Method, apparatus and recording medium for coding of separated tone and noise characteristic spectral components of an acoustic sibnal

• US 5,717,821 A
• Filed: 03/20/1995
• Issued: 02/10/1998
• Est. Priority Date: 05/31/1993
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Tsutsui, Kyoya, Heddle, Robert
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): Smits, Talivaldis Ivars
• Application Number:
  US08/374,518
• Time in Patent Office: 1,058 Days
• Field of Search: 395/2.14, 395/2.15, 395/2.17, 395/2.29, 395/2.67, 395/2.77, 381/41, 381/51
• US Class Current: 704/200.1
• CPC Class Codes: G10L 19/002 Dynamic bit allocation for perceptual audio coders G10L19/032
• CPC Class Codes: G10L 19/0212 using orthogonal transformation
• CPC Class Codes: G10L 21/0232 Processing in the frequency domain
• CPC Class Codes: G10L 21/0264 characterised by the type of parameter measurement, e.g. correlation techniques, zero crossing techniques or predictive techniques
• CPC Class Codes: H04B 1/665 using psychoacoustic properties of the ear, e.g. masking effect

Abstract

A method, and the apparatus for its implementation, for encoding an input signal includes the steps of transforming an input signal into frequency components, as for example by spectrum transform processing, separating the frequency components into a first signal consisting of tone characteristic components and a second signal consisting of other components, encoding the first signal, and encoding a signal based on the second signal.

Claims

1. A method of encoding an input acoustic signal, comprising the steps of:

transforming the input acoustic signal into a block of frequency components, each frequency component characterized by an energy distribution;

separating the frequency components into a first signal consisting of tone characteristic components, each tone characteristic component having a sharp energy distribution relative to energy distributions of frequency components in a local neighborhood of the tone characteristic component and to energy distributions of frequency components in the block, and a second signal consisting of components other than the tone characteristic components;

encoding the first signal; and

encoding a signal based on the second signal.

2. A signal encoding method as set forth in claim 1, wherein the transform processing is spectrum transform processing.

3. A signal encoding method as set forth in claim 1, wherein position information on the frequency base of the tone characteristic component is encoded.

4. A signal encoding method as set forth in claim 3, wherein the position information includes information indicating a difference between position information of a current block and position information of any other time block.

5. A signal encoding method as set forth in claim 1, wherein numeric information within a predetermined range of the tone characteristic component is encoded.

6. A signal encoding method as set forth in claim 1, wherein the signal based on the second signal is a signal in which a signal of an encoding unit including the principal portion of the tone characteristic components of the first signal of frequency components is caused to be zero.

7. A signal encoding method as set forth in claim 1, wherein the signal based on the second signal is a signal in which the first signal and neighboring frequency components thereof of the frequency component are caused to be zero.

8. A signal encoding method as set forth in claim 7, wherein the number of frequency components caused to be zero with respect to respective frequency components of the first signal is such that the number of such frequency component on a higher frequency band side is greater than that on a lower frequency band side.

9. A signal encoding method as set forth in claim 7, wherein the number of frequency components caused to be zero with respect to one frequency component of the first signal is asymmetric with respect to a higher frequency band side and a lower frequency band side with one frequency component of the first signal being as a center.

10. A signal encoding method as set forth in claim 1,

wherein the step of separation includes the steps of encoding the tone characteristic component, decoding the encoded tone characteristic component, and subtracting the decoded tone characteristic component from any frequency component of the input signal to generate a difference signal, and

wherein the second signal is the difference signal.

11. A signal encoding method as set forth in claim 10

wherein the step of separation at least once includes the steps of encoding a tone characteristic component of the difference signal, decoding the encoded tone characteristic component, and subtracting the decoded tone characteristic component from the difference signal to generate a new difference signal to allow the new difference signal to be the difference signal, and

wherein the second signal is the new difference signal.

12. A signal encoding method as set forth in claim 1,

wherein at least the step of encoding of the signal based on the second signal of the step of encoding of the first signal and the step of encoding of the signal based on the second signal includes the steps of:

normalizing an inputted signal every encoding units; and

quantizing the normalized signals.

13. A signal encoding method as set forth in claim 1,

wherein at least the step of encoding of the signal based on the second signal of the step of encoding of the first signal and the step of encoding of the signal based on the second signal includes:

a step of implementing variable length encoding to an inputted signal.

14. A signal encoding method as set forth in claim 1, wherein the step of separation separates the first signal only from a high frequency band of the frequency components.

15. A signal encoding method as set forth in claim 1, wherein the step of transform processing carries out transform processing so that frequency resolution on a lower frequency band side is higher than frequency resolution on a higher frequency band side.

16. A signal encoding method as set forth in claim 1, wherein the input signal is audio signal.

17. A signal encoding apparatus adapted for encoding an input acoustic signal, comprising:

means for transforming the input acoustic signal into a block of frequency components, each frequency component characterized by an energy distribution;

means for separating the frequency components into a first signal consisting of tone characteristic components, each tone characteristic component having a sharp energy distribution relative to energy distributions of frequency components in a local neighborhood of the tone characteristic component and to energy distributions of frequency components in the block, and a second signal consisting of components other than the tone characteristic components;

first encoding means for encoding the first signal; and

second encoding means for encoding a signal based on the second signal.

18. A signal encoding apparatus as set forth in claim 17, wherein the transform processing is spectrum transform processing.

19. A signal encoding apparatus as set forth in claim 17, which includes:

position information encoding means for encoding position information on the frequency base of the tone characteristic component.

20. A signal encoding apparatus as set forth in claim 19, wherein the position information includes position information indicating a difference between position information of a current block and position information of any other time block.

21. A signal encoding apparatus as set forth in claim 17, which includes;

encoding means for encoding numeric information within a predetermined range of the tone characteristic components.

22. A signal encoding apparatus as set forth in claim 17, wherein the signal based on the second signal is a signal in which a signal of an encoding unit including the principal portion of tone characteristic components of the first signal of frequency components is caused to be zero.

23. A signal encoding apparatus as set forth in claim 17, wherein the signal based on the second signal is a signal in which the first signal and neighboring frequency components thereof of the frequency component are caused to be zero.

24. A signal encoding apparatus as set forth in claim 23,

wherein the number of frequency components caused to be zero with respect to respective frequency components of the first signal is such that the number of such components on a higher frequency band side is greater than that on a lower frequency band side.

25. A signal encoding apparatus as set forth in claim 23, wherein the number of frequency components caused to be zero with respect to one frequency component of the first signal is asymmetric with respect to a higher frequency band side and a lower frequency band side with one frequency component of the first signal being as a center.

26. A signal encoding apparatus as set forth in claim 17,

wherein the separating means includes:

encoding means for encoding the tone characteristic component;

decoding means for decoding the encoded tone characteristic component; and

means for subtracting the decoded tone characteristic component from any frequency component of the input signal to generate a difference signal, and

wherein the second signal is the difference signal.

27. A signal encoding apparatus as set forth in claim 26,

wherein the separating means includes:

encoding means for encoding a tone characteristic component of the difference signal;

decoding means for decoding the encoded tone characteristic component; and

means for subtracting the decoded tone characteristic component from the difference signal to generate a new difference signal to output the new difference signal as the difference signal, and

wherein the second signal is the new difference signal.

28. A signal encoding apparatus as set forth in claim 17,

wherein at least the second encoding means of the first encoding means and the second encoding means includes:

normalizing means for normalizing an inputted signal every encoding units; and

quantizing means for quantizing the normalized signals.

29. A signal encoding apparatus as set forth in claim 17,

wherein at least the second encoding means of the first encoding means and the second encoding means includes:

variable length encoding means for implementing variable length encoding to an inputted signal.

30. A signal encoding apparatus as set forth in claim 17,

wherein the separating means separates the first signal only from a higher frequency band of the frequency components.

31. A signal encoding apparatus as set forth in claim 17,

wherein the transform means carries out transform processing so that frequency resolution on a lower frequency band side is higher than frequency resolution on a higher frequency band side.

32. A signal encoding apparatus as set forth in claim 17, wherein the input signal is audio signal.

33. A signal recording medium on which encoded signals are recorded,

wherein the signal recording medium is formed by the steps of:

transforming an input acoustic signal into a block of frequency components, each frequency component characterized by an energy distribution;

separating the frequency components into a first signal consisting of tone characteristic components, each tone characteristic component having a sharp energy distribution relative to energy distributions of frequency components in a local neighborhood of the tone characteristic component and to energy distributions of frequency components in the block, and a second signal consisting of components other than the tone characteristic components;

encoding the first signal; and

encoding a signal based on the second signal.

recording the encoded first and second signals.

34. A signal recording medium as set forth in claim 33, wherein the transform processing is spectrum transform processing.

35. A signal recording medium as set forth in claim 33, the signal recording medium being formed by the steps of:

encoding position information on the frequency base of the tone characteristic component; and

recording the encoded position information.

36. A signal recording medium as set forth in claim 35, wherein the position information includes information indicating a difference between position information of a current block and position information of any other time block.

37. A signal recording medium as set forth in claim 33, the signal recording medium being formed by the steps of:

encoding numeric information within a predetermined range of the tone characteristic components; and

recording the encoded numeric information.

38. A signal recording medium as set forth in claim 33, wherein the signal based on the second signal is a signal in which a signal of an encoding unit including the principal portion of tone characteristic components of the first signal of frequency components is caused to be zero.

39. A signal recording medium as set forth in claim 33,

wherein the signal based on the second signal is a signal in which the first signal and neighboring frequency components thereof of the frequency components are caused to be zero.

40. A signal recording medium as set forth in claim 39, wherein the number of frequency components caused to be zero with respect to respective frequency components of the first signal is such that the number of such frequency components on a higher frequency band side is greater than that on a lower frequency band side.

41. A signal recording medium as set forth in claim 39, wherein the number of frequency components caused to be zero with respect to one frequency component of the first signal is asymmetric with respect to a higher frequency band side and a lower frequency band side with one frequency component of the first signal being as a center.

42. A signal recording medium as set forth in claim 33,

wherein the step of separation includes the steps of

encoding the tone characteristic component, decoding the encoded tone characteristic component, and subtracting the decoded tone characteristic component from any frequency component of the input signal to generate a difference signal, and

wherein the second signal is the difference signal.

43. A signal recording medium as set forth in claim 42,

wherein the step of separation includes at least once the steps of encoding a tone characteristic component of the difference signal, decoding the encoded tone characteristic component, and subtracting the decoded tone characteristic component from the difference signal to generate a new difference signal to allow the new difference signal to be the difference signal, and

wherein the second signal is the new difference signal.

44. A signal recording medium as set forth in claim 33,

wherein at least the step of encoding of the signal based on the second signal of the step of encoding of the first signal and the step of encoding of the signal based on the second signal includes a step of:

normalizing an inputted signal every encoding units; and

quantizing the normalized signals.

45. A signal recording medium as set forth in claim 33,

wherein at least the step of encoding of the signal based on the second signal of the step of encoding of the first signal and the step of encoding of the signal based on the second signal includes:

a step of implementing variable length encoding to an inputted signal.

46. A signal recording medium as set forth in claim 33,

wherein the step of separation separates the first signal only from a higher frequency band of the frequency components.

47. A signal recording medium as set forth in claim 33,

wherein the step of transform processing carries out transform processing so that frequency resolution on a lower frequency band side is higher than frequency resolution on a higher frequency band side.

48. A signal recording medium as set forth in claim 33, wherein the input signal is audio signal.

49. A signal recording medium on which encoded signals are recorded,

wherein tone characteristic component information relating to tone characteristic components and noise characteristic component information relating to noise characteristic components are separately recorded, each tone characteristic component having a sharp energy distribution relative to energy distributions of frequency components in a local neighborhood of the tone characteristic component and to energy distributions of frequency components in the block and the noise characteristic components being components other than the tonal components.

50. A signal recording medium as set forth in claim 49, wherein the tone characteristic information includes position information on the frequency base of the tone characteristic component.

51. A signal recording medium as set forth in claim 49, wherein the tone characteristic component information includes numeric information within a predetermined range of the tone characteristic component.

52. A signal recording medium as set forth in claim 49, wherein at least the noise characteristic component information of the tone characteristic component information and the noise characteristic component information includes normalization coefficient information and quantization accuracy information.

53. A signal decoding method for an encoded signal, comprising the steps of:

decoding a first signal consisting of tone characteristic components to generate a first decoded signal, each tone characteristic component having a sharp energy distribution relative to energy distributions of frequency components in a local neighborhood of the tone characteristic component and to energy distributions of frequency components in a time block;

decoding a second signal consisting of noise characteristic components to generate a second decoded signal, the noise characteristic components consisting of components other than the tone characteristic components; and

carrying out synthetic inverse transform processing to synthesize the first and second decoded signals to implement inverse transform processing thereto, or synthetic inverse transform processing to respectively inversely transform the first and second decoded signals to synthesize them.

54. A signal decoding method as set forth in claim 53, wherein the step of synthetic inverse transform processing carries out synthesis or inverse transform processing on the basis of position information on the frequency base of the tone characteristic component.

55. A signal decoding method as set forth in claim 53, which includes a step of carrying out separation between the first signal and the second signal on the basis of numeric information within a predetermined range of the tone characteristic components.

56. A signal decoding method as set forth in claim 53, wherein at least the step of generating the second decoded signal of the step of generating the first decoded signal and the step of generating the second decoded signal carries out inverse quantization and release of normalization with respect to an inputted signal on the basis of normalization information and quantization accuracy information.

57. A signal decoding method as set forth in claim 53, wherein the encoded signal is audio signal.

58. A signal decoding apparatus for encoded signals, comprising:

first decoding means for decoding a first signal consisting of tone characteristic components to generate a first decoded signal, each tone characteristic component having a sharp energy distribution relative to energy distributions of frequency components in a local neighborhood of the tone characteristic component and to energy distributions of frequency components in a time block;

second decoding means for decoding a second signal consisting of noise characteristic components to generate a second decoded signal, the noise characteristic components consisting of components other than the tone characteristic components; and

synthetic inverse transforming means for carrying out synthetic inverse transform processing to synthesize the first and second decoded signals to implement inverse transform processing thereto, or synthetic inverse transform processing to respectively inversely transform the first and second decoded signals to synthesize them.

59. A signal decoding apparatus as set forth in claim 58, wherein the synthetic inverse transform means carries out synthesis or inverse transform processing on the basis of position information on the frequency base of the tone characteristic component.

60. A signal decoding apparatus as set forth in claim 58, which includes separating means for carrying out separation between the first and second signals on the basis of numeric information within a predetermined range of the tone characteristic components.

61. A signal decoding apparatus as set forth in claim 58, wherein the first decoding means and the second decoding means include:

means for carrying out inverse quantization and releases of normalization with respect to an inputted signal on the basis of normalization information and quantization accuracy information.

62. A signal decoding apparatus as set forth in claim 58, wherein the encoded signal is audio signal.