Multichannel audio extension

• US 7,620,554 B2
• Filed: 05/26/2005
• Issued: 11/17/2009
• Est. Priority Date: 05/28/2004
• Status: Expired due to Fees

• Current Assignee: Nokia Corporation
• Original Assignee: Nokia Corporation
• Inventors: Ojanpera, Juha
• Primary Examiner(s): Sked; Matthew J
• Application Number:
  US11/138,711
• Publication Number:
  US 20050267763A1
• Time in Patent Office: 1,636 Days
• Field of Search: 704/229
• US Class Current: 704/500
• CPC Class Codes: G10L 19/008 Multichannel audio signal coding or decoding using interchannel correlation to reduce redundancy, e.g. joint-stereo, intensity-coding or matrixing
• CPC Class Codes: G10L 19/0204 using subband decomposition

Abstract

A method is shown for supporting a multichannel audio extension at an encoding end of a multichannel audio coding system. In order to improve the audio quality over a large frequency range, the method comprises transforming each channel of a multichannel audio signal into the frequency domain and dividing a bandwidth of the frequency domain signals into a first region of lower frequencies and at least one further region of higher frequencies. Then, the frequency domain signals are encoded in each of the frequency regions with another type of coding to obtain parametric multichannel extension information for the respective frequency region. The invention relates equally to a method for supporting in a corresponding manner a multichannel audio extension at a decoding end. Also shown are a corresponding encoder, a corresponding decoder, and corresponding devices, systems and software program products.

Claims

1. Method comprising:

generating from a multichannel audio signal an encoded mono audio signal in a first processing chain; and

generating from said multichannel audio signal encoded parametric multichannel extension information in a second processing chain distinct from said first processing chain, said generating of encoded parametric multichannel extension information comprising;

transforming each channel of said multichannel audio signal into the frequency domain;

dividing a bandwidth of said frequency domain channel signals into a first region of lower frequencies and at least one further region of higher frequencies; and

encoding said frequency domain channel signals in each of said frequency regions with another type of coding to obtain a parametric multichannel extension information for the respective frequency region, wherein encoding said frequency domain signals in said first region comprises combining corresponding samples of all channels in said first region, quantizing said combined samples and encoding said quantized samples, and wherein encoding said quantized samples comprises dividing said quantized samples into subblocks and encoding each subblock separately.

2. Method according to claim 1, wherein encoding said quantized samples comprises applying a plurality of coding schemes to said quantized samples and selecting a coding scheme which results in a lowest number of bits for said parametric multichannel extension information.

3. Method according to claim 2, wherein said plurality of coding schemes comprise a plurality of Huffman coding schemes.

4. Method according to claim 1, wherein, in case encoding said quantized samples results in more bits for said parametric multichannel extension information than are available for said first region, said quantization comprises modifying said quantized samples to obtain quantized samples which result in said encoding of quantized samples at the most in the number of bits for said parametric multichannel extension information that are available for said first region.

5. Method according to claim 1, wherein said quantization employs a selectable quantization gain for quantizing combined samples of a respective frame, said quantization comprising selecting a quantization gain for a respective frame which avoids sudden changes in the quantization gain from one frame to the next.

6. Method according to claim 1, wherein in case encoding said quantized samples results in a number of bits for said parametric multichannel extension information which is lower than a number of bits which are available for said first region, said method further comprising generating refinement bits representing information which allows to compensate for quantization errors.

7. Method according to claim 1, wherein said at least one further region comprises a middle frequency region and a high frequency region.

8. Method according to claim 7, wherein said type of coding employed for encoding said frequency domain signals in said middle frequency region comprises:

determining for each of a plurality of adjacent frequency bands within said middle frequency region whether a spectral first channel signal of said multichannel signal, a spectral second channel signal of said multichannel signal or none of said spectral channel signals is dominant in the respective frequency band; and

encoding a corresponding state information for each of said frequency bands as a parametric multichannel extension information.

9. Method according to claim 8, further comprising post-processing said determined state information such that short-time changes in said state information are avoided before encoding said state information.

10. Method according to claim 7, wherein said type of coding employed for encoding said frequency domain signals in said high frequency region comprises:

determining for each of a plurality of adjacent frequency bands within said high frequency region whether a spectral first channel signal of said multichannel signal, a spectral second channel signal of said multichannel signal or none of said spectral channel signals is dominant in the respective frequency band; and

selecting a first approach or a second approach for encoding a corresponding state information for each of said frequency bands as a parametric multichannel extension information, wherein said first approach includes encoding a corresponding state information for each of said frequency bands, and wherein said second approach includes comparing said state information for a current frame to state information for a previous frame, encoding a result of this comparison and encoding state information for a current frame only in case there was a change in said state information from said previous frame to said current frame.

11. Method according to claim 10, further comprising post-processing said determined state information such that short-time changes in said state information are avoided before encoding said state information.

12. Method comprising:

decoding an encoded mono signal;

decoding an encoded parametric multichannel extension information which is provided separately for a first region of lower frequencies and for at least one further region of higher frequencies using different types of coding, wherein said encoded parametric multichannel extension information comprises for said first region encoded subblocks, said encoded subblocks having been obtained at an extension encoder by combining corresponding samples of all channels in said first region, quantizing said combined samples, dividing said quantized samples into subblocks and encoding each subblock separately;

reconstructing a multichannel signal based on said decoded mono signal and on said decoded parametric multichannel extension information separately for said first region and said at least one further region;

combining said reconstructed multichannel signals in said first and said at least one further region; and

transforming each channel of said combined multichannel signal into the time domain.

13. Apparatus comprising:

an encoder configured to generate from a multichannel audio signal an encoded mono audio signal in a first processing chain; and

an extension encoder configured to generate from said multichannel audio signal encoded parametric multichannel extension information in a second processing chain distinct from said first processing chain, said extension encoder comprising;

a transforming portion configured to transform each channel of a multichannel audio signal into the frequency domain;

a separation portion configured to divide a bandwidth of frequency domain channel signals provided by said transforming portion into a first region of lower frequencies and at least one further region of higher frequencies;

a low frequency encoder configured to encode frequency domain signals provided by said grouping portion for said first frequency region with a first type of coding to obtain a parametric multichannel extension information for said first frequency region, said low frequency encoder comprising a combining portion configured to combine corresponding samples of all channels in said first region, a quantization portion configured to quantize combined samples provided by said combining portion and an encoding portion configured to encode quantized samples provided by said quantization portion, wherein the encoding portion is configured to divide said quantized samples into subblocks and to encode each subblock separately; and

at least one higher frequency encoder configured to encode frequency domain signals provided by said grouping portion for said at least one further frequency region with at least one further type of coding to obtain a parametric multichannel extension information for said at least one further frequency region.

14. Apparatus according to claim 13, wherein the encoding portion is configured apply a plurality of coding schemes to said quantized samples and to select a coding scheme which results in the lowest number of bits for said parametric multichannel extension information.

15. Apparatus according to claim 14, wherein said plurality of coding schemes comprise a plurality of Huffman coding schemes.

16. Apparatus according to claim 13, wherein said quantization portion is configured to modifying said quantized samples, in case encoding said quantized samples by said encoding portion results in more bits for said parametric multichannel extension information than are available for said first region, to obtain quantized samples which result in said encoding of quantized samples by said encoding portion at the most in the number of bits for said parametric multichannel extension information that are available for said first region.

17. Apparatus according to claim 13, wherein said quantization portion is configured to employ a selectable quantization gain for quantizing combined samples of a respective frame, and wherein said quantization portion is further configured to select a quantization gain for a respective frame which avoids sudden changes in the quantization gain from one frame to the next.

18. Apparatus according to claim 13, wherein said low frequency encoder further comprises a refinement portion which is configured to generate refinement bits representing information which allows to compensate for quantization errors in a quantization by said quantization portion, in case encoding said quantized samples by said encoding portion results in a number of bits for said parametric multichannel extension information which is lower than a number of bits which are available for said first region.

19. Apparatus according to claim 13, wherein said at least one higher frequency encoder comprises a middle frequency encoder configured to encode frequency domain signals in a middle frequency region and a high frequency encoder configured to encode frequency domain signals in a high frequency region.

20. Apparatus according to claim 19, wherein said middle frequency encoder comprises:

a processing portion configured to determine for each of a plurality of adjacent frequency bands within said middle frequency region whether a spectral first channel signal of said multichannel signal, a spectral second channel signal of said multichannel signal or none of said spectral channel signals is dominant in the respective frequency band and to provide for each frequency band a corresponding state information; and

an encoding portion configured to encode state information provided by said processing portion to obtain a parametric multichannel extension information.

21. Apparatus according to claim 20, further comprising a post-processing portion configured to post-process state information determined by said processing portion such that short-time changes in said state information are avoided before said state information is encoded by said encoding portion.

22. Apparatus according to claim 19, wherein said high frequency encoder comprises:

a processing portion configured to determine for each of a plurality of adjacent frequency bands within said middle frequency region whether a spectral first channel signal of said multichannel signal, a spectral second channel signal of said multichannel signal or none of said spectral channel signals is dominant in the respective frequency band and to provide for each frequency band a corresponding state information; and

an encoding portion configured to select and to apply a first approach or a second approach for encoding a state information provided by said processing portion to obtain a parametric multichannel extension information, wherein said first approach includes encoding a state information for each of said frequency bands provided by said processing portion, and wherein said second approach includes comparing state information provided by said processing portion for a current frame to state information provided by said processing portion for a previous frame, encoding a result of this comparison and encoding state information for a current frame only in case there was a change in said state information from said previous frame to said current frame.

23. Apparatus according to claim 22, further comprising a post-processing portion configured to post-process state information determined by said processing portion such that short-time changes in said state information are avoided before said state information is encoded by said encoding portion.

24. Apparatus according to claim 13, wherein said apparatus is one of a multichannel encoder and a mobile terminal.

25. Audio coding system comprising an apparatus according to claim 13 and an electronic device with a multichannel decoder, said multichannel decoder comprising a decoder configured to decode a provided encoded mono signal and an extension decoder, said extension decoder including:

a first decoding portion configured to decode an encoded parametric multichannel extension information which is provided for a first region of lower frequencies using a first type of coding, and to reconstruct a multichannel signal based on said decoded mono signal and on said decoded parametric multichannel extension information;

at least one further decoding portion configured to decode an encoded parametric multichannel extension information which is provided for at least one further region of higher frequencies using at least one further type of coding, and to reconstruct a multichannel signal based on said decoded mono signal and on said decoded parametric multichannel extension information;

a combining portion configured to combine reconstructed multichannel signals provided by said first decoding portion and said at least one further decoding portion; and

a transforming portion configured to transform each channel of a combined multichannel signal into a time domain.

26. Apparatus comprising:

a decoder configured to decode a provided encoded mono signal; and

an extension decoder including;

a first decoding portion configured to decode an encoded parametric multichannel extension information which is provided for a first region of lower frequencies using a first type of coding, wherein said encoded parametric multichannel extension information comprises encoded subblocks, said encoded subblocks having been obtained at an extension encoder by combining corresponding samples of all channels in said first region, quantizing said combined samples, dividing said quantized samples into subblocks and encoding each subblock separately, said first decoding portion being further configured to reconstruct a multichannel signal based on said decoded mono signal and on said decoded parametric multichannel extension information;

at least one further decoding portion configured to decode an encoded parametric multichannel extension information which is provided for at least one further region of higher frequencies using at least one further type of coding, and to reconstruct a multichannel signal based on said decoded mono signal and on said decoded parametric multichannel extension information;

a combining portion configured to combine reconstructed multichannel signals provided by said first decoding portion and said at least one further decoding portion; and

a transforming portion configured to transform each channel of a combined multichannel signal into a time domain.

27. Apparatus according to claim 26, wherein said apparatus is one of a multichannel decoder and a mobile terminal.

28. Encoder in which a software code is stored, said software code realizing the following when running in a processing component of said encoder:

generating from a multichannel audio signal an encoded mono audio signal in a first processing chain; and

generating from said multichannel audio signal encoded parametric multichannel extension information in a second processing chain distinct from said first processing chain, said generating of encoded parametric multichannel extension information comprising;

transforming each channel of said multichannel audio signal into the frequency domain;

dividing a bandwidth of said frequency domain channel signals into a first region of lower frequencies and at least one further region of higher frequencies; and

encoding said frequency domain signals in each of said frequency regions with another type of coding to obtain a parametric multichannel extension information for the respective frequency region, wherein encoding said frequency domain signals in said first region comprises combining corresponding samples of all channels in said first region, quantizing said combined samples and encoding said quantized samples, and wherein encoding said quantized samples comprises dividing said quantized samples into subblocks and encoding each subblock separately.

29. Decoder in which a software code is stored, said software code realizing the following when running in a processing component of said decoder:

decoding an encoded mono signal;

decoding an encoded parametric multichannel extension information which is provided separately for a first region of lower frequencies and for at least one further region of higher frequencies, wherein said encoded parametric multichannel extension information comprises for said first region encoded subblocks, said encoded subblocks having been obtained at an extension encoder by combining corresponding samples of all channels in said first region, quantizing said combined samples, dividing said quantized samples into subblocks and encoding each subblock separately;

reconstructing a multichannel signal based on said decoded mono signal and on said decoded parametric multichannel extension information separately for said first region and said at least one further region;

combining said reconstructed multichannel signals in said first and said at least one further region; and

transforming each channel of said combined multichannel signal into the time domain.