Audio Sample Audibility Estimation in Masked Broadcast Mixes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Measuring the audibility of audio samples in broadcast media programs is challenging due to their low volume or masking by other sounds, making it difficult to account for them in copyright purposes, especially in real-world broadcast production where pre-measurement is not common.
Innovation Solution
A method involving time-frequency representation of audio mixes and audio samples, with time-alignment and gain adjustment using spectral peaks, to generate a difference matrix, which is then processed using an audibility model to estimate the audibility of audio samples, accounting for psychoacoustic factors and masking effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If audio samples are mixed with other sounds in broadcast media programs, then the program becomes more engaging and complete, but the audibility of individual audio samples becomes difficult to measure and assess
Solution Approach 1:
The patent segments the audio mix into individual components by separating the audio sample signal from the additional audio signal. This is achieved through signal processing techniques that isolate the audio sample track, allowing independent analysis of its audibility characteristics within the mixed program context.
Solution Approach 2:
The patent introduces an intermediary measurement process that operates on the separated audio sample signal. This intermediary analysis includes calculating temporal and spectral characteristics, comparing them against threshold values, and determining audibility without requiring direct measurement of the mixed signal, thus resolving the measurement difficulty.
2Measurement precision
If pre-measurement of audio sample power is performed in production stage, then accurate audibility data can be obtained, but this requires additional workflow steps that are not commonly implemented in real-world broadcast production
Solution Approach 1:
The patent performs preliminary extraction and identification of the audio sample signal from the mixed program audio. By identifying temporal boundaries and separating the audio sample track beforehand, the system prepares the data in advance for audibility assessment, enabling accurate measurement without requiring complex pre-production measurement workflows.
Solution Approach 2:
The system uses the audio sample signal itself and its inherent characteristics (temporal boundaries, spectral features) to perform the audibility measurement. The method is self-contained, requiring only the mixed program audio as input, and automatically extracts necessary information without needing external measurement equipment or additional production workflow steps.
3Measurement precision
If high resolution time-frequency representation is used for audibility analysis, then measurement precision improves, but processing time and computational resources increase significantly
Solution Approach 1:
The patent applies local quality analysis by examining audibility characteristics in specific time-frequency regions where the audio sample is present. Instead of uniformly analyzing the entire audio spectrum at high resolution, the method focuses computational resources on relevant segments identified through temporal boundary detection and spectral feature analysis, achieving precise measurement with reduced processing overhead.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Methods and devices for estimating audibility of audio samples in audio mixes of broadcast media programs are proposed. Example methods comprise the steps of providing a representation matrix of an audio mix, the audio mix comprising the audio sample and additional audio, providing a representation matrix of the audio sample, subtracting the representation matrix of the audio sample from representation matrix of the audio mix to generate a difference matrix, applying an audibility model to the difference matrix to generate an audibility matrix, determining an audibility level for each element of the audibility matrix, and averaging the determined audibility levels of the matrix to estimate the audibility of the audio sample.