Concept for encoding an audio signal and decoding an audio signal using speech related spectral shaping information

• US 10,373,625 B2
• Filed: 04/18/2016
• Issued: 08/06/2019
• Est. Priority Date: 10/18/2013
• Status: Active Grant

• Current Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V.
• Original Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V.
• Inventors: Fuchs, Guillaume, Multrus, Markus, Ravelli, Emmanuel, Schnell, Markus
• Primary Examiner(s): Neway, Samuel G
• Application Number:
  US15/131,681
• Publication Number:
  US 20160232909A1
• Time in Patent Office: 1,205 Days
• CPC Class Codes: G10L 19/032 Quantisation or dequantisation of spectral components
• CPC Class Codes: G10L 19/06 Determination or coding of the spectral characteristics, e.g. of the short-term prediction coefficients
• CPC Class Codes: G10L 19/07 Line spectrum pair [LSP] vocoders
• CPC Class Codes: G10L 19/083 the excitation function being an excitation gain G10L25/90 takes precedence
• CPC Class Codes: G10L 19/12 the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders
• CPC Class Codes: G10L 19/20 using sound class specific coding, hybrid encoders or object based coding
• CPC Class Codes: G10L 2019/0016 Codebook for LPC parameters
• CPC Class Codes: G10L 25/93 Discriminating between voiced and unvoiced parts of speech signals G10L25/90 takes precedence

Abstract

According to an aspect of the present invention an encoder for encoding an audio signal has an analyzer configured for deriving prediction coefficients and a residual signal from a frame of the audio signal. The encoder has a formant information calculator configured for calculating a speech related spectral shaping information from the prediction coefficients, a gain parameter calculator configured for calculating a gain parameter from an unvoiced residual signal and the spectral shaping information and a bitstream former configured for forming an output signal based on an information related to a voiced signal frame, the gain parameter or a quantized gain parameter and the prediction coefficients.

Claims

1. An encoder for encoding an audio signal, the encoder comprising

an analyzer configured for deriving prediction coefficients and a residual signal from a frame of the audio signal;

a formant information calculator configured for calculating a speech related spectral shaping information from the prediction coefficients;

a gain parameter calculator configured for calculating a gain parameter from an unvoiced residual signal and the spectral shaping information; and

a bitstream former configured for forming an output signal based on an information related to a voiced signal frame, the gain parameter or a quantized gain parameter and the prediction coefficients;

wherein the gain parameter calculator comprises a noise generator configured for generating an encoding noise-like signal; and

wherein the gain parameter calculator comprises a shaping processor configured for shaping a spectrum of the encoding noise-like signal using the speech related spectral shaping information and a variable amplifier configured for amplifying the spectrally shaped encoding noise-like signal to obtain an amplified shaped encoding noise-like signal and a controller for calculating the gain parameter based on the amplified shaped encoding noise-like signal;

wherein one or more of the analyzer, the formant information calculator, the gain parameter calculator and the bitstream former is implemented, at least in part, by one or more hardware elements of the encoder;

wherein the shaping processor is configured for combining a spectrum of the encoding noise-like signal or a spectrum derived thereof and a transfer function comprising

2. The encoder according to claim 1, further comprising a decider configured for determining if the residual signal was determined from an unvoiced signal audio frame.

3. The encoder according to claim 1, wherein the gain parameter calculator is configured to:

use the gain parameter as temporary gain parameter to acquire the amplified shaped encoding noise-like signal;

wherein the gain parameter calculator comprises a comparer configured for comparing the unvoiced residual signal and the amplified shaped encoding noise-like signal to acquire a measure for a likeness between the unvoiced residual signal and the amplified shaped encoding noise-like signal; and

wherein the controller is configured for determining the gain parameter and to adapt the temporary gain parameter based on the comparison result;

wherein the controller is configured to provide the gain parameter to the bitstream former, when a value of the measure for the likeness is above a threshold value.

4. The encoder according to claim 1, wherein the gain parameter calculator is configured to:

use the gain parameter as temporary gain parameter to acquire an amplified shaped encoding noise-like signal;

wherein the gain parameter calculator comprises a synthesizer configured for synthesizing a synthesized signal from the amplified shaped encoding noise-like signal and the prediction coefficients and to provide the synthesized signal;

wherein the gain parameter calculator comprises a comparer configured for comparing the audio signal and the synthesized signal to acquire a measure for a likeness between the audio signal and the synthesized signal; and

wherein the controller is configured for determining the gain parameter and to adapt the temporary gain parameter based on the comparison result;

wherein the controller is configured to provide the gain parameter to the bitstream former, when a value of the measure for the likeness is above a threshold value.

5. The encoder according to claim 1, further comprising a gain memory configured for recording an encoding information comprising the gain parameter or an information ĝ

_n related thereto, wherein the controller is configured to record the encoding information during processing of the audio frame and for determining the gain parameter for a subsequent frame of the audio signal based on the encoding information of the preceding frame of the audio signal.

6. The encoder according to claim 1, wherein the noise generator is configured for generating a plurality of random signals and to combine the plurality of random signals to acquire the encoding noise-like signal.

7. The encoder according to claim 1, further comprising a quantizer configured for receiving the gain parameter, for quantizing the gain parameter to acquire the quantized gain parameter.

8. The encoder according to claim 1, wherein a shaper is configured for combining a spectrum of the encoding noise-like signal or a spectrum derived thereof with a transfer function comprising

Ft(z)=1−

β

z<sup>−

1</sup>

wherein z indicates a representation in the z-domain, wherein β

represents a measure (voicing) for a voicing determined by relating an energy of a past frame of the audio signal and an energy of a present frame of the audio signal, wherein the measure β

is determined in function of a voicing value.

16. Encoder according to claim 1, wherein the gain parameter calculator comprises a comparer configured for comparing the unvoiced residual signal and the amplified shaped encoding noise-like signal to obtain a comparison result, wherein the controller is configured for determining the gain parameter based on the comparison result.

9. A decoder for decoding a received signal comprising information related to prediction coefficients, the decoder comprising

a formant information calculator configured for calculating a speech related spectral shaping information from the prediction coefficients;

a noise generator configured for generating a decoding noise-like signal;

a shaper configured for shaping a spectrum of the decoding noise-like signal using the spectral shaping information to acquire a shaped spectrum of the decoding noise-like signal;

a synthesizer configured for synthesizing a synthesized signal from the shaped spectrum of the decoding noise-like signal and the prediction coefficients; and

a variable amplifier configured for receiving a gain parameter and for amplifying the shaped spectrum of the decoding noise-like signal to obtain an amplified shaped decoding noise-like signal;

wherein one or more of the formant information calculator, the noise generator, the shaper and the synthesizer is implemented, at least in part, by one or more hardware elements of the decoder;

wherein the shaper is configured for combining a spectrum of the decoding noise-like signal or a spectrum derived thereof and a transfer function comprising

10. The decoder according to claim 9, wherein the received signal comprises an information related to a gain parameter and wherein the shaper comprises an amplifier configured for amplifying the decoding noise-like signal or the shaped decoding noise-like signal.

11. The decoder according to claim 9, wherein the received signal further comprises a voiced information related to a voiced frame of an encoded audio signal and wherein the decoder further comprises a voiced frame processor configured for determining a voiced signal based on the voiced information, wherein the decoder further comprises a combiner configured for combining the synthesized signal and the voiced signal to acquire a frame of an audio signal sequence.

17. Decoder according to claim 9, comprising a signal generator configured to generate a code excited excitation signal using the prediction coefficients and comprising a further shaper configured for shaping the code excited excitation signal using the speech related shaping information and for amplifying the spectrum of the shaped code excited excitation signal to obtain an amplified shaped code excited excitation signal.

18. Decoder according to claim 17, wherein the formant information calculator is configured to provide different speech related shaping information to the shaper and to the further shaper.

12. A method for encoding an audio signal, comprising

deriving, using an analyzer, prediction coefficients and a residual signal from an audio signal frame;

calculating, using a formant information calculator, a speech related spectral shaping information from the prediction coefficients;

calculating, using a gain parameter calculator, a gain parameter from an unvoiced residual signal and the spectral shaping information; and

forming, using a bitstream former, an output signal based on an information related to a voiced signal frame, the gain parameter or a quantized gain parameter and the prediction coefficients; and

generating an encoding noise-like signal, comprising;

shaping a spectrum of the encoding noise-like signal using the speech related spectral shaping information; and

amplifying the shaped spectrum of the encoding noise-like signal to obtain an amplified shaped encoding noise-like signal;

calculating a gain parameter based on the amplified shaped encoding noise-like signal;

wherein one or more of the analyzer, the formant information calculator, the gain parameter calculator and the bitstream former is implemented, at least in part, by one or more hardware elements;

combining a spectrum of the encoding noise-like signal or a spectrum derived thereof and a transfer function comprising

14. A non-transitory digital storage medium having stored thereon a computer program for performing a method for encoding an audio signal according to claim 12.

13. A method for decoding a received audio signal comprising an information related prediction coefficients and a gain parameter, the method comprising

calculating, using a formant information calculator, a speech related spectral shaping information from the prediction coefficients;

generating, using a noise generator, a decoding noise-like signal;

shaping, using a shaper, a spectrum of the decoding noise-like signal using the spectral shaping information to acquire a shaped decoding noise-like signal;

receiving a gain parameter and amplifying the shaped spectrum of the decoding noise-like signal with a variable amplifier, to obtain an amplified spectrum of the shaped noise signal; and

synthesizing, using a synthesizer, a synthesized signal from the amplified shaped decoding noise-like signal and the prediction coefficients;

wherein one or more of the formant information calculator, the noise generator, the shaper and the synthesizer is implemented, at least in part, by one or more hardware elements;

combining a spectrum of the decoding noise-like signal or a spectrum derived thereof and a transfer function comprising

15. A non-transitory digital storage medium having stored thereon a computer program for performing a method for decoding a received audio signal according to claim 13.

19. An encoder for encoding an audio signal, the encoder comprising:

an analyzer configured for deriving prediction coefficients and a residual signal from a frame of the audio signal;

a formant information calculator configured for calculating a speech related spectral shaping information from the prediction coefficients;

a gain parameter calculator configured for calculating a gain parameter from an unvoiced residual signal and the spectral shaping information; and

a bitstream former configured for forming an output signal based on an information related to a voiced signal frame, the gain parameter or a quantized gain parameter and the prediction coefficients;

wherein the gain parameter calculator comprises a noise generator configured for generating an encoding noise-like signal; and

wherein the gain parameter calculator comprises a shaping processor configured for spectral shaping the encoding noise-like signal using the speech related spectral shaping information and a variable amplifier configured for amplifying the spectrally shaped encoding noise-like signal to obtain an amplified shaped encoding noise-like signal and a controller for calculating the gain parameter based on the amplified shaped encoding noise-like signal;

wherein one or more of the analyzer, the formant information calculator, the gain parameter calculator and the bitstream former is implemented, at least in part, by one or more hardware elements of the encoder;

wherein the gain parameter calculator comprises a comparer configured for comparing the unvoiced residual signal and the amplified shaped encoding noise-like signal to obtain a comparison result, wherein the controller is configured for determining the gain parameter based on the comparison result.

20. A method for encoding an audio signal, comprising

deriving, using an analyzer, prediction coefficients and a residual signal from an audio signal frame;

calculating, using a formant information calculator, a speech related spectral shaping information from the prediction coefficients;

calculating, using a gain parameter calculator, a gain parameter from an unvoiced residual signal and the spectral shaping information; and

forming, using a bitstream former, an output signal based on an information related to a voiced signal frame, the gain parameter or a quantized gain parameter and the prediction coefficients;

generating an encoding noise-like signal, comprising;

shaping a spectrum of the encoding noise-like signal using the speech related spectral shaping information; and

amplifying the spectrally shaped encoding noise-like signal to obtain an amplified shaped encoding noise-like signal;

calculating a gain parameter based on the amplified shaped encoding noise-like signal;

wherein one or more of the analyzer, the formant information calculator, the gain parameter calculator and the bitstream former is implemented, at least in part, by one or more hardware elements;

comparing, using a comparer, the unvoiced residual signal and the amplified shaped encoding noise-like signal to obtain a comparison result, wherein the gain parameter is determined based on the comparison result.

21. A decoder for decoding a received signal comprising information related to prediction coefficients, the decoder comprising

a formant information calculator configured for calculating a speech related spectral shaping information from the prediction coefficients;

a noise generator configured for generating a decoding noise-like signal;

a shaper configured for shaping a spectrum of the decoding noise-like signal using the spectral shaping information to acquire a spectrally shaped decoding noise-like signal;

a synthesizer configured for synthesizing a synthesized signal from the spectrally shaped decoding noise-like signal and the prediction coefficients; and

a variable amplifier configured for receiving a gain parameter and for amplifying the spectrally shaped decoding noise-like signal to obtain an amplified shaped decoding noise-like signal;

wherein one or more of the formant information calculator, the noise generator, the shaper and the synthesizer is implemented, at least in part, by one or more hardware elements of the decoder; and

wherein the gain parameter is generated by a gain parameter calculator comprising;

a comparer configured for comparing the unvoiced residual signal and the amplified shaped decoding noise-like signal to obtain a comparison result, wherein the gain parameter is determined based on the comparison result.

22. A method for decoding a received audio signal comprising an information related prediction coefficients and a gain parameter, the method comprising

calculating, using a formant information calculator, a speech related spectral shaping information from the prediction coefficients;

generating, using a noise generator, a decoding noise-like signal;

shaping, using a shaper, a spectrum of the decoding noise-like signal using the spectral shaping information to acquire a spectrally shaped decoding noise-like signal;

receiving a gain parameter and amplifying the spectrally shaped decoding noise-like signal with a variable amplifier, to obtain an amplified shaped decoding noise signal; and

synthesizing, using a synthesizer, a synthesized signal from the amplified shaped decoding noise-like signal and the prediction coefficients;

wherein one or more of the formant information calculator, the noise generator, the shaper and the synthesizer is implemented, at least in part, by one or more hardware elements; and

wherein the gain parameter is determined by comparing, using a comparer, the unvoiced residual signal and the amplified shaped decoding noise-like signal to obtain a comparison result, wherein the gain parameter is determined based on the comparison result.