Method and apparatus for encoding decoding and compression of audio-type data

• US 5,394,508 A
• Filed: 01/17/1992
• Issued: 02/28/1995
• Est. Priority Date: 01/17/1992
• Status: Expired due to Term

• Current Assignee: Massachusetts Institute of Technology
• Original Assignee: Massachusetts Institute of Technology
• Inventors: Lim, Jae S.
• Primary Examiner(s): Knepper, David D.
• Application Number:
  US07/822,247
• Time in Patent Office: 1,138 Days
• Field of Search: 381/29, 381/37, 381/39, 395/2.14, 395/2.18, 395/2.33, 395/2.38, 395/2.39, 395/2.12
• US Class Current: 704/200.1
• CPC Class Codes: G10L 19/0204 using subband decomposition
• CPC Class Codes: G10L 19/0212 using orthogonal transformation
• CPC Class Codes: G10L 2021/02163 Only one microphone
• CPC Class Codes: G10L 25/27 characterised by the analysis technique
• CPC Class Codes: H04B 1/667 using a division in frequency subbands for TV signals H04N19/63

Abstract

An audio type signal is encoded. The signal is first divided into bands. For each band, a yardstick signal element is selected. Its magnitude is quantized using a first level of accuracy. This magnitude is used for various purposes, including assigning bits to the different bands, and for establishing reconstruction levels within a band. The magnitude of non yardstick signal elements is quantized with less accuracy than are the yardstick signal elements. The encoded signal is also decoded. Apparatus for both encoding and decoding are also disclosed. The location of the yardstick element within its band may also be recorded and encoded, and used for efficiently allocating bits to non-yardstick signal elements.

Claims

1. A method for encoding a selected signal element of a signal that is defined by signal elements that are discrete in at least one dimension, said method comprising the steps of:

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements, one of said bands including said selected signal element;

b. in each band, identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element only one time;

d. allocating to said selected signal element a signal element bit allocation that is a function of the quantized magnitudes of said yardstick signal elements.

2. The method of claim 1, further comprising the step, before said step of dividing said signal into a plurality of bands, the step of transforming said signal from a first domain into another domain, such that said signal is defined by a different set of signal elements that are discrete in at least one dimension, all of said subsequent steps being performed with respect to said transformed signal defined by said different set of signal elements.

3. The method of claim 1, further comprising, before said step of allocating a signal element bit allocation, the step of:

for said band including said selected signal element, allocating to said band from the total available number of bits N, a band bit allocation that is a function of;

a. the available remaining number of bits;

b. the number of signal elements in said band and the quantized magnitude of said respective yardstick signal element; and

c. for every other band;

i. the number of signal elements in said other band; and

ii. the quantized magnitude of the respective yardstick

signal element for said other band.

4. The method of claim 1, further comprising the steps of:

a. specifying the location of said respective yardstick signal element with respect to at least one dimension within its band as well as the sign of said yardstick signal element; and

b. encoding said yardstick signal element using the quantized magnitude, the specified location and the specified sign.

5. The method of encoding of claim 1, wherein said yardstick signal element is quantized to a first degree of accuracy, further comprising the step of quantizing the magnitude of said selected signal element to a second degree of accuracy that is less accurate than said first degree of accuracy.

6. The method of claim 5, further comprising, before said step of allocating said signal element bit allocation, the step of specifying the location of said respective yardstick signal element within its band with respect to said at least one dimension in which said signal elements are discrete and where said function applied during said step of allocating said bit allocation is further based on the location of said selected signal element within said respective band relative to the location of said respective yardstick signal element.

7. The method of claim 5, said step of quantizing the magnitude of said selected signal element comprising the step of specifying a plurality of non-yardstick reconstruction levels said plurality being divided within a range based on the magnitude of said yardstick signal element of said respective band.

8. A method for encoding a selected signal element of a signal that is defined by signal elements that are discrete in at least one dimension, said method comprising the steps of:

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements;

b. in each band, identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element to a first degree of accuracy;

d. specifying the location of said respective yardstick signal element with respect to at least one dimension within its band in which said signal elements are discrete;

e. allocating to said selected signal element a signal element bit allocation that is a function of;

i. the quantized magnitudes of said yardstick signal elements;

ii. the location of said selected signal element within said respective band relative to the location of said respective yardstick signal element, such that;

iii. bits are allocated to said selected signal element more preferentially if the location of said selected signal element is between the location of said respective yardstick signal element and a selected end of said band as compared to if said signal element is not located between said respective yardstick signal element and said selected end of said band; and

iv. quantization of said selected signal element using said signal element bit allocation is to a second degree of accuracy, which is less accurate than said first degree of accuracy; and

f. quantizing the magnitude of said selected signal element to said second degree of accuracy.

9. A method for encoding a selected signal element of a signal that is defined by signal elements that are discrete in at least one dimension, said method comprising the steps of:

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements;

b. in each band identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element to a first degree of accuracy;

d. measuring the sign of said respective yardstick signal element;

e. quantizing the magnitude of said selected signal element to a second degree of accuracy that is less accurate than said first degree of accuracy by specifying a plurality of non-yardstick reconstruction levels of non-uniform ranges, said non-uniform arrangement being based on said sign of said yardstick signal element, said plurality of non-yardstick reconstruction levels being divided within a range based on the magnitude of said yardstick signal element of said respective band; and

f. allocating to said selected signal element a signal element bit allocation that is a function of the quantized magnitudes of said yardstick signal elements, said signal element bit allocation chosen such that quantization of said selected signal element using said signal element bit location is to a second degree of accuracy, which is less accurate than said first degree of accuracy.

10. The method of claim 9, said step of specifying reconstruction levels of non-uniform range comprising the step of specifying more reconstruction levels having the same sign as the yardstick signal element than are specified having the sign opposite from the yardstick signal element.

11. A method for encoding a selected signal element of a signal that is defined by signal elements that are discrete in at least one dimension, said method comprising the steps

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements;

b. in each band, identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element to a first degree of accuracy;

quantizing the magnitude of said selected signal element to a second degree of accuracy that is less accurate than said first degree of accuracy by specifying a plurality of non-yardstick reconstruction levels, said plurality of reconstruction levels being divided within a range based on the magnitude of said yardstick signal element of said respective band, wherein said step of specifying a plurality of non-yardstick reconstruction levels further comprises the steps of;

i. dividing said plurality of non-yardstick reconstruction, levels within a range of less than the magnitude of said yardstick signal element of said respective band if said yardstick signal element is of a positive amplitude; and

ii. dividing said plurality of non-yardstick reconstruction levels within a range of less than twice the magnitude of said yardstick signal element of said respective band if said yardstick signal element is of a negative amplitude; and

e. allocating to said selected signal element a signal element bit allocation that is a function of the quantized magnitudes of said yardstick signal elements, said signal element bit allocation chosen such that quantization of said selected signal element using said signal element bit allocation is to a second degree of accuracy, which is less accurate than said first degree of accuracy.

12. A method for encoding a signal that is defined by signal elements that are discrete in at least one dimension, said method comprising the steps of:

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements;

b. in each band, identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element to a first degree of accuracy;

d. for at least one non-yardstick signal element, allocating a non-yardstick signal element bit allocation that is a function of the quantized magnitudes of said yardstick signal elements, said signal element bit allocation being chosen such that quantization of said at least one non-yardstick signal element using said signal element bit allocation is to a second degree of accuracy, which is less accurate than said first degree of accuracy;

e. for each said at least one non-yardstick signal element measuring the sign of said respective yardstick signal element, and quantizing the magnitude of said non-yardstick signal element to said second degree of accuracy that is less accurate than said first degree of accuracy by specifying a plurality of non-yardstick reconstruction levels of unequal spacing between consecutive reconstruction levels, said unequal arrangement being based on said sign of said yardstick signal element.

13. A method for decoding a code representing a selected signal element of a signal that is defined by signal elements that are discrete in at least one dimension, which has been encoded by a method comprising the steps of:

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements, one of said bands including said selected signal element;

b. in each band, identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element only one time;

d. allocating to said selected signal element a signal element bit allocation that is a function of the quantized magnitudes of the yardstick signal elements; and

e. quantizing the magnitude of said selected signal element and providing a codeword for said quantized selected signal;

said method of decoding comprising the step of of translating said codeword based on a function of the quantized magnitudes of said yardstick signal elements that is appropriately inversely related to said function of the quantized magnitudes used to allocate bits to said selected signal element.

14. A method for encoding a selected signal element of a signal that is defined by signal elements that are discrete in at least one dimension, said method comprising the steps of:

a. dividing the signal into a plurality of bands, at least one band having a plurality of adjacent signal elements;

b. in each band, identifying a signal element having the greatest magnitude of any signal element in said band, and designating said signal element as a "yardstick" signal element for said band;

c. quantizing the magnitude of each yardstick signal element to a first degree of accuracy;

d. allocating to said selected signal element a signal element bit allocation that is a function of the quantized magnitudes of said yardstick signal elements, said signal element bit allocation chosen such that quantization of said selected signal element using said signal element bit allocation is to a second degree of accuracy, which is less accurate than said first degree of accuracy;

e. specifying the location of said respective yardstick signal element with respect to at least one dimension within its band as well as the sign of said yardstick signal element; and

f. encoding said yardstick signal element using the accurately quantized magnitude, the specified location and the specified sign.