Automatic Audio Mixing Device Using Neural Network Feature Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing music mixing technologies face inefficiencies and unnatural results due to manual mixing point calculation, high error rates, and lack of consideration for multi-track music, particularly in vocal tracks.
Innovation Solution
An automatic mixing device that calculates music features such as beat point time, chord, chroma vector, sound energy, tonality, and tempo using neural networks and conditional random fields, allowing for automated mixing point determination across melody, bass, percussion, and vocal tracks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual mixing mode is used with DJ software, then the user can control the mixing process, but the mixing efficiency is low and the cost is high
Solution Approach 1:
The system automatically analyzes music tracks, calculates mixing points, and generates mixing sequences without requiring manual DJ intervention. The computer executes the mixing process autonomously based on pre-calculated parameters, making the system self-sufficient in the mixing operation.
Solution Approach 2:
The patent replaces manual mechanical mixing operations with automated computer-based processing. The system uses algorithms to analyze audio signals, detect beats and chords, and determine mixing points, substituting the DJ's manual ear-based and hand-based operations with automated digital signal processing.
2Extent of automation
If commercial solutions based on music similarity are used, then some automation is provided, but the mixing point calculation cannot be completely automatic and multi-track music is not considered
Solution Approach 1:
The patent divides music tracks into multiple independent analysis channels including melody track, bass track, percussion track, and vocal track. Each track is analyzed separately to extract specific features, and the results are integrated to determine the final mixing points. This segmentation allows comprehensive consideration of multi-track music structures.
Solution Approach 2:
The system performs preliminary analysis of music tracks before mixing by pre-calculating beat points, chords, chroma vectors, and other musical features. These pre-computed features are stored and used during the mixing process, enabling completely automatic mixing point calculation without real-time manual intervention.
3Ease of operation
If whole music replacement is used, then the mixing process is simple, but the mixing result is excessively unnatural
Solution Approach 1:
The patent applies different processing strategies to different sections of music tracks based on their local characteristics. Vocal sections are handled differently from instrumental sections, with special attention to maintaining vocal continuity. Mixing points are selected at locations that preserve the natural flow and emotional continuity of the music, rather than applying uniform replacement rules.
4Productivity
If chord comparison is used without special processing on vocal track, then the mixing can proceed, but the chord detection error rate is extremely high
Solution Approach 1:
The patent separates vocal track analysis from instrumental track analysis. The vocal track is processed independently to detect vocal phrases and emotional characteristics, while chord detection is performed on instrumental tracks. This segmentation prevents vocal elements from interfering with chord detection accuracy.
Solution Approach 2:
The system introduces an intermediate analysis layer that processes vocal tracks separately to identify vocal phrases and emotional content. This intermediate analysis serves as a mediator that guides the mixing process without relying on potentially erroneous chord detection from vocal sections, thereby improving overall detection accuracy.
Data Source
AI summary
The present invention provides an automatic mixing device, including: a music feature calculator. Input music of the music feature calculator includes melody, bass, percussion music, and vocal tracks; the music feature calculator selects one or more of the melody track, bass track, percussion track, and vocal tracks, and calculates one or more features of the input music, including beat point time, a chord at a downbeat, a chroma vector at a downbeat, sound energy at a downbeat, tonality and tempo. The automatic mixing device of the present invention can calculate music features in the music according to different audio tracks and automatically calculate mixing points according to the music features, thereby achieving the automation of mixing and solving the problem of low mixing efficiency, unnatural mixing effect, and the like in the prior art.