Selective Harmonic Prediction in Transform-Domain Audio Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio coding methods, such as frequency-domain prediction (FDP), suffer from high computational complexity and limited prediction gain due to the need for backward adaptivity and prediction of noisy components, leading to reconstruction errors.

Innovation Solution

Implement a predictive encoding and decoding scheme that selectively applies prediction to spectral coefficients based on a spacing value, such as the fundamental frequency or its multiples, reducing complexity by applying it only to harmonic signal components, and transmitting prediction errors rather than coefficients.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If predictive encoding is applied to all spectral coefficients, then coding efficiency is improved, but computational complexity increases significantly

Engineering Contradiction:
Improvecoding efficiencyVSAvoidcomputational complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent divides the spectrum into multiple bands and applies predictive encoding selectively to different bands based on their characteristics. This segmentation allows the system to apply complex predictive encoding only where necessary (in bands with high correlation) while using simpler methods in other bands, thereby resolving the contradiction between coding efficiency and computational complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements different encoding strategies for different spectral regions. By analyzing the correlation properties of each band and applying predictive encoding locally only where it provides significant benefit, the system achieves high coding efficiency in critical regions while maintaining overall computational feasibility.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If complex predictive encoding is applied to improve audio quality, then perceptual quality is improved, but processing time increases

Engineering Contradiction:
Improveaudio qualityVSAvoidprocessing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary analysis of spectral correlation properties before applying predictive encoding. By pre-identifying bands that benefit from predictive encoding and preparing prediction models in advance, the system reduces actual encoding processing time while maintaining high audio quality where it matters most.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent applies predictive encoding selectively to only those spectral bands where it provides perceptible quality improvement, rather than applying it uniformly across the entire spectrum. This partial application approach achieves high audio quality in critical regions while significantly reducing overall processing time.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4297027B1Audio encoder, audio decoder, method for encoding an audio signal and method for decoding an encoded audio signal
Publication Date: 2026.05.06 FRAUNHOFER GESELLSCHAFT ZUR FORDERUNG DER ANGEWANDTEN FORSCHUNG EV
  • EP4297027B1 patent drawingFigure 1
  • EP4297027B1 patent drawingFigure 2
  • EP4297027B1 patent drawingFigure 3

AI summary

An encoder for encoding an audio signal. The encoder is configured to encode the audio signal in a transform domain or filter-bank domain, wherein the encoder is configured to determine spectral coefficients of the audio signal for a current frame and at least one previous frame, wherein the encoder is configured to selectively apply predictive encoding to a plurality of individual spectral coefficients or groups of spectral coefficients which are separated by at least one spectral coefficient.