Reverberation Kernel Matching for Consistent Voice Recording Acoustics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound recording technologies fail to effectively match the reverberation of a sound recorded in one environment to another, leading to inconsistent audio quality when portions of a voiceover or narration are recorded in different environments, as dereverberation techniques remove all reverberation, which can result in a loss of desired warmth and inability to simulate a different acoustic environment.
Innovation Solution
The method involves decomposing a sound recording into a clean signal and a reverb kernel, then replacing or weighting the reverb kernel with that of a desired environment to enhance the sound, allowing it to simulate the reverberation of the target environment, using techniques like convolutive non-negative matrix factorization to separate the clean signal from the reverberation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If dereverberation techniques are used to remove reverberation from sound recordings, then the clarity of the sound is improved, but the natural warmth and acoustic characteristics of the environment are lost
Solution Approach 1:
The patent extracts the reverberation kernel from the original sound recording by decomposing it into a clean signal and a reverberation component. This allows selective removal and replacement of reverberation characteristics while preserving the core sound content, thus improving clarity without completely eliminating the acoustic warmth when desired.
Solution Approach 2:
The patent changes the reverberation parameters by replacing the original reverberation kernel with a target reverberation kernel from a different environment. This allows control over the degree and type of reverberation applied, enabling optimization between clarity and warmth based on the desired outcome.
2Object-affected harmful factors
If reverberation is completely removed from sound recordings, then the noise and persistence of sound are reduced, but the ability to simulate different environment acoustics is lost
Solution Approach 1:
The patent copies the reverberation characteristics from a target environment's sound recording by extracting its reverberation kernel. This copied kernel is then applied to the source sound recording, enabling accurate simulation of the target environment's acoustics without being limited by the original recording environment.
Solution Approach 2:
The patent segments the sound recording into distinct components (clean signal and reverberation kernel), allowing independent manipulation of each. This segmentation enables selective removal of unwanted reverberation while preserving the ability to add simulated reverberation from different environments.
3Stability of the object's composition
If reverberation kernels from different environments are combined, then a consistent reverberation effect can be achieved across multiple recordings, but artifacts may be introduced
Solution Approach 1:
The patent applies weighted combinations of reverberation kernels rather than using full-strength kernels. By using weights less than 1.0, the reverberation effect is applied partially, which reduces the introduction of artifacts while still achieving the desired consistent reverberation characteristics across multiple recordings.
Solution Approach 2:
The patent creates a composite reverberation kernel by combining multiple reverberation kernels with different weights. This composite approach allows balancing the characteristics from different environments while minimizing artifacts through the weighted averaging effect.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables sound recordings to be enhanced to sound as if they were recorded in a different environment, maintaining desired reverberation characteristics and improving audio quality by allowing for consistent acoustic simulation across different recording locations.
Implementation Method 1
Sounds may persist after production in a process known as reverberation, which is caused by reflection of the sound in an environment
Implementation Method 2
reverberation, which is caused by reflection of the sound in an environment
Data Source
AI summary
Embodiments of the present invention relate to enhancing sound through reverberation matching. In sonic implementations, a first sound recording recorded in a first environment is received. The first sound recording is decomposed to a first clean signal and a first reverb kernel. A second reverb kernel corresponding with a second sound recording recorded in a second environment is accessed, for example, based on a user indication to enhance the first sound recording to sound as though recorded in the second environment. An enhanced sound recording is generated based on the first clean signal and the second reverb kernel. The enhanced sound recording is a modification of the first sound recording to sound as though recorded in the second environment.