Equal-Loudness Audio Compression for Perceived Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio compression systems face challenges in achieving high perceived quality due to the complex characteristics of human sound perception, making it difficult to adjust compression parameters effectively.
Innovation Solution
An audio compression system that filters input audio signals using a digital filter with a frequency transfer function based on the equal loudness curve of the human ear, amplifying low loudness sensitivity areas and attenuating high loudness sensitivity areas, followed by compression to enhance perceived quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If compression parameters are adjusted to reduce dynamic range, then the audio signal can be adapted to equipment capabilities, but the perceived quality deteriorates due to complex human sound perception characteristics
Solution Approach 1:
The patent applies different compression strategies to different frequency regions based on human ear sensitivity. The equal-loudness contour identifies frequency regions where the human ear is less sensitive (typically low and high frequencies), and compression is selectively applied in these regions while preserving quality in mid-frequency regions where the ear is most sensitive. This local differentiation resolves the contradiction by adapting compression to local perceptual characteristics.
Solution Approach 2:
The patent dynamically adjusts compression parameters (threshold, ratio, attack, release) based on the input signal's frequency content and loudness characteristics. By using the equal-loudness contour as a reference, the system modifies compression parameters to match human perceptual sensitivity at different frequencies and loudness levels, thereby maintaining perceived quality while achieving dynamic range reduction.
2Adaptability or versatility
If compression is applied to reduce dynamic range, then the audio signal adapts to equipment capabilities, but compression artifacts increase reducing sound clarity
Solution Approach 1:
The patent converts the human ear's natural insensitivity to certain frequencies (which could be seen as a limitation) into a benefit by applying compression preferentially in those frequency regions. The equal-loudness contour identifies regions where compression artifacts will be least perceptible, and the system exploits this by concentrating compression effort there, thereby reducing overall dynamic range while minimizing audible artifacts.
Solution Approach 2:
Different compression settings are applied to different frequency bands based on their perceptual importance. Frequency regions where the human ear is less sensitive receive stronger compression with lower thresholds and higher ratios, while sensitive mid-frequency regions receive gentler compression. This localized approach reduces artifacts in the overall output while maintaining clarity in perceptually critical regions.
3Manufacturing precision
If complex compression algorithms are used to account for human sound perception, then perceived quality improves, but computational complexity and power consumption increase
Solution Approach 1:
The patent pre-calculates or pre-stores equal-loudness contour data and uses it to guide compression parameter selection. Instead of performing complex real-time perceptual analysis, the system uses the pre-established equal-loudness contour as a lookup reference to quickly determine appropriate compression settings for different frequency regions and loudness levels, significantly reducing computational burden while maintaining perceptual quality.
Solution Approach 2:
The frequency spectrum is segmented into different regions based on human ear sensitivity characteristics derived from the equal-loudness contour. Each frequency region is processed independently with tailored compression parameters, allowing the system to apply complex perceptual processing only where necessary while using simpler processing in other regions, thereby reducing overall computational complexity and power consumption.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to an audio compression system (100) for compressing an input audio signal, the audio compression system (100) comprising a digital filter (101) for filtering the input audio signal, the digital filter (101) comprising a frequency transfer function having a magnitude over frequency, the magnitude being formed by an equal loudness curve of a human ear to obtain a filtered audio signal, and a compressor (103) being configured to compress the input audio signal upon the basis of the filtered audio signal to obtain a compressed audio signal.