Multi-Channel Audio Generation Using Voice Band Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional multi-channel audio signal generation apparatuses require complex calculations and often result in unreliable channel separation, particularly for signals not optimally encoded, and tend to concentrate audio excessively on the center-channel for mono-type signals.
Innovation Solution
The method involves a voice signal removal unit to remove components corresponding to the voice frequency range from an input signal, a voice signal amplification unit to amplify these components, and a control filter to compensate for the level of the signal, generating high-quality multi-channel audio signals with improved channel separation by processing the input signals through a series of filters and calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional correlation measurement and matrix generation methods are used, then multi-channel audio signals can be generated, but the calculations become complex and channel separation becomes unreliable
Solution Approach 1:
The patent extracts and removes voice frequency components from the audio signal using a voice signal removal unit. By separating the voice frequency range (typically 85-255 Hz for male voices and 165-420 Hz for female voices) from the rest of the audio spectrum, the system simplifies subsequent processing and improves channel separation reliability without requiring complex correlation measurements and matrix generations.
2Ease of operation
If conventional methods are used for mono-type signals, then multi-channel signals are generated, but the audio becomes excessively concentrated on the center-channel
Solution Approach 1:
The patent applies different processing to different frequency bands. By specifically targeting voice frequency components for removal and treating them differently from other frequency ranges, the system achieves proper audio distribution across channels. This local quality approach prevents excessive center-channel concentration by ensuring voice frequencies are handled separately and distributed appropriately.
3Reliability
If voice frequency components are removed from the signal, then channel separation is improved, but the overall audio signal may lose important components
Solution Approach 1:
The patent temporarily removes voice frequency components during the channel separation process, then recovers and reconstructs them in the final output. The voice signal removal unit extracts these components, processes them separately through the channel separation logic, and the multi-channel audio signal generation unit reconstructs the complete signal by combining the processed voice components with the separated channels, thus maintaining both channel separation and audio content integrity.
Data Source
AI summary
An apparatus and method of generating multi-channel audio signals includes a voice signal removal unit to generate a first signal by removing one or more components of a plurality of frequency bands corresponding to a voice frequency range from an input signal, a voice signal amplification unit to generate a second signal by calculating a sum of channel signals of the input signal and amplifying one or more components of the plurality of the frequency bands corresponding to the voice frequency range of the sum signal, a control filter to generate a third signal by compensating for a level of the first signal, and a multi-channel audio generation unit to generate a center-channel audio signal and a front-channel audio signal using the second, third, and fourth signals.