# US 9,437,204 B2 · Transform encoding/decoding of harmonic audio signals > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 9,437,204 B2 - Document ID: US9437204B2 - Status: Granted - Kind: B2 - Application number: 14/387,367 - Filed: 2012-10-30 - Published: 2015-02-12 - Granted: 2016-09-06 - Earliest claimed: 2012-03-29 - Assignees: Telefonaktiebolaget LM Ericsson (publ) - Applicants: Telefonaktiebolaget L M Ericsson (publ) - Inventors: Volodya Grancharov; Tomas Jansson Toftgård; Sebastian Näslund; Harald Pobloth - Web page: https://patentgrep.com/patent/US9437204B2 ## CPC - G10L 19/0212: {using orthogonal transformation} - G10L 19/002: Dynamic bit allocation (for perceptual audio coders G10L19/032) - G10L 19/028: Noise substitution, i.e. substituting non-tonal spectral components by noisy source (comfort noise for discontinuous speech transmission G10L19/012) - G10L 19/038: Vector quantisation, e.g. TwinVQ audio - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING ## Abstract An encoder (20) for encoding frequency transform coefficients (Y(k)) of a harmonic audio signal include the following elements: A peak locator (22) configured to locate spectral peaks having magnitudes exceeding a predetermined frequency dependent threshold. A peak region encoder (24) configured to encode peak regions including and surrounding the located peaks. A low-frequency set encoder (26) configured to encode at least one low-frequency set of coefficients outside the peak regions and below a crossover frequency that depends on the number of bits used to encode the peak regions. A noise-floor gain encoder (28) configured to encode a noise-floor gain of at least one high-frequency set of not yet encoded coefficients outside the peak regions. ## Claims ### Claim 1 A method of encoding a frequency transformed harmonic audio signal, comprising: receiving the frequency transformed harmonic audio signal; generating an encoded frequency transformed harmonic audio signal corresponding to the frequency transformed harmonic audio signal, based on: locating spectral peaks in the frequency transformed harmonic audio signal that have magnitudes exceeding a predetermined frequency dependent threshold; encoding peak regions including and surrounding the located spectral peaks; encoding at least one low-frequency set of Modified Discrete Cosine Transform (MDCT) coefficients outside the peak regions and below a crossover frequency that depends on a number of bits used to encode the peak regions; encoding a noise-floor gain of at least one high-frequency set of not yet encoded MDCT coefficients outside the peak regions; and outputting the encoded frequency transformed harmonic audio signal. ### Claim 2 (depends on 1) The encoding method of claim 1, wherein a peak region is encoded by: encoding spectrum position and sign of a peak; quantizing peak gain; encoding the quantized peak gain; scaling predetermined frequency bins surrounding the peak by the inverse of the quantized peak gain; and shape encoding the scaled frequency bins. ### Claim 3 (depends on 1) The encoding method of claim 1, wherein encoding a low-frequency set of MDCT coefficients includes encoding the low-frequency set based on a gain-shape encoding scheme. ### Claim 4 (depends on 3) The encoding method of claim 3, wherein the gain-shape encoding scheme is based on scalar gain quantization and factorial pulse shape encoding. ### Claim 5 (depends on 1) The encoding method of claim 1, comprising encoding a noise-floor gain for each of two high-frequency sets. ### Claim 6 A method of audio signal reconstruction comprising: receiving an encoded frequency transformed harmonic audio signal; decoding the encoded frequency transformed harmonic audio signal and thereby obtaining a reconstructed frequency transformed harmonic audio signal, based on: decoding spectral peak regions of the encoded frequency transformed harmonic audio signal, said spectral peak regions comprising spectral peaks having magnitudes exceeding a predetermined frequency dependent threshold; decoding at least one low-frequency set of Modified Discrete Cosine Transform (MDCT) coefficients of the encoded frequency transformed harmonic audio signal; distributing the MDCT coefficients of each low-frequency set outside the spectral peak regions and below a crossover frequency that depends on a number of bits used to encode the peak regions; decoding a noise-floor gain of at least one high-frequency set of MDCT coefficients of the encoded frequency transformed harmonic audio signal that are outside of the spectral peak regions; filling each high-frequency set of MDCT coefficients with noise having the corresponding decoded noise-floor gain; and outputting the reconstructed frequency transform harmonic audio signal. ### Claim 7 (depends on 6) The reconstruction method of claim 6, wherein a peak region is decoded by: decoding spectrum position and sign of a peak; decoding peak gain; decoding a shape of predetermined frequency bins surrounding the peak; and scaling the decoded shape by the decoded peak gain. ### Claim 8 (depends on 6) The reconstruction method of claim 6, wherein decoding a low-frequency set includes decoding the low-frequency set based on a gain-shape decoding scheme. ### Claim 9 (depends on 8) The reconstruction method of claim 8, wherein the gain-shape decoding scheme is based on scalar gain decoding and factorial pulse shape decoding. ### Claim 10 (depends on 6) The reconstruction method of claim 6, comprising decoding a noise-floor gain for each of two high-frequency sets. ### Claim 11 An encoder for encoding a frequency transformed harmonic audio signal, said encoder configured to obtain the frequency transformed harmonic audio signal and comprising a processing circuit configured to: generate an encoded frequency transformed harmonic audio signal corresponding to the frequency transformed harmonic audio signal, based on being configured to: locate spectral peaks in the frequency transformed harmonic audio signal that have magnitudes exceeding a predetermined frequency dependent threshold; encode peak regions including and surrounding the located spectral peaks; encode at least one low-frequency set of Modified Discrete Cosine Transform (MDCT) coefficients outside the peak regions and below a crossover frequency that depends on a number of bits used to encode the peak regions; and encode a noise-floor gain of at least one high-frequency set of not yet encoded MDCT coefficients outside the peak regions; and output the encoded frequency transformed harmonic audio signal. ### Claim 12 (depends on 11) The encoder of claim 11, wherein the processing circuit is configured to: encode a spectrum position and sign of a peak; quantize peak gain and encode the quantized peak gain; scale predetermined frequency bins surrounding the peak by the inverse of the quantized peak gain; and shape encode the scaled frequency bins. ### Claim 13 (depends on 11) A user equipment (UE) comprising the encoder of claim 11, said encoder configured to output the encoded frequency transformed harmonic audio signal to radio circuitry of the UE, for transmission to a remote receiver. ### Claim 14 A decoder configured for audio signal reconstruction, said decoder configured to receive an encoded frequency transformed harmonic audio signal and comprising a processing circuit configured to: decode the encoded frequency transformed harmonic audio signal and thereby obtain a reconstructed frequency transformed harmonic audio signal, based on being configured to: decode spectral peak regions of the encoded frequency transformed harmonic audio signal, said spectral peak regions including spectral peaks having magnitudes exceeding a predetermined frequency dependent threshold; decode at least one low-frequency set of Modified Discrete Cosine Transform (MDCT) coefficients; distribute the MDCT coefficients of each low-frequency set outside the spectral peak regions and below a crossover frequency that depends on a number of bits used to encode the peak regions; decode a noise-floor gain of at least one high-frequency set of MDCT coefficients outside of the spectral peak regions; and fill each high-frequency set of MDCT coefficients with noise having the corresponding noise-floor gain; and output the reconstructed frequency transformed harmonic audio signal. ### Claim 15 (depends on 14) The decoder of claim 14, wherein the processing circuit is configured to: decode spectrum position and sign of a peak; decode peak gain; decode a shape of predetermined frequency bins surrounding the peak; and scale the decoded shape by the decoded peak gain. ### Claim 16 (depends on 14) A user equipment (UE) comprising the decoder of claim 14, said decoder configured to output the reconstructed transformed harmonic audio signal to further audio signal processing circuitry of the UE, for generating a corresponding audio signal.