Multi-Rate Audio Processing With Real-Time Phase Compensation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio processing techniques at high sampling rates require significant computational resources and introduce amplitude and phase changes, especially when processing the lowest frequency subband, which are difficult to correct due to reliance on a priori knowledge or offline measurements.

Innovation Solution

A multi-rate arbitrary audio processing system that decimates high sampling rate signals into subbands, applies processing only to the lowest frequency subband, and uses real-time amplitude and phase measurement and compensation methods to correct for changes in the transition band, reducing computational complexity and aliasing noise.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If audio processing is applied to the lowest frequency subband to reduce computational complexity, then computational resources are reduced, but amplitude and phase changes are introduced that are difficult to correct

Engineering Contradiction:
Improvecomputational resourcesVSAvoidamplitude and phase accuracy
Core Design Contradiction:
Use of energy by moving objectVSManufacturing precision

Solution Approach 1:

The audio signal is divided into multiple frequency subbands, with processing applied only to the lowest frequency subband. This segmentation allows computational resources to be reduced while isolating the processing effects to a specific frequency range that can be compensated for in other subbands.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system changes the sampling rate parameter for different subbands, using a lower sampling rate for the lowest frequency subband where processing is applied, while maintaining higher sampling rates for other subbands. This parameter variation optimizes computational efficiency while preserving audio quality through selective processing.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If existing compensation filters are used that rely on a priori knowledge or offline measurements, then amplitude and phase changes can be corrected, but the system cannot adapt to arbitrary or time-varying audio processing

Engineering Contradiction:
Improveamplitude and phase correctionVSAvoidadaptability to arbitrary processing
Core Design Contradiction:
Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The system employs a feedback mechanism where the output of the audio processing is fed back through analysis filters to measure the actual amplitude and phase changes. This real-time feedback allows the system to adapt to any type of audio processing (linear, non-linear, time-variant) without requiring prior knowledge of the processing characteristics.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The compensation system is designed to be universal and adaptable to any type of audio processing by using measured amplitude and phase changes from the actual processing output. Rather than requiring specific compensation filters for each processing type, the system universally applies measured corrections that work for arbitrary, linear, non-linear, and time-variant processing.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Manufacturing precision

If high sampling rate processing is used to preserve full bandwidth, then audio quality is maintained, but computational complexity increases significantly

Engineering Contradiction:
Improveaudio quality and bandwidth preservationVSAvoidcomputational complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The audio signal is segmented into multiple frequency subbands, each processed at appropriate sampling rates. This allows the system to preserve full bandwidth audio quality while reducing computational complexity by processing different subbands at different rates rather than processing the entire spectrum at the highest sampling rate.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different sampling rates and processing qualities are applied to different frequency subbands based on their specific requirements. The lowest frequency subband receives processing at a lower sampling rate with subsequent compensation, while other subbands maintain higher sampling rates, optimizing the balance between audio quality and computational complexity locally for each subband.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3257173B1Multi-rate system for audio processing
Publication Date: 2019.04.10 DTS INC(US)
  • EP3257173B1 patent drawingFigure 1
  • EP3257173B1 patent drawingFigure 2
  • EP3257173B1 patent drawingFigure 3~5

AI summary

A multi-rate audio processing system and method provides real-time measurement and processing of amplitude/phase changes in the transition band of the lowest frequency subband caused by the audio processing that can be used to apply amplitude/phase compensation to the higher subband(s). Tone signals may be injected into the transition band to provide strong tonal content for measurement and processing. The real-time measurement and compensation adapts to time- varying amplitude/phase changes regardless of the source of the change (e.g. non-linear time-varying linear or user control parameters) and provides universal applicability for any linear audio processing.