Instrument Separation for Multi-Speaker Sound Field Depth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio systems fail to enhance the sense of depth in sound field reproduction when playing music with multiple instruments, as they primarily support stereo or mono signal transmission, lacking multi-channel and low-latency audio capabilities.
Innovation Solution
A method and system using an instrument separation model to process a mixture audio source into separate instrument spectrograms, which are then modulated into multi-channel broadcast signals and reproduced by multiple speakers for synchronized playback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If multi-speaker broadcasting is used to increase sound pressure level, then the listening experience is enhanced, but the sense of depth of the sound field is not improved
Solution Approach 1:
The patent segments the mixed audio signal into separate instrument channels using spectral masking and source separation algorithms. Each instrument (e.g., vocals, guitar, bass, drums) is extracted as an independent audio stream, which is then assigned to different spatial positions in the sound field, thereby creating depth perception while maintaining high sound pressure level through multi-speaker playback.
Solution Approach 2:
The patent transitions from traditional stereo (2-channel) audio to multi-channel spatial audio by mapping separated instrument sources to multiple speakers arranged in three-dimensional space. This dimensional expansion allows instruments to be positioned at different locations and depths, creating an immersive sound field that preserves spatial information lost in conventional broadcasting.
2Adaptability or versatility
If stereo or mono signal transmission is used in existing speakers, then broadcasting is supported, but multi-channel and low-latency audio transmission requirements cannot be met
Solution Approach 1:
The patent implements a multi-functional audio transmission system that can operate in both traditional broadcasting mode (mono/stereo) and advanced spatial audio mode (multi-channel). The system universally supports multiple transmission protocols and channel configurations, allowing it to adapt to different speaker capabilities while maintaining high data throughput through efficient multi-channel audio coding and parallel transmission paths.
3Manufacturing precision
If instrument separation is implemented to reproduce timbre of different channels, then sound quality is enhanced, but system complexity increases
Solution Approach 1:
The patent replaces complex hardware-based instrument separation systems with software-based signal processing algorithms running on general-purpose processors or DSPs. The separation functionality is implemented through digital signal processing techniques (spectral masking, non-negative matrix factorization, or deep learning models) that can be updated via software, eliminating the need for complex physical separation hardware while achieving high sound quality.
Solution Approach 2:
The patent introduces an intermediary audio processing layer that sits between the audio source and the speakers. This intermediary layer performs instrument separation, spatial mapping, and channel assignment, acting as a mediator that translates mixed audio into spatially-positioned instrument channels. This modular intermediary approach simplifies the overall system architecture by centralizing complex processing functions in a dedicated software layer.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a method and a system for instrument separating and reproducing for a mixture audio source, including inputting selected music into an instrument separation model for extracting features therefrom, determining audio source signals of multiple channels for the separation of all instruments, each channel containing sound of one instrument, and feeding the signals of the different channels to multiple speakers placed at designated positions for playing, which can reproduce or recreate an immersive sound field listening experience for users.