Granted Patent B2

US 12,725,620 · App. 18/448,020 · Granted Sep 1, 2026

Inventors: Sascha Disch (Fürth, DE); Martin Dietz (Nuremberg, DE); Markus Multrus (Nuremberg, DE); Guillaume Fuchs (Bubenreuth, DE); Emmanuel Ravelli (Erlangen, DE); Matthias Neusinger (Rohr, DE); Markus Schnell (Nuremberg, DE); Benjamin Schubert (Nuremberg, DE); Bernhard Grill (Rückersdorf, DE)

Assignee: Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.

G10L19/0208G10L19/022G10L19/18G10L19/24G10L2019/0001G10L19/02G10L19/028G10L19/04G10L19/083G10L19/26G10L21/038

Abstract

Claims (32)

1 . An audio decoder for decoding an encoded audio signal to acquire a decoded audio signal, comprising:

a first decoding processor configured for decoding a first encoded audio signal portion in a frequency domain to obtain a decoded spectral representation of the first encoded audio signal portion, wherein the first decoding processor is configured to reconstruct a first set of first spectral portions in a waveform-preserving manner to generate a spectrum comprising a first set of reconstructed first spectral portion and at least one gap between two reconstructed first spectral portions of the first set of reconstructed first spectral portions, is configured to fill the at least one gap between two reconstructed first spectral portions of the first set of reconstructed first spectral portions in the spectrum using a frequency regeneration technique, the frequency regeneration technique being configured for applying parametric data to one or more reconstructed first spectral portions of the first set of reconstructed first spectral portions to obtain a decoded spectral representation of the first encoded audio signal portion, and

comprises a frequency-time converter configured for converting the decoded spectral representation of the first encoded audio signal portion into a time domain to acquire a decoded first audio signal portion;

a second decoding processor configured for decoding a second encoded audio signal portion in the time domain to acquire a decoded second audio signal portion;

a cross-processor configured

for calculating, from the decoded spectral representation of the first encoded audio signal portion, initialization data for the second decoding processor, and

for initializing the second decoding processor using the initialization data to decode the second encoded audio signal portion following in time the first encoded audio signal portion in the encoded audio signal; and

a combiner configured for combining the decoded first audio signal portion and the decoded second audio signal portion to acquire the decoded audio signal.

2 . The audio decoder of claim 1 , wherein the decoded spectral representation of the first encoded audio signal portion extends until a maximum frequency of a time representation of the decoded audio signal, a spectral value for the maximum frequency being zero or different from zero.

3 . The audio decoder of claim 1 , wherein the second decoding processor comprises at least one element of the group of elements comprising:

a stage configured for decoding ACELP gains and an innovative codebook;

an adaptive codebook synthesis stage;

an ACELP post-processor;

a prediction synthesis filter; and

a de-emphasis stage.

4 . A method of decoding an encoded audio signal to acquire a decoded audio signal, the method comprising:

decoding a first encoded audio signal portion in a frequency domain to obtain a decoded spectral representation of the first encoded audio signal portion, wherein the decoding the first encoded audio signal portion comprises

reconstructing a first set of first spectral portions in a waveform-preserving manner to generate a spectrum comprising a first set of reconstructed first spectral portions and at least one gap between two reconstructed first spectral portions of the first set of reconstructed first spectral portions,

filing at least one gap between the two reconstructed first spectral portions of the first set of reconstructed first spectral portions in the spectrum using a frequency regeneration technique, the frequency regeneration technique comprising applying parametric data to one or more reconstructed first spectral portions of the first set of reconstructed first spectral portions to obtain the decoded spectral representation of the first encoded audio signal portion, and

converting the decoded spectral representation of the first encoded audio signal portion into a time domain to acquire a decoded first audio signal portion,

decoding a second encoded audio signal portion in the time domain to acquire a decoded second audio signal portion;

calculating, from the decoded spectral representation of the first encoded audio signal portion, initialization data for the step of decoding the second encoded audio signal portion; initializing the step of decoding the second encoded audio signal portion using the initialization data to decode the second encoded audio signal portion following in time the first encoded audio signal portion in the encoded audio signal; and

combining the decoded first audio signal portion and the decoded second audio signal portion to acquire the decoded audio signal.

5 . A non-transitory digital storage medium having a computer program stored thereon to perform the method of decoding an encoded audio signal wherein said computer program is run by a computer to acquire a decoded audio signal, the method comprising:

decoding a first encoded audio signal portion in a frequency domain to obtain a decoded spectral representation of the first encoded audio signal portion, wherein the decoding the first encoded audio signal portion comprises

reconstructing a first set of first spectral portions in a waveform-preserving manner to generate a spectrum comprising a first set of reconstructed first spectral portions and at least one gap between two reconstructed first spectral portions of the first set of reconstructed first spectral portions,

filing the at least one gap between the two reconstructed first spectral portions of the first set of reconstructed first spectral portions in the spectrum using a frequency regeneration technique, the frequency regeneration technique comprising applying parametric data to one or more reconstructed first spectral portions of the first set of reconstructed first spectral portions to obtain the decoded spectral representation of the first encoded audio signal portion, and

converting the decoded spectral representation into a time domain to acquire a decoded first audio signal portion;

decoding a second encoded audio signal portion in the time domain to acquire a decoded second audio signal portion;

calculating, from the decoded spectral representation of the first encoded audio signal portion, initialization data for the step of decoding the second encoded audio signal portion;

initializing the step of decoding the second encoded audio signal portion using the initialization data to decode the second encoded audio signal portion following in time the first encoded audio signal portion in the encoded audio signal; and

combining the decoded first audio signal portion and the decoded second audio signal portion to acquire the decoded audio signal.

Assignments (1)

Priority Claims (1)

Continuity (5)

References Cited (305)