Audio Loudness Metadata for Cross-Device Dynamic Range Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The varying dynamic range and loudness across different end-user devices and audio programs lead to an inconsistent user experience during digital audio playback, as existing technologies fail to effectively manage dynamic range control and loudness normalization across different devices and content types.
Innovation Solution
The method involves applying an audio normalization gain value to an audio signal, computing dynamic range control (DRC) gain values based on predefined characteristics, and embedding these gain values as metadata with the encoded signal. During decoding, these metadata values are used to adjust the dynamic range of the audio signal in real-time, allowing for dynamic range control without per-frame loudness calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If dynamic range control is applied to increase average loudness, then the average loudness is improved, but the peak levels must be reduced which may cause clipping and distortion
Solution Approach 1:
The patent applies preliminary dynamic range control during the encoding stage to pre-adjust the audio signal's dynamic range. By computing DRC gain values and applying them before encoding, the system prepares the signal to fit within the desired envelope, preventing clipping during playback while maintaining improved average loudness.
Solution Approach 2:
The patent changes the dynamic range parameter of the audio signal by applying DRC gain values that compress the difference between peak and average levels. This parameter transformation allows the signal to maintain higher average loudness while keeping peak levels within safe boundaries, resolving the contradiction between loudness enhancement and distortion prevention.
2Adaptability or versatility
If different encoding standards and devices are used, then compatibility and versatility are improved, but loudness consistency across devices deteriorates
Solution Approach 1:
The patent performs preliminary loudness normalization and DRC processing during encoding, embedding the processed signal and associated metadata into the encoded audio stream. This preliminary action ensures that when the audio is decoded on any compatible device, the loudness characteristics are already optimized, maintaining consistency across different playback devices without requiring device-specific processing.
Solution Approach 2:
The patent introduces metadata as an intermediary element that carries DRC gain values and loudness information from the encoding stage to the decoding stage. This metadata acts as a bridge, enabling consistent loudness reproduction across different devices by providing the necessary control parameters regardless of the specific playback device or encoding standard used.
3Measurement precision
If per-frame loudness calculations are performed during decoding, then loudness accuracy is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent performs the computationally intensive loudness measurement and DRC gain calculation during the encoding stage, before the audio signal is compressed and stored. By completing these calculations in advance, the encoded audio stream contains pre-computed DRC gain values that can be directly applied during decoding without requiring complex real-time calculations, thus maintaining loudness accuracy while reducing decoding complexity.
Solution Approach 2:
The patent extracts the computationally demanding loudness analysis and DRC processing from the decoding stage and relocates it to the encoding stage. This extraction eliminates the need for complex per-frame loudness calculations during playback, reducing the processing burden on decoding devices while preserving the precision of loudness measurement through the pre-computed gain values.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An audio normalization gain value is applied to an audio signal to produce a normalized signal. The normalized signal is processed to compute dynamic range control (DRC) gain values in accordance with a selected one of several pre-defined DRC characteristics. The audio signal is encoded, and the DRC gain values are provided as metadata associated with the encoded audio signal. Several other embodiments are also described and claimed.