domain.

17. The method of claim 13, wherein the filtering is performed in a time domain.

18. The method of claim 13, further comprising reducing a range of the set of values in the frequency domain before the clipping.

19. The method of claim 18, wherein reducing a range of the set of values in the frequency domain before the clipping comprises normalizing the values in the frequency domain.

20. The method of claim 18, wherein reducing a range of the set of values in the frequency domain before the clipping comprises performing nonlinear compression on values in the frequency domain.

21. The method of claim 13, wherein capping the set of values in the frequency domain at a maximum clip value comprises performing one or more calculations in a logarithmic domain.

22. The method of claim 13, further comprising:

receiving a reconstructed composite signal synthesized from plural reconstructed frequency sub-band signals, the plural reconstructed frequency sub-band signals including a reconstructed first frequency sub-band signal for a first frequency band and a reconstructed second frequency sub-band signal for a second frequency band; and

selectively enhancing the reconstructed composite signal at a frequency region around an intersection between the first frequency band and the second frequency band, wherein the enhancing comprises increasing signal energy in the frequency region.

23. A computer-implemented method comprising:

receiving a reconstructed composite signal synthesized from plural reconstructed frequency sub-band signals, the plural reconstructed frequency sub-band signals including a reconstructed first frequency sub-band signal for a first frequency band and a reconstructed second frequency sub-band signal for a second frequency band; and

selectively enhancing the reconstructed composite signal at a frequency region around an intersection between the first frequency band and the second frequency band, wherein enhancing the reconstructed composite signal comprises passing the reconstructed composite signal through a baud pass filter, wherein a pass band of the band pass filter corresponds to the frequency region around the intersection between the first frequency band and the second frequency band.

24. The method of claim 23 further comprising:

decoding coded information to produce the plural reconstructed frequency sub-band signals; and

synthesizing the plural reconstructed frequency sub-band signals to produce the reconstructed composite signal.

25. The method of claim 23, wherein the band pass filter comprises a low pass filter in series with a high pass filter.

26. The method of claim 23, wherein the band pass filter has unity gain at one or more stop bands and greater than unity gain at the pass band.

27. The method of claim 23, wherein the enhancing further comprises increasing signal energy in the frequency region.

28. A method comprising:

producing a set of filter coefficients for application to a reconstructed audio signal, including processing a set of coefficient values representing one or more peaks and one or more valleys, wherein the processing the set of coefficient values comprises clipping one or more of the peaks or valleys such that only those coefficient values which exceed a maximum clip value are clipped, and wherein the set of coefficient values is based at least in part on a set of linear prediction coefficient values; and

filtering at least a portion of the reconstructed audio signal using the filter coefficients.

29. A method comprising:

producing a set of filter coefficients for application to a reconstructed audio signal, including processing a set of coefficient values representing one or more peaks and one or more valleys, wherein the processing the set of coefficient values comprises clipping one or more of the peaks or valleys such that only those coefficient values which exceed a maximum clip value are clipped, and wherein the clipping is performed in a frequency domain; and

filtering at least a portion of the reconstructed audio signal using the filter coefficients.

30. A method comprising:

producing a set of filter coefficients for application to a reconstructed audio signal, including processing a set of coefficient values representing one or more peaks and one or more valleys, wherein the processing the set of coefficient values comprises clipping one or more of the peaks or valleys such that only those coefficient values which exceed a maximum clip value are clipped; and

filtering at least a portion of the reconstructed audio signal using the filter coefficients, wherein the filtering is performed in a time domain.

31. A method comprising:

producing a set of filter coefficients for application to a reconstructed audio signal, including processing a set of coefficient values representing one or more peaks and one or more valleys, wherein the processing the set of coefficient values comprises;

reducing a range of the set of coefficient values; and

clipping one or more of the peaks or valleys such that only those coefficient values which exceed a maximum clip value are clipped; and

filtering at least a portion of the reconstructed audio signal using the filter coefficients.