Per-Source Audio Rendering for Speech Intelligibility

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio rendering technologies do not effectively customize the listening experience to enhance speech intelligibility or accessibility for users with hearing impairments, particularly by managing the ratio of indirect audio to direct audio based on the nature and position of the audio source.

Innovation Solution

An apparatus and method that dynamically adjust the ratio of indirect audio to direct audio for each audio source, prioritizing speech audio sources with a first reduction in this ratio when active, and non-speech sources with a lesser or no reduction, based on metadata and user position, using reverb attenuation to improve speech intelligibility and accessibility.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a uniform reduction in the ratio of indirect audio to direct audio is applied to all audio sources, then speech intelligibility is improved, but the listening experience for non-speech audio sources deteriorates

Engineering Contradiction:
Improvespeech intelligibilityVSAvoidlistening experience for non-speech sources
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent applies different audio rendering treatments to different audio sources based on their type. Speech audio sources receive a first reduction in the indirect-to-direct audio ratio, while non-speech audio sources receive a second reduction (less than or equal to the first). This localized differentiation ensures speech intelligibility is prioritized without compromising the listening experience for other audio types.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system dynamically adjusts the indirect-to-direct audio ratio reduction based on the identified type of each audio source. The audio processing apparatus identifies whether each source is speech or non-speech and applies corresponding dynamic rendering parameters, allowing the system to adapt its audio processing in real-time according to the content being played.

Inventive Principle:
Principle #15Dynamics

2Reliability

If indirect audio is reduced for all sources to improve accessibility, then speech accessibility for hearing-impaired users is improved, but the overall audio quality and spatial realism deteriorates

Engineering Contradiction:
Improveaudio accessibilityVSAvoidaudio quality and spatial realism
Core Design Contradiction:
ReliabilityVSEase of manufacture

Solution Approach 1:

The patent implements accessibility enhancements selectively for speech audio sources only, applying a first reduction in indirect-to-direct ratio specifically to speech. Non-speech audio sources receive a different treatment (second reduction), which maintains their spatial realism and audio quality while still providing some accessibility benefit. This localized approach ensures hearing-impaired users can understand speech clearly without sacrificing overall audio fidelity.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system changes the audio rendering parameters (indirect-to-direct ratio reduction levels) based on the identified characteristics of each audio source. By modifying these parameters selectively - applying stronger reduction to speech sources for accessibility while applying milder or no reduction to non-speech sources for quality preservation - the system achieves both accessibility and audio quality goals simultaneously.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4657896A1Customized audio rendering
Publication Date: 2025.12.03 NOKIA TECHNOLOGIES OY
  • EP4657896A1 patent drawingFigure 1~3
  • EP4657896A1 patent drawingFigure 4A~5
  • EP4657896A1 patent drawingFigure 6~8

AI summary

An apparatus comprising means for: providing an audio intelligibility mode or audio accessibility mode in which a ratio of indirect audio to direct audio for an audio source is reduced compared to another mode on a per audio source basis wherein a priority audio source is rendered: with a first reduction in the ratio of indirect audio to direct audio, if it is a speech audio source; and with a second reduction, or no reduction, in the ratio of indirect audio to direct audio, if it is not a speech audio source, wherein the second reduction is less than the first reduction.