Dynamic Range Compression With Adaptive Release to Reduce Audio Artifacts

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing dynamic range compression (DRC) technologies often introduce undesirable artifacts such as 'pumping' and 'breathing' in audio playback, particularly in systems with limited power handling capabilities, and fail to optimize loudness and prevent loss of quieter audio elements.

Innovation Solution

Implementing a dynamic range compressor that applies reduced DRC when the average loudness of the input audio approaches a target level, using a slow smoother to determine average loudness over a longer time than traditional DRC application times, and adjusting gain values to unity when the input is well-compressed, along with controlling release time constants based on transient and loudness slope to minimize artifacts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If traditional dynamic range compression is applied to maximize loudness in low-powered systems, then average loudness is improved, but pumping and breathing artifacts are introduced

Engineering Contradiction:
Improveaverage loudnessVSAvoidpumping and breathing artifacts
Core Design Contradiction:
Illumination intensityVSObject-generated harmful factors

Solution Approach 1:

The patent implements dynamic adjustment of the release time constant based on the detected presence of regular transients. When transients are detected, the release time constant is increased (slower release) to prevent pumping artifacts. When transients are absent, the release time constant is decreased (faster release) to maintain loudness optimization. This dynamic adaptation allows the system to resolve the contradiction between maximizing loudness and preventing artifacts.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the release time constant parameter in response to different audio conditions. By detecting regular transients and adjusting the release time constant accordingly, the system adapts its compression behavior to prevent artifacts while maintaining loudness optimization. This parameter change strategy enables the system to resolve the technical contradiction between loudness maximization and artifact prevention.

Inventive Principle:
Principle #35Parameter changes

2Speed

If fast attack and release times are used to respond quickly to audio level changes, then responsiveness is improved, but breathing artifacts are introduced

Engineering Contradiction:
Improveresponsiveness to audio level changesVSAvoidbreathing artifacts
Core Design Contradiction:
SpeedVSObject-generated harmful factors

Solution Approach 1:

The patent dynamically changes the release time constant parameter based on the loudness slope of the audio signal. When a steep loudness slope is detected (indicating rapid changes), the release time constant is increased to slow down the release and prevent breathing artifacts. When the loudness slope is gentle, the release time constant can be decreased for faster response. This adaptive parameter adjustment resolves the contradiction between responsiveness and artifact prevention.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system implements dynamic control of the release time constant in response to real-time analysis of the audio signal's loudness slope. This dynamic adaptation allows the system to be responsive when needed while preventing breathing artifacts during rapid transitions, thus resolving the technical contradiction between speed and artifact generation.

Inventive Principle:
Principle #15Dynamics

3Device complexity

If monotonic release time constant is used to simplify system design, then device complexity is reduced, but ability to prevent both pumping and breathing artifacts is worsened

Engineering Contradiction:
Improvesystem design simplicityVSAvoidpumping and breathing artifacts
Core Design Contradiction:
Device complexityVSObject-generated harmful factors

Solution Approach 1:

The patent implements dynamic adjustment of the release time constant based on detected audio characteristics (regular transients or loudness slope). This dynamic behavior allows a single system to adapt to different audio conditions and prevent both pumping and breathing artifacts without requiring separate systems for each artifact type, thus managing complexity while improving artifact prevention.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the release time constant parameter in response to different audio conditions detected through transient analysis or loudness slope measurement. This parameter adaptation enables the system to prevent multiple types of artifacts (both pumping and breathing) using a unified approach, resolving the contradiction between simplicity and effectiveness.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4029144B1Dynamic range compression with reduced artifacts
Publication Date: 2025.07.23 DOLBY LABORATORIES LICENSING CORP
  • EP4029144B1 patent drawingFigure 1~2A
  • EP4029144B1 patent drawingFigure 3~4
  • EP4029144B1 patent drawing

AI summary

Methods for performing dynamic range compression (DRC) on audio in a manner intended to produce output audio for playback by systems or devices with limited power handling capabilities and preferably also to reduce or prevent undesirable artifacts (e.g., pumping and/or breathing) in the output audio. Some embodiments perform the DRC so as to maximize average loudness (while preventing loss of quieter elements) during playback, and also to reduce or prevent distortion. Other aspects are systems or devices configured to perform embodiments of the method. In some embodiments, reduced DRC is applied when average loudness of the input audio approaches (or matches or exceeds) a target (e.g., a knee point for DRC, or a signal level near to a maximum playback level of the intended playback system), since such input audio is assumed to have already been compressed, and otherwise applying full DRC to the input audio.