Compressed Audio Restoration Using Harmonic and Spatial Enhancement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Compressed audio signals often suffer from undesirable audible differences due to data compression, which discards perceptually less important components, leading to reduced quality, especially at lower bitrates.

Innovation Solution

The Signal Enhancer system modifies compressed audio signals by restoring and enhancing discarded or altered components, such as bandwidth, harmonics, transients, reverb, and masked signals, while phase-aligning harmonic components, operating on individual spatial slices and Mid-Side components of stereo signals.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data compression is applied to reduce audio signal size, then storage and transmission efficiency is improved, but audible quality deteriorates due to discarded signal components

Engineering Contradiction:
Improvedata compression efficiencyVSAvoidaudible signal components
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent applies the principle of discarding and recovering by first identifying and discarding inaudible signal components during compression, then recovering and reconstructing the discarded audible components through analysis of the compressed signal and synthesis of missing harmonics, transients, and spatial information. This allows the system to achieve compression while restoring the perceived audio quality.

Inventive Principle:
Principle #34Discarding and recovering

Solution Approach 2:

The patent uses an intermediary approach by introducing a perceptual model as a mediator between the compressed signal and the final output. This perceptual model analyzes the compressed signal to identify missing components and guides the reconstruction process, acting as an intermediary that translates compressed data back into perceptually accurate audio.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Quantity of substance

If perceptual audio codec discards less important components, then data rate is reduced, but perceived audio quality deteriorates

Engineering Contradiction:
Improvedata rateVSAvoidperceived audio quality
Core Design Contradiction:
Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies parameter changes by modifying the perceptual parameters of the audio signal during reconstruction. It adjusts parameters such as harmonic content, transient characteristics, and spatial distribution to match the original perceived quality while maintaining reduced data rates. The system changes parameters like bandwidth extension and reverb characteristics to restore quality.

Inventive Principle:
Principle #35Parameter changes

3Loss of information

If compression artifacts are reduced by maintaining more signal components, then audio quality is improved, but compression efficiency deteriorates

Engineering Contradiction:
Improvesignal component retentionVSAvoidcompression efficiency
Core Design Contradiction:
Loss of informationVSProductivity

Solution Approach 1:

The patent applies segmentation by dividing the audio signal into different perceptual components (harmonics, transients, reverb, spatial information) and processing each segment separately. This allows selective reconstruction of only the necessary components for perceived quality, maintaining compression efficiency while restoring audio fidelity through targeted component recovery.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS9576584B2System for perceived enhancement and restoration of compressed audio signals
Publication Date: 2017.02.21 HARMAN INT IND INC
  • US9576584B2 patent drawing
  • US9576584B2 patent drawing
  • US9576584B2 patent drawing

AI summary

A system for processing compressed audio includes a signal enhancer module configured to generate one or more signal treatments. The one or more signal treatments may be generated by the signal enhancer module based on analysis of the incoming audio signal. Alternatively, or in addition, characteristics of the incoming audio signal may be provided to the signal enhancer module for use in generating the one or more signal treatments. The one or more signal treatments may be added to file audio signals.