Abstract
With the development of vocal training in music art, singing skills have been continuously explored and innovated. At present, the articulation skills of singing mostly focus on the surface operations of pronunciation skills such as tongue position, mouth shape, and lip movements, lacking the consideration of throat muscles and their coordination in clear pronunciation. This paper utilizes a biomechanical model to analyze the coordination of throat muscles and proposes a systematic and scientific singing pronunciation and articulation training method. The study first constructs a biomechanical model of throat muscles based on Mooney-Rivlin and CAD, introducing Finite Element Analysis (FEA) to perform dynamic simulation on the model to simulate the mechanical behavior of muscle groups under different vocalization states. Then, it takes the pharyngeal bones and muscle groups as the research objects, constructs a multi-rigid body dynamics model, and establishes the dynamic relationship between muscle drive and skeletal movement. Finally, the paper designs personalized singing pronunciation and clarity techniques based on Coordination Indexes (CI) and muscle tension distribution. Experimental results from 60 healthy singers show that the pronunciation clarity of the experimental group reached 7 points and the Short-Time Objective Intelligibility (STOI) reached 0.84, significantly outperforming traditional oral resonance training methods.
Introduction
Traditional vocal pedagogy often treats the throat as a "black box," focusing instead on observable "surface" adjustments: tongue position, lip movement, and breath control. While effective for some, this approach ignores the complex, internal orchestration of the laryngeal and pharyngeal muscles. Unclear singing often stems from poor coordination among these deep muscles or individual physiological differences that standard training fails to address. Without a scientific way to quantify and optimize this internal synergy, singers frequently hit plateaus in articulation clarity, especially during fast or high-pitched passages. This research moves beyond surface-level techniques to model the deep biomechanical coordination of throat muscles, providing a personalized training system based on engineering principles.
Methodology
Biomechanical Modeling and Simulation
The researchers combined several advanced engineering tools to create a "digital twin" of the vocal apparatus. Using CAD, they built a 3D anatomical model of 14 major muscle groups. To simulate the non-linear, hyperelastic nature of human tissue, the Mooney-Rivlin material model was employed. Dynamic simulations were performed using Finite Element Analysis (FEA) on the ABAQUS platform to observe muscle deformation and interaction under various vocal loads.
Multi-Rigid Body Dynamics and Optimization
The cricoid cartilage, thyroid cartilage, and hyoid bone were treated as rigid units connected by joints. Using Lagrange equations, the team mapped how muscle drive translates into skeletal movement. To ensure the model matched reality, Particle Swarm Optimization (PSO) was used to fine-tune parameters like muscle tension and coordination indexes. The specific parameters for the PSO algorithm are detailed in the table below.
Data Collection
The study involved 60 healthy singers, collecting 2,150 sets of measurements including Electromyography (EMG) signals from the cricoid, thyroid, and tongue muscles, laryngeal movement trajectories via 3D motion capture, and aerodynamic data (airflow and air pressure) in the pharynx.
Results and Discussion
Quantitative Improvements
The personalized biomechanical training outperformed traditional methods across every metric:
- Objective Intelligibility (STOI): The experimental group reached 0.84, significantly higher than the oral resonance control group (0.72).
- Expert Scoring: On a 10-point scale for clarity, the experimental group averaged 7 points, while the traditional group scored 5 points.
- Signal Quality: Metrics like PESQ (speech quality) and SegSNR (signal-to-noise ratio) showed marked improvements, with the experimental group reaching a SegSNR of 18.32 compared to roughly 17.25 for traditional breath control training.
Biomechanical Insights
The data revealed a "Synergy Ceiling"—a non-linear relationship between muscle coordination (CI) and clarity. While improving coordination initially boosts clarity significantly, there is a point of diminishing returns. Furthermore, the training successfully redistributed muscle stress; many singers previously overworked their pharyngeal muscles while their cricoid muscles remained weak. The study also identified distinct "pronunciation types" (Tense, Relaxed, Resonant, Asymmetric), finding that "Tense" types suffered the most in clarity, while "Resonant" types exhibited the best biomechanical efficiency.
Conclusion
This study demonstrates that a singing pronunciation technique based on the biomechanical coordination of throat muscles can significantly improve articulation clarity and enhance the listening experience. While the results are promising, future work should expand the sample diversity to include different age groups and individuals with vocal pathologies. Additionally, integrating CT or MRI scans could further personalize the anatomical models, and long-term studies are needed to determine the durability of these biomechanical improvements.
Summary updated