Audio Watermarking Down-Mix Compensation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Audio watermarks embedded in multichannel audio signals become perceptible when media devices down-mix channels, leading to reduced masking capability and potential audibility in the resulting stereo signal.

Innovation Solution

Implementing down-mixing compensation techniques, such as determining and applying attenuation factors and phase shifts to watermarks in specific audio channels, to reduce perceptibility during the down-mixing process, ensuring the watermark remains masked in the down-mixed audio.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If audio watermarks are embedded in multichannel audio signals, then media identification information can be carried for distribution monitoring, but the watermarks become perceptible when channels are down-mixed to stereo

Engineering Contradiction:
Improvewatermark detectabilityVSAvoidwatermark perceptibility
Core Design Contradiction:
Loss of informationVSObject-affected harmful factors

Solution Approach 1:

The patent applies different processing to different audio channels based on their down-mixing behavior. Specifically, it identifies which channels will be down-mixed and applies compensation only to those channels, leaving others unchanged. This localized approach ensures watermarks in down-mixed channels remain imperceptible while maintaining watermark integrity in channels that retain their original characteristics.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent modifies watermark parameters (amplitude/energy) based on the down-mixing characteristics of each channel. By calculating the energy ratio between the original multichannel signal and the down-mixed stereo signal, the system adjusts the watermark amplitude in each channel to compensate for the energy reduction that occurs during down-mixing, thereby maintaining constant watermark perceptibility across different playback configurations.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If watermark amplitude is increased to maintain detectability after down-mixing, then watermark robustness improves, but watermark perceptibility in the original multichannel signal increases

Engineering Contradiction:
Improvewatermark robustnessVSAvoidwatermark perceptibility
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent performs compensation in advance during the watermark embedding stage, before down-mixing occurs. By calculating the expected down-mixing energy ratio and pre-adjusting the watermark amplitude in each channel, the system ensures that after down-mixing, the watermark maintains its intended detectability without needing to increase amplitude beyond what is necessary for the specific channel's down-mixing behavior.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent dynamically adjusts watermark amplitude parameters based on channel-specific down-mixing characteristics. Channels that will be down-mixed receive higher amplitude watermarks to compensate for energy loss, while channels that retain their original characteristics receive lower amplitude watermarks. This parameter adaptation ensures robustness only where necessary, maintaining imperceptibility in the original multichannel signal.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If watermark is embedded in all channels, then detection reliability improves, but complexity of watermark embedding and processing increases

Engineering Contradiction:
Improvewatermark detection reliabilityVSAvoidwatermark embedding complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent divides the audio signal into separate channel groups based on their down-mixing behavior. Instead of treating all channels uniformly, it segments them into those that will be down-mixed and those that will not, applying different watermarking strategies to each segment. This segmentation reduces processing complexity by focusing computational resources only on channels that require compensation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies watermark compensation only to the extent necessary for down-mixed channels, rather than uniformly to all channels. By identifying and processing only the channels that will undergo down-mixing, the system achieves sufficient detection reliability for stereo playback without the unnecessary complexity of processing all channels equally.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2973553B1Down-mixing compensation for audio watermarking
Publication Date: 2023.11.29 THE NIELSEN CO (US) LLC
  • EP2973553B1 patent drawingFigure 1
  • EP2973553B1 patent drawingFigure 2~7
  • EP2973553B1 patent drawingFigure 3

AI summary

Example methods, apparatus, systems and articles of manufacture to implement down-mixing compensation for audio watermarking are disclosed. Example methods disclosed herein to compensate for audio channel down-mixing when embedding watermarks in a multichannel audio signal include obtaining a watermark to be embedded in respective ones of a plurality of audio channels of the multichannel audio signal. Such example methods also include embedding the watermark in a first one of the plurality of audio channels based on a compensation factor that is to reduce perceptibility of the watermark when the first one of the plurality of audio channels is down-mixed with a second one of the plurality of audio channels after the watermark has been applied to the first and second ones of the plurality of audio channels.