Spectrum-Compensated Audio Compression for Distortion Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio compression methods using dynamic range compressors (DRCs) fail to adequately adjust the dynamic range of audio signals with varying amplitudes, leading to distortion and reduced audibility due to inconsistent compression across frequency bands.
Innovation Solution
Incorporating a spectrum compensation filter before the DRC to adjust the amplitude of distorted frequency bands, allowing for expanded compression ranges without significantly reducing loud sound frequencies, thereby improving audibility and reducing distortion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the compression range is expanded and the amplitude of the audio signal is reduced, then the peak values are controlled within the acoustic output threshold, but low-frequency gain is reduced and distortion may occur
Solution Approach 1:
The patent applies different compression strategies to different frequency bands. The spectrum compensation filter selectively compensates for specific frequency bands that are prone to distortion, while the DRC applies compression. This local differentiation allows peak values to be controlled without uniformly reducing low-frequency gain, thereby preventing distortion in specific bands while maintaining overall reliability.
Solution Approach 2:
The spectrum compensation filter is applied before the DRC to pre-compensate frequency bands that are likely to experience distortion. By performing this preliminary action on the spectrum, the subsequent compression can be applied more effectively without causing distortion, as the vulnerable frequency bands have already been prepared to withstand the compression process.
2Reliability
If the dynamic range is compressed to the same range for different audio signals, then the maximum amplitude is controlled, but some frequency bands still exceed the acoustic output threshold and become distorted
Solution Approach 1:
The patent introduces frequency-band-specific compensation through the spectrum compensation filter. Different frequency bands receive different compensation levels based on their susceptibility to distortion. This local quality approach ensures that each frequency band is compressed accurately to the threshold without exceeding it, improving manufacturing precision in the audio processing context.
Solution Approach 2:
The patent changes the spectral parameters of the audio signal by applying the spectrum compensation filter before DRC. This parameter transformation adjusts the distribution of energy across frequency bands, allowing the subsequent compression to more precisely control each band's amplitude relative to the acoustic output threshold, thereby improving compression accuracy.
3Reliability
If the amplitude of audio signals with loud sound is reduced, then the dynamic range is compressed, but low-frequency gain is reduced and audibility is affected
Solution Approach 1:
The spectrum compensation filter identifies and compensates specific frequency bands that contain important auditory information. By applying local quality enhancement to these bands before compression, the overall compression can be applied more aggressively without sacrificing audibility of critical frequencies. This selective approach maintains audibility while achieving reliable dynamic range compression.
Solution Approach 2:
The spectrum compensation filter acts as an intermediary between the original audio signal and the DRC. It modifies the signal in a way that protects important frequency components from excessive compression, thereby maintaining audibility. The filter mediates the compression process to ensure that dynamic range control does not come at the cost of reduced audibility.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The present application is applicable to the technical field of audio processing, and provides an audio compression method and an audio compression apparatus. The method includes: obtaining a first audio signal to be processed; performing, by using a spectrum compensation filter, filtering processing on the first audio signal based on a preset target spectrum to obtain a second audio signal, where the target spectrum is correlated with a distorted frequency band of the audio playing device; using a preset dynamic range compressor (DRC) to process the second audio signal to obtain a compression coefficient; and outputting a compressed signal according to the compression coefficient. The method may expand a compression range, reduce an occurrence rate of distortion when playing the audio signal and improve a playing quality of the audio signal, without significantly reducing the gain of the loud sound frequency band.