Table controlled dynamic bit allocation in a variable rate sub-band speech coder

• US 4,899,384 A
• Filed: 08/25/1986
• Issued: 02/06/1990
• Est. Priority Date: 08/25/1986
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Ware, Malcolm S., Crouse, William G.
• Primary Examiner(s): Harkcom, Gary V.
• Assistant Examiner(s): NOT, DEFINED
• Application Number:
  US06/900,113
• Time in Patent Office: 1,261 Days
• Field of Search: 381/29-32, 381/36-40, 381/41, 364/513.5, 375/122, 375/25-26, 375/34
• US Class Current: 704/229
• CPC Class Codes: G10L 19/24 Variable rate codecs, e.g. for generating different qualities using a scalable representation such as hierarchical encoding or layered encoding
• CPC Class Codes: H04B 1/667 using a division in frequency subbands for TV signals H04N19/63

Abstract

A sub-band coder utilizes one or more bit allocation tables to dynamically distribute the channel bit capacity bandwidth among the frequency bands according to the desired output quality of speech rather than by means of complex algorithms or simulation techniques. Multiple bit assignment tables are provided to allow various quality levels to be traded off as increasing bit rate demands are placed upon the transmission system. The technique is used for a single coder to achieve a minimum bit rate for a desired given level of subjective quality in speech output or may be used in a shared bit resource to maintain equal and minimum quality degradation for all users. The quality tables determine the number of bits to be dropped from the encoded representation of each signal sample to minimize the transmission load for a given coder without sacrificing speech quality to an unacceptable degree. Table entries are arranged based on the overall band peak energy level and on the sub-band peak energy distribution or spectrum as it is known in the field.

Claims

1. In a sub-band speech coder system having means for sampling input speech signals, means for converting the input speech signal samples to digital sample values, means for subdividing said digital sample values into spectral frequency sub-bands and means for transmitting normalized and compacted values of said speech signal samples together with encoded indicia descriptive of the normalization and compaction used to a remote receiver, an improved method of bit allocation for compacted input signal sample data comprising steps of:

sampling the input speech signals and converting the resulting samples to digital data samples;

dividing the sampled input signals into time blocks of digital data samples;

measuring the block peak energy level of said digital data samples in each said time block of digital data samples;

measuring the individual sub-band peak energy levels in each said spectral frequency sub-band in said digital data samples;

deleting a variable number of bits from each of said digital data samples in said time block in accordance with the values of said block peak energy level and of said sub-band peak energy level measurements and the spectral distribution of said sub-band peak energy levels in said time block, thereby allocating a variable total number of bits for each said time block.

2. A method as described in claim 1, wherein said deleting of bits in accordance with said values of said block peak energy level measurement and of said sub-band peak energy level measurements and the spectral distribution of said sub-band peak energy levels is accomplished by:

accessing a stored table of values representing the number of bits to be deleted from each digital data sample in said time block, said accessing being accomplished by addressing said table of values using representations of said peak energy measurements, and deleting the number of bits found in said table from each said digital data sample.

3. A method as described in claim 1 or 2, wherein:

said deletion of bits is performed on the least significant bits in each of said digital data samples.

4. A method as described in 2, wherein multiple tables of said values are included and further including a step of:

selectively accessing only one of said tables during a said time block, said accessed table being selected in accordance with whether higher or lower quality of speech representation is desired.

5. A method as described in 2 wherein multiple tables of said values are included and further including a step of:

selectively accessing only one of said tables during a said time block, said accessed table being selected in accordance with the currently available bit transmission bandwidth and the current total data traffic transmission demand in said system.

6. An improved sub-band speech coder apparatus having means for sampling input analog speech signals, means for converting the input analog signal samples to digital data sample values, means for subdividing the input signal frequency spectrum samples into frequency sub-bands and means for transmitting normalized and compacted values of said speech signal samples together with encoded indicia representing and describing the normalization and compaction used on said speech signal samples, further comprising:

means for dividing the sample digital data signals into time blocks each containing an equal number of samples;

means for measuring the block peak energy sample occurring in each said time block of digital data signal samples;

means for measuring the individual frequency sub-band peak energy levels in each said time block of digital data signal samples;

means for deleting a variable number of bits from each said digital data signal sample in said time block in accordance with the values of said block peak energy and of said sub-band peak energy level measurements and the spectral distribution of said sub-band peak energy levels in said time block.

7. Apparatus as described in claim 6, wherein:

said means for deleting bits from each of said digital data signal samples comprises at least one stored table of values representing the number of bits to be deleted, said values being arranged in said table in accordance with the values of said peak energy measurements.

8. Apparatus as described in 7, wherein:

said stored table of values contains a plurality of sets of values, only one of which may be selected during transmission of a given time block of said digital data signal samples, said sets of values being arranged to provide varying degrees of accuracy of speech analog signal representation.