Communication Device Spatial Source Separation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Multi-talker communication systems fail to effectively separate and position audio sources spatially, leading to loss of directional cues and difficulty in distinguishing between different audio signals.

Innovation Solution

A communication device with a processor that combines input signals with complementing signals and applies head-related transfer functions to generate left and right output signals, enhancing spatial separation by extending the effective bandwidth and adding binaural cues.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multiple audio signals are mixed into a single signal for presentation over headphones, then the communication system can transmit multiple talkers, but the listeners lose the ability to distinguish spatial positions and separate different sources

Engineering Contradiction:
Improvemulti-talker communication capabilityVSAvoidspatial position information
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent segments the mixed audio signal into multiple independent audio channels, each corresponding to a different talker or sound source. By applying individual spatial processing to each channel rather than processing the mixed signal as a whole, the system preserves the ability to distinguish spatial positions while maintaining multi-talker communication capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different spatial processing characteristics to different audio channels based on their respective spatial positions. Each audio channel receives customized spatial cues (such as different impulse responses or transfer functions) that correspond to its specific spatial location, enabling listeners to perceive the spatial distribution of multiple talkers.

Inventive Principle:
Principle #3Local quality

2Loss of information

If audio signals are processed with spatial separation, then listeners can distinguish different sources, but the system complexity increases due to multiple processing channels

Engineering Contradiction:
Improvesource distinction capabilityVSAvoidsignal processing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent performs preliminary spatial processing on audio channels during the encoding or transmission phase, preparing spatial cues in advance. This allows the spatial separation functionality to be implemented without adding significant complexity to the real-time decoding and playback system, as the spatial processing framework is established beforehand.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements a universal spatial processing framework that can handle multiple audio channels simultaneously using the same processing architecture. By designing the system to process N audio channels through a unified approach rather than requiring separate processing paths for each channel, the patent reduces overall system complexity while maintaining source distinction capability.

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

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Improves directionality and separation of audio sources, allowing users to better distinguish between multiple audio signals by providing more detailed binaural cues and accurate spatial positioning.

Implementation Method 1

a first noise generator for generating one or more noise signals

Methodology Applied
Scientific EffectNoise generation:

Implementation Method 2

apply a first filter function to the first combined signal for provision of a first left output signal and a first right output signal, the first filter function comprising one or more head related transfer functions

Methodology Applied
Scientific EffectHead-related transfer function filtering:

Data Source

PatentEP3585076B1Communication device with spatial source separation, communication system, and related method
Publication Date: 2023.12.27 FALCOM AS
  • EP3585076B1 patent drawingFigure 1~3
  • EP3585076B1 patent drawingFigure 4
  • EP3585076B1 patent drawingFigure 5

AI summary

A communication system, a communication device and related method, the communication device comprising a processor, a source interface comprising a first source interface for a first external source, and an output interface, wherein the processor is configured to obtain, e.g. receive, a first input signal via the first source interface, obtain a first complementing signal, combine the first input signal and the first complementing signal to a first combined signal, apply a first filter function to the first combined signal for provision of a first left output signal and a first right output signal, the first filter function comprising one or more head related transfer functions, and output a left output signal and a right output signal via the output interface, wherein the left output signal is based on the first left output signal and the right output signal is based on the first right output signal.