Audio Watermark Encoding for Compression Robustness

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio watermarking technologies are not robust to audio compression and format changes, leading to errors and unusability of embedded information in compressed audio files like MP3 formats.

Innovation Solution

A system and method that generate an audio signal by superposing an imperceptible audio watermark with audio data at the playback site, using a frequency range above human hearing capabilities, ensuring the additional information is not corrupted by audio data manipulations and allowing flexible customization of information for different playback systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If audio watermarking is used to embed additional information in audio data, then identification and ordering information can be provided to listeners, but the watermark becomes vulnerable to corruption during audio compression and format changes

Engineering Contradiction:
Improveintegrity of embedded informationVSAvoidrobustness of watermark to compression
Core Design Contradiction:
Loss of informationVSReliability

Solution Approach 1:

The patent applies preliminary action by embedding the audio watermark into the uncompressed audio data at the source before any compression or format conversion occurs. This ensures the watermark is established in its most robust form, allowing it to survive subsequent processing steps that would otherwise corrupt watermarks embedded later in the chain.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system segments the information delivery process by separating the watermark embedding function from the audio processing functions. The watermark is embedded independently in the original audio data, then the audio undergoes compression and format conversion. This segmentation protects the watermark from being corrupted by these subsequent processing steps.

Inventive Principle:
Principle #1Segmentation

2Loss of information

If additional information is embedded in audio data using traditional watermarking, then content identification is possible, but the embedding process becomes complex and computationally intensive

Engineering Contradiction:
Improvepreservation of additional informationVSAvoidcomplexity of watermarking system
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent uses copying by creating a separate audio watermark data structure that mirrors the essential information needed for identification and ordering. Rather than complexly embedding data directly into the audio waveform, the system generates a parallel watermark representation that can be independently processed and is simpler to implement while achieving the same information preservation goal.

Inventive Principle:
Principle #26Copying

3Quantity of substance

If audio watermarks are embedded in compressed audio files, then data volume is reduced for easy distribution, but the watermarks become unreadable or error-prone

Engineering Contradiction:
Improvedata volume of audio fileVSAvoidaccuracy of watermark detection
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The system performs the watermark embedding action preliminarily, before compression reduces the data volume. By establishing the watermark in the full-resolution audio data first, the system ensures accurate embedding with sufficient bit depth and dynamic range. The subsequent compression then preserves this pre-established watermark rather than attempting to embed it in the already-compressed, lower-precision data.

Inventive Principle:
Principle #10Preliminary action

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

Ensures reliable and imperceptible embedding of additional information in audio signals, maintaining data integrity and flexibility across various playback systems, with the ability to provide identification and ordering information to listeners.

Implementation Method 1

An audio watermark is a data signal within an audio signal that is essentially inaudible due to the psycho-acoustic properties of the human hearing

Methodology Applied
Scientific EffectPsycho-acoustic properties of human hearing:

Implementation Method 2

The encoding component is configured to generate a first intermediate audio signal from the additional information and a second intermediate audio signal from the audio data and to superpose the second intermediate audio signal with the first intermediate audio signal

Methodology Applied
Scientific EffectFrequency modulation: Phase Modulation

Data Source

PatentEP1973103B1System and method for providing additional information related to an audio piece
Publication Date: 2017.08.30 VODAFONE HOLDING GMBH
  • EP1973103B1 patent drawingFigure 1

AI summary

The invention relates to a playback system (20) for playing back an audio piece at a playback site and for providing additional information related to the audio piece. The proposed playback system (20) comprises: - an acquisition component (40) for acquiring audio data (10) of the audio piece, - a means (40, 55) for acquiring the additional information related to the audio piece, - an encoding component (60), which is located at the playback site of the audio piece and which is configured to combine the audio data and the additional information and to generate an audio signal containing the audio piece and the additional information, and - a playback component (80), which is connected to the encoding component (60) and which is configured to generate an acoustic signal (75) from the audio data. The invention further relates to a method for playing back an audio piece at a playback site and for providing additional information related to the audio piece.