Dynamic Audio Compression Using Multi-Window Gain Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio compression methods often distort sound or cause unpleasant sharp compressions due to inappropriate gain application and sudden changes, failing to smoothly vary sound volume and limit pumping effects, while also compromising the feeling of dynamics in the audio signal.
Innovation Solution
A method and system for compressing audio dynamics that involves calculating partial gains for observation windows following a geometric progression, summing these gains into corrective terms, and applying a variable gain linearly over time to smooth out volume variations and reduce pumping effects, while preserving the audio signal's dynamic feel.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a traditional gain curve is applied to compress audio dynamics, then the dynamic range is reduced, but the sound becomes distorted or saturated
Solution Approach 1:
The audio signal is divided into multiple observation windows of different widths following a geometric progression. Each window calculates a partial gain independently, and these partial gains are summed to create the final corrective term. This segmentation allows different time scales to be processed separately, preventing distortion while maintaining dynamic range compression.
Solution Approach 2:
The patent introduces a multi-dimensional approach by using observation windows of varying widths (different time dimensions) rather than a single fixed window. This creates a hierarchical structure where multiple scales coexist, allowing the system to capture both rapid transients and sustained levels simultaneously, thereby preserving sound quality during compression.
2Adaptability or versatility
If a high compression rate is applied, then the dynamic range is effectively reduced, but sudden changes in gain curve produce unpleasant sharp compressions
Solution Approach 1:
The system dynamically adjusts the gain curve by calculating partial gains for multiple observation windows with different widths. The summation of these partial gains creates a smooth, adaptive corrective term that responds naturally to signal variations. This dynamic approach eliminates sudden gain changes while maintaining effective compression control.
Solution Approach 2:
The patent changes the parameter of observation window width by using a geometric progression of different scales. This parameter variation allows the system to adapt to different signal characteristics, producing smooth compressions without sharp transitions while maintaining high compression effectiveness where needed.
3Object-generated harmful factors
If traditional compression methods are used, then the pumping effects are reduced, but the feeling of dynamics is lost
Solution Approach 1:
By segmenting the audio signal into multiple observation windows of different widths and calculating partial gains independently, the system preserves transient information that traditional single-window methods lose. This segmentation maintains the natural feeling of dynamics while reducing pumping effects through the summation of multi-scale partial gains.
Solution Approach 2:
The patent applies partial compression actions through multiple observation windows rather than a single aggressive compression. The summation of partial gains from different windows creates a more gradual, natural compression effect that preserves dynamic feeling while effectively reducing pumping artifacts.
Data Source
Figure 1~2
Figure 3~4
AI summary
A method for compressing the dynamic range of an audio signal, this method comprising: - a step of acquiring an audio signal; - a step of selecting a first instant of the audio signal; - a step of calculating a plurality of partial gains corresponding, respectively, to a plurality of observation windows of the audio signal centered on said first instant, the width of these observation windows following a geometric progression with a predefined ratio and first term; - a step of summing the calculated partial gains into a first corrective term; - a step of applying the first corrective term to the audio signal at said first instant.