Bit allocation of digital audio signal blocks by non-linear processing

• US 5,608,713 A
• Filed: 02/08/1995
• Issued: 03/04/1997
• Est. Priority Date: 02/09/1994
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Oikawa, Yoshiaki, Suzuki, Hiroshi, Akagiri, Kenzo
• Primary Examiner(s): Epps, Georgia Y.
• Assistant Examiner(s): CHU, KIM KWOK
• Application Number:
  US08/385,287
• Time in Patent Office: 755 Days
• Field of Search: 369/124, 369/54, 369/275.1, 369/107, 341/51, 381/37, 381/29, 381/30, 381/45, 381/31, 370/109, 370/118, 375/1, 375/122, 375/240, 395/2.38, 395/2.12, 395/2.35, 395/2.33, 395/2.39, 395/2.13, 380/4, 380/3, 380/28, 380/46, 380/49
• US Class Current: 369/124.08
• CPC Class Codes: G11B 20/00007 Time or data compression or expansion audio compression based on psychoacoustics G10L19/00; data processing for reproducing audio data at different playback speeds G10L21/04; video compression H04N19/00; data compression per se H03M7/30
• CPC Class Codes: H04B 1/665 using psychoacoustic properties of the ear, e.g. masking effect

Abstract

A method and an apparatus for transmitting digital signals are disclosed. The input digital signal is converted into a plurality of blocks each having a finite time width and an finite frequency width. Signal components of at least some of the blocks are processed in a non-linear fashion, and the resulting non-linearly processed signal components are quantized. A recording medium on which encoded digital signals are recorded is also disclosed.

Claims

1. A recording medium having an encoded digital signal recorded thereon, said recording medium being prepared by the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing signal components of at least some of the blocks; and

quantizing the non-linearly processed signal components for generating recording data, and recording said recording data on the recording medium, wherein said signal components are spectral components and said step of non-linearly processing includes enlarging the spectral components except at least the spectral component which gives the maximum value in each block.

2. A digital signal processing method, comprising the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing the signal components of at least some of the blocks; and

quantizing the non-linearly processed signal components, wherein said signal components are spectral components and said step of non-linearly processing includes enlarging the spectral components except at least the spectral component which gives the maximum value in each block.

3. A digital signal processing method, comprising the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing the signal components of at least some of the blocks; and

quantizing the non-linearly processed signal components, wherein said step of non-linearly processing includes reducing the quantization value by said quantization of the signal component excluding at least the signal component having the maximum signal-to-noise ratio in each block to zero.

4. A digital signal processing method, comprising the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing the signal components of at least some of the blocks; and

quantizing the non-linearly processed signal components, wherein said signal components are spectral components and the digital signal processing method includes normalizing said signal components, and wherein said step of non-linearly processing includes enlarging the spectral components having magnitudes intermediate between a first comparison level lower than the normalization level for said normalization and a second comparison level smaller than said first comparison level, or reducing the quantization value of said spectral component by said quantization to zero, and

reducing the quantization value by said quantization of the spectral component smaller in magnitude than said second comparison level to zero.

5. The digital signal processing method as claimed in claim 4, wherein said first comparison level and the second comparison level are variable depending on the maximum value of the spectral component in each block.

6. The digital signal processing method as claimed in claim 5, wherein the larger the maximum spectral value in each block, at least one of said first comparison level and said second comparison level becomes lower and higher, respectively.

7. A digital signal processing method, comprising the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width, a finite frequency width and signal components;

non-linearly processing the signal components of at least some of the blocks;

quantizing the non-linearly processed signal components; and

selecting a block having the word length as determined by bit allocation found on the basis of said signal component prior to said non-linear processing shorter than a pre-set word length as a block to be processed non-linearly.

8. A digital signal processing method, comprising the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width, wherein said signal components are spectral components;

non-linearly processing the signal components of at least some of the blocks;

quantizing the non-linearly processed signal components; and

selecting the block to be processed non-linearly on the basis of the maximum spectral value in each block.

9. The digital signal processing method as claimed in claim 8, wherein, if the maximum spectral value of a block is not less than a pre-set value, the block is selected as a block to be processed non-linearly.

10. A digital signal processing method, comprising the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing the signal components of at least some of the blocks;

quantizing the non-linearly processed signal components; and

selecting the block to be processed non-linearly on the basis of tonality in each block.

11. The digital signal processing method as claimed in claim 10, wherein said signal components are spectral components, and wherein said tonality is found on the basis of a first component which at least is the spectral component among the spectral components in each block having the maximum signal-to-noise ratio and a second component having spectral components in each block excluding said first component.

12. The digital signal processing method as claimed in claim 11, wherein said tonality is a ratio of a first value derived from said first component and a second value derived from said second component.

13. The digital signal processing method as claimed in claim 12, wherein said first value is an effective value of said first component and said second value is an effective value of said second component.

14. A recording medium having an encoded digital signal recorded thereon, said recording medium being prepared by the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing signal components of at least some of the blocks; and

quantizing the non-linearly processed signal components for generating recording data, and recording said recording data on the recording medium, wherein said step of non-linearly processing includes reducing the quantization value by said quantization of the signal component excluding at least the signal component having the maximum signal-to-noise ratio in each block to zero.

15. A digital signal processing apparatus, comprising:

means for converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linear processing means for non-linearly processing the signal components in at least some of said blocks; and

encoding means for quantizing the non-linearly processed signal components, wherein said signal components are spectral components and said non-linear processing means enlarges the spectral components except at least the spectral component which gives the maximum value in each block.

16. A digital signal processing apparatus, comprising:

means for converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linear processing means for non-linearly processing the signal components in at least some of said blocks; and

encoding means for quantizing the non-linearly processed signal components, wherein said non-linear processing means reduces the quantization value by said quantization of the signal component excluding at least the signal component having the maximum signal-to-noise ratio in each block to zero.

17. A digital signal processing apparatus, comprising:

means for converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linear processing means for non-linearly processing the signal components in at least some of said blocks; and

encoding means for quantizing the non-linearly processed signal components, wherein said signal components are spectral components and said encoding means normalizes said signal components, and wherein said non-linear processing means enlarges the spectral components having magnitudes intermediate between a first comparison level lower than the normalization level for said normalization and a second comparison level smaller than said first comparison level, or reduces the quantization value of said spectral component by said quantization to zero, said non-linear processing means reducing the quantization value by said quantization of the spectral component smaller in magnitude than said second comparison level to zero.

18. The digital signal processing apparatus as claimed in claim 17, wherein said first comparison level and the second comparison level are variable depending on the maximum value of the spectral component in each block.

19. The digital signal processing apparatus as claimed in claim 18, wherein the larger the maximum spectral value in each block, at least one of said first comparison level and said second comparison level becomes lower and higher, respectively.

20. A digital signal processing apparatus, comprising:

means for converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linear processing means for non-linearly processing the signal components in at least some of said blocks;

encoding means for quantizing the non-linearly processed signal components; and

means for selecting a block having a word length as determined by bit allocation found on the basis of said signal component prior to said non-linearly processing a shorter than a pre-set word length as a block to be processed non-linearly.

21. A digital signal processing apparatus, comprising:

means for converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linear processing means for non-linearly processing the signal components in at least some of said blocks;

encoding means for quantizing the non-linearly processed signal components; and

means for selecting the block to be processed non-linearly on the basis of the maximum spectral value in each block.

22. The digital signal processing apparatus as claimed in claim 21, wherein, if the maximum spectral value of a block is not less than a pre-set value, the block is selected as a block to be processed non-linearly.

23. A digital signal processing apparatus, comprising:

means for converting a digital signal into signal components in a a plurality of blocks, each block having a finite time width and finite frequency width;

non-linear processing means for non-linearly processing the signal components in at least some of said blocks;

encoding means for quantizing the non-linearly processed signal components; and

means for selecting the block to be processed non-linearly on the basis of tonality in each block.

24. The digital signal processing apparatus as claimed in claim 23, wherein said signal components are spectral components, and wherein said tonality is found on the basis of a first component which at least is the spectral component among the spectral components in each block having the maximum signal-to-noise ratio and a second component composed of spectral components in each block excluding said first component.

25. The digital signal processing apparatus as claimed in claim 24, wherein said tonality is a ratio of a first value derived from said first component and a second value derived from said second component.

26. The digital signal processing apparatus as claimed in claim 25, wherein said first value is an effective value of said first component and said second value is an effective value of said second component.

27. A recording medium having an encoded digital signal recorded thereon, said recording medium being prepared by the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width, wherein the signal components are spectral components;

non-linearly processing signal components of at least some of the blocks;

quantizing the non-linearly processed signal components for generating recording data, and recording said recording data on the recording medium; and

selecting the block to be processed non-linearly on the basis of the maximum spectral value in each block.

28. The recording medium as claimed in claim 27, wherein, if the maximum spectral value of a block is not less than a pre-set value, the block is selected as a block to be processed non-linearly.

29. The recording medium as claimed in claim 28, prepared by a further step of:

selecting the block to be processed non-linearly on the basis of tonality in each block.

30. The recording medium as claimed in claim 29, wherein said signal components are spectral components, and wherein said tonality is found on the basis of a first component which at least is the spectral component among the spectral components in each block having the maximum signal-to-noise ratio and a second component composed of spectral components in each block excluding said first component.

31. The recording medium as claimed in claim 30, wherein said tonality is a ratio of a first value derived from said first component and a second value derived from said second component.

32. The recording medium as claimed in claim 31, wherein said first value is an effective value of said first component and said second value is an effective value of said second component.

33. A recording medium having an encoded digital signal recorded thereon, said recording medium being prepared by the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing signal components of at least some of the blocks;

quantizing the non-linearly processed signal components for generating recording data, and recording said recording data on the recording medium; and

selecting a block having the word length as determined by bit allocation found on the basis of said signal component prior to said non-linear processing shorter than a preset word length as a block to be processed non-linearly.

34. A recording medium having an encoded digital signal recorded thereon, said recording medium being prepared by the steps of:

converting a digital signal into signal components in a plurality of blocks, each block having a finite time width and a finite frequency width;

non-linearly processing signal components of at least some of the blocks; and

quantizing the non-linearly processed signal components for generating recording data, and recording said recording data on the recording medium, wherein said signal components are spectral components and the recording data is generated by normalizing said signal components, and wherein said non-linear processing includes

enlarging the spectral components having magnitudes intermediate between a first comparison level lower than the normalization level for said normalization and a second comparison level smaller than said first comparison level, or reducing the quantization value of said spectral component by said quantization to zero, and

reducing the quantization value by said quantization of the spectral component smaller in magnitude than said second comparison level to zero.

35. The recording medium as claimed in claim 34, wherein said first comparison level and the second comparison level are variable depending on the maximum value of the spectral component in each block.

36. The recording medium as claimed in claim 35, wherein the larger the maximum spectral value in each block, at least one of said first comparison level and said second comparison level becomes lower and higher, respectively.