Excitation for higher band coding in a codec utilising band split coding methods

• US 7,376,554 B2
• Filed: 07/14/2004
• Issued: 05/20/2008
• Est. Priority Date: 07/14/2003
• Status: Active Grant

• Current Assignee: Vivo Mobile Communication Co., Ltd. (Vivo Communication Technology Co. Ltd.)
• Original Assignee: Nokia Corporation
• Inventors: Ojala, Pasi S., Vainio, Janne, Mikkola, Hannu J.
• Primary Examiner(s): Vo; Huyen X.
• Application Number:
  US10/891,846
• Publication Number:
  US 20050065783A1
• Time in Patent Office: 1,406 Days
• Field of Search: 704/203, 704/201, 704/205, 704/206, 704/207, 375/240
• US Class Current: 704/207
• CPC Class Codes: G10L 21/038 using band spreading techniques

Abstract

Methods and arrangements are disclosed for digitally encoding and decoding sound—An input signal is split (811) into a primary frequency band and at least one secondary frequency band. The parts of the input signal in the frequency bands are separately encoded. Certain characteristics of the input signal in the primary frequency band and corresponding characteristics of the input signal in at least one secondary frequency band are examined (302, 303, 814) in order to find out, whether there is certain resemblance therebetween. Alternatively certain characteristic features of the process applied to encoding the primary frequency band extracted (305, 813) and used (307) in encoding the secondary frequency band, or such extracted characteristic features are replaced (306, 501, 701, 815) with a locally generated, independent set of corresponding features.

Claims

1. A method for digitally encoding sound, comprising:

a) splitting an input signal into a primary frequency band and at least one secondary frequency band;

b) digitally encoding the part of the input signal in the primary frequency band;

c) examining certain characteristics of the input signal in the primary frequency band and corresponding characteristics of the input signal in at least one secondary frequency band in order to find out, whether there is certain resemblance therebetween; and

d1) if there is said certain resemblance, extracting certain characteristic features of the process applied to encoding the input signal in the primary frequency band and digitally encoding the part of the input signal in the secondary frequency band or bands using such extracted characteristic features;

d2) otherwise replacing such extracted characteristic features with a locally generated, independent set of corresponding features and digitally encoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features.

2. A method according to claim 1:

wherein the digitally encoding the part of the input signal in the primary frequency band corresponds to applying linear predictive coding to the input signal in the primary frequency band, involving the generation of a primary frequency band excitation signal,

wherein digitally encoding the part of the input signal in the secondary frequency band or bands corresponds to applying linear predictive coding to the input signal in a secondary frequency band, involving the use of a secondary frequency band excitation signal,

wherein the extracting certain characteristic features of the process applied to encoding the input signal in the primary frequency band and digitally encoding the part of the input signal in the secondary frequency band or bands using such extracted characteristic features, corresponds to extracting the primary frequency band excitation signal and delivering the primary frequency band excitation signal or a derivative thereof for use as the secondary frequency band excitation signal, and

wherein the replacing such extracted characteristic features with a locally generated, independent set of corresponding features and digitally encoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features, corresponds to generating the secondary frequency band excitation signal independently of the primary frequency band excitation signal.

3. A method according to claim 2, wherein the replacing such extracted characteristic features with a locally generated, independent set of corresponding features and digitally encoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features, corresponds to generating a random excitation signal.

4. A method according to claim 2, wherein the replacing such extracted characteristic features with a locally generated, independent set of corresponding features and digitally encoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features, comprises:

examining whether the input signal in a secondary frequency band exhibits voicedness, and depending on the results of such examining;

generating a periodic excitation signal, if the input signal in a secondary frequency band was found to exhibit voicedness, or

generating a random excitation signal, if the input signal in a secondary frequency band was not found to exhibit voicedness.

5. A method according to claim 2, wherein the extracting certain characteristic features of the process applied to encoding the input signal in the primary frequency band and digitally encoding the part of the input signal in the secondary frequency band or bands using such extracted characteristic features, corresponds to extracting the primary frequency band excitation signal, resampling the primary frequency band excitation signal and using the resampled primary frequency band excitation signal as the secondary frequency band excitation signal.

6. A method according to claim 2, comprising modifying the secondary frequency band excitation signal generated in the replacing such extracted characteristic features with a locally generated, independent set of corresponding features and digitally encoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features, in order to match its signal energy with a signal energy of said primary frequency band excitation signal.

7. A method according to claim 6, comprising:

extracting the primary frequency band excitation signal,

calculating a first energy value representative of a signal energy of said primary frequency band excitation signal,

generating the secondary frequency band excitation signal,

calculating a second energy value representative of a signal energy of said secondary frequency band excitation signal, and

scaling said secondary frequency band excitation signal with a ratio of the first energy value and the second energy value.

8. A method according to claim 1, wherein:

the examining certain characteristics of the input signal in the primary frequency band and corresponding characteristics of the input signal in at least one secondary frequency band in order to find out, whether there is certain resemblance therebetween corresponds to examining, whether the input signal in the primary frequency band exhibits voicedness and whether the input signal in a secondary frequency band exhibits voicedness,

the extracting certain characteristic features of the process applied to encoding the input signal in the primary frequency band and digitally encoding the part of the input signal in the secondary frequency band or bands using such extracted characteristic features, is executed if both the input signal in the primary frequency band and the input signal in a secondary frequency band are found to exhibit voicedness or if the input signal in the primary frequency band is found to not exhibit voicedness, and

the replacing such extracted characteristic features with a locally generated, independent set of corresponding features and digitally encoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features, is executed if the input signal in the primary frequency band is found to exhibit voicedness and the input signal in a secondary frequency band is found to not exhibit voicedness.

9. A method according to claim 8, wherein the examination of whether the input signal in a frequency band exhibits voicedness comprises:

calculating a long-term correlation gain for the input signal in question and

comparing the calculated long-term correlation gain to a threshold value;

so that the input signal in a frequency band is found to exhibit voicedness if the calculated long-term correlation gain is found to be greater than a corresponding threshold value.

10. A method for decoding digitally encoded sound, comprising:

a) receiving an encoded input signal split into a primary frequency band and at least one secondary frequency band, which secondary frequency band has been encoded separately from the primary frequency band;

b) decoding the part of the input signal in the primary frequency band;

c) examining the input signal in order to find out, what indication does the input signal contain about utilising characteristic features of the process applied to encoding the primary frequency band in the process applied to encoding the secondary frequency band; and

d1) if there is said certain resemblance, extracting certain characteristic features of the process applied to decoding the input signal in the primary frequency band and decoding the part of the input signal in the secondary frequency band or bands using such extracted characteristic features, or

d2) otherwise replacing such extracted characteristic features with a locally generated, independent set of corresponding features and decoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features.

11. A method according to claim 10,

wherein decoding the part of the input signal in the primary frequency band corresponds to decoding a linear-predictive-coded input signal in the primary frequency band, involving the generation of a primary frequency band excitation signal,

wherein decoding the part of the input signal in the secondary frequency band or bands corresponds to decoding a linear-predictive-coded input signal in that secondary frequency band, involving the use of a secondary frequency band excitation signal,

wherein extracting certain characteristic features of the process applied to decoding the input signal in the primary frequency band and decoding the part of the input signal in the secondary frequency band or bands, using such extracted characteristic features, corresponds to extracting the primary frequency band excitation signal and delivering the primary frequency band excitation signal or a derivative thereof for use as the secondary frequency band excitation signal, and

wherein replacing such extracted characteristic features with a locally generated, independent set of corresponding features and decoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features, corresponds to generating the secondary frequency band excitation signal independently of the primary frequency band excitation signal.

12. A method according to claim 11, wherein replacing such extracted characteristic features with a locally generated, independent set of corresponding features and decoding the part of the input signal in the secondary frequency band or bands using said locally generated independent set of corresponding features, corresponds to generating a random excitation signal.

13. A method according to claim 11, wherein replacing such extracted characteristic features with a locally generated, independent set of corresponding features and decoding the part of the input signal in the secondary frequency band or bands using said locally generated, independent set of corresponding features,

examining whether the input signal contains an indication about periodicity in the secondary frequency band, and depending on the results of such examining:

generating a periodic excitation signal, if the input signal contains an indication about periodicity in the secondary frequency band, or

generating a random excitation signal, if the input signal does not contain any indication about periodicity in the secondary frequency band.

14. A method according to claim 11, wherein extractin certain characteristic features of the process applied to decoding the input signal in the primary frequency band and decoding the part of the input signal in the secondary frequency band or bands using such extracted characteristic features, corresponds to extracting the primary frequency band excitation signal, resampling the primary frequency band excitation signal and using the resampled primary frequency band excitation signal as the secondary frequency band excitation signal.

15. A transmitter apparatus for transmitting digitally encoded sound, comprising:

a band splitter adapted to split an input signal into a primary frequency band and at least one secondary frequency band,

a primary encoder adapted to digitally encode the part of the input signal in the primary frequency band,

a secondary encoder adapted to digitally encode the part of the input signal in a secondary frequency band,

an examining portion adapted to examine certain characteristics of the input signal in the primary frequency band and corresponding characteristics of the input signal in at least one secondary frequency band and to indicate, whether there is certain resemblance therebetween,

an extracting portion adapted to extract certain characteristic features of a process applied to encoding the input signal in the primary frequency band, for use in a process applied to encoding the input signal in the secondary frequency band, and

a replacing portion adapted to replace such extracted characteristic features with a locally generated, independent set of corresponding features in the process applied to encoding the input signal in the secondary frequency band;

wherein said extracting portion and said replacing portion are arranged to be operationally alternative to each other depending on an indication produced by said examining portion.

16. A transmitter apparatus according to claim 15, wherein:

the primary encoder is a linear predictive coder capable of generating a primary frequency band excitation signal,

the secondary encoder is a linear predictive coder capable of using a secondary frequency band excitation signal,

said extracting portion is adapted to extract a primary frequency band excitation signal and to deliver the primary frequency band excitation signal or a derivative thereof to the secondary encoder for use as the secondary frequency band excitation signal, and

said replacing portion is adapted to generate a secondary frequency band excitation signal independently of the primary frequency band excitation signal.

17. A transmitter apparatus according to claim 16, wherein said replacing portion is adapted to generate a random excitation signal.

18. A transmitter apparatus according to claim 16, wherein said replacing portion is adapted to examine whether the input signal in a secondary frequency band exhibits voicedness, and to generate a periodic excitation signal, if the input signal in a secondary frequency band was found to exhibit voicedness, and to generate a random excitation signal, if the input signal in a secondary frequency band was not found to exhibit voicedness.

19. A transmitter apparatus according to claim 16, wherein said extracting portion comprises a resampler adapted to resample the primary frequency band excitation signal and to deliver the resampled primary frequency band excitation signal for use as the secondary frequency band excitation signal.

20. A transmitter apparatus according to claim 16, comprising a signal modifier adapted to modify the secondary frequency band excitation signal generated by said replacing portion in order to match its signal energy with a signal energy of said primary frequency band excitation signal.

21. A transmitter apparatus according to claim 20, comprising:

extractor for extracting the primary frequency band excitation signal,

calculator for calculating a first energy value representative of a signal energy of said primary frequency band excitation signal,

generator for generating the secondary frequency band excitation signal,

second calculator for calculating a second energy value representative of a signal energy of said secondary frequency band excitation signal, and

scaler for scaling said secondary frequency band excitation signal with a ratio of the first energy value and the second energy value.

22. A transmitter apparatus according to claim 15, wherein:

said examining portion is adapted to examine, whether the input signal in the primary frequency band exhibits voicedness and whether the input signal in a secondary frequency band exhibits voicedness,

said extracting portion is adapted to be selected for operation if both the input signal in the primary frequency band and the input signal in a secondary frequency band are found to exhibit voicedness or if the input signal in the primary frequency band is found to not exhibit voicedness, and

said replacing portion is adapted to be selected for operation if the input signal in the primary frequency band is found to exhibit voicedness and the input signal in a secondary frequency band is found to not exhibit voicedness.

23. A transmitter apparatus according to claim 22, wherein the examining portion is adapted to calculate long-term correlation gains for input signals and to compare the calculated long-term correlation gains to threshold values, so that the input signal in a frequency band is found to exhibit voicedness if the calculated long-term correlation gain is found to be greater than a corresponding threshold value.

24. A receiver apparatus for receiving and decoding digitally encoded sound, comprising:

a receiver adapted to receive an encoded input signal split into a primary frequency band and at least one secondary frequency band, which secondary frequency band has been encoded separately from the primary frequency band,

a primary decoder adapted to decode the part of the input signal in the primary frequency band,

a secondary decoder adapted to decode the part of the input signal in a secondary frequency band,

a examining portion adapted to examine the input signal and to find out what indication does the input signal contain about utilising characteristic features of the process applied to encoding the primary frequency band in the process applied to encoding the secondary frequency band,

an extracting portion adapted to extract certain characteristic features of a process applied to decoding the input signal in the primary frequency band and to use such extracted characteristic features in a process applied to decoding the input signal in the primary frequency band, and

a replacing portion adapted to replace such extracted characteristic features with a locally generated, independent set of corresponding features in the process applied to decoding the input signal in the primary frequency band;

wherein said extracting portion and said replacing portion are arranged to be operationally alternative to each other depending on an indication found by said examining portion.

25. A receiver apparatus according to claim 24, wherein:

the primary decoder is adapted to decode a linear-predictive-coded input signal in the primary frequency band and to generate a primary frequency band excitation signal,

the secondary decoder is adapted to decode a linear-predictive-coded input signal in a secondary frequency band and to use a secondary frequency band excitation signal,

said extracting portion is adapted to extract the primary frequency band excitation signal and to deliver the primary frequency band excitation signal or a derivative thereof to the secondary decoder as the secondary frequency band excitation signal, and

said replacing portion is adapted to generate the secondary frequency band excitation signal independently of the primary frequency band excitation signal.

26. A receiver apparatus according to claim 25, wherein said replacing portion is adapted to generate a random excitation signal.

27. A receiver apparatus according to claim 25, wherein said replacing portion is adapted to examine whether the input signal contains an indication about periodicity in the secondary frequency band, and depending on the results of such examining to generate a periodic excitation signal, if the input signal contains an indication about periodicity in the secondary frequency band, or to generate a random excitation signal, if the input signal does not contain any indication about periodicity in the secondary frequency band.

28. A receiver apparatus according to claim 24, wherein said extracting portion comprises a resampler adapted to resample the primary frequency band excitation signal and to deliver the resampled primary frequency band excitation signal for use as the secondary frequency band excitation signal.

29. A transmitter apparatus for transmitting digitally encoded sound, comprising:

means for splitting an input signal into a primary frequency band and at least one secondary frequency band,

means for digitally encoding the part of the input signal in the primary frequency band,

means for digitally encoding the part of the input signal in a secondary frequency band,

means for examining certain characteristics of the input signal in the primary frequency band and corresponding characteristics of the input signal in at least one secondary frequency band and indicating whether there is certain resemblance therebetween,

means for extracting certain characteristic features of a process applied to encoding the input signal in the primary frequency band, for use in a process applied to encoding the input signal in the secondary frequency band, and

means for replacing such extracted characteristic features with a locally generated, independent set of corresponding features in the process applied to encoding the input signal in the secondary frequency band;

wherein said means for extracting and said means for replacing are arranged to be operationally alternative to each other depending on an indication produced by said examining means.