Data for Experimental Assessment of Bowed-String Transient Playability Limits

This dataset provides the raw recordings and analysis code used in the research article titled “Assessing Playability Limits of Bowed-String Transients using Experimental Measurements” (currently under review in Acta Acustica). The data explores the characteristics of bow-string interaction during transients, focusing on achieving successful attacks (Helmholtz motion). The recordings capture bridge force signals at different bow forces and bow accelerations.

This dataset offers resources for researchers interested in:

• Investigating bow-string interaction phenomena
• Analysing characteristics of bowed-string transients
• Developing or validating models of bowed-string instruments

Data Organization:

The main folder of this repository contains the following:

• raw_data: This subfolder stores the raw multi-channel audio recordings (.wav format) captured during the measurement sessions (a more detailed description later).
• main.m: This script serves as the central processing unit for the analysis. It guides the processing of raw data from the "raw_data" folder, including:
  • Feature extraction: Calculates relevant features related to bow-string interaction in transients, such as transient time and friction coefficients.
  • Guettler diagram generation: Creates Guettler diagrams, which are visualizations commonly used in bowed-string research.
• Additional Analysis Scripts: Alongside the "main.m" script, several other MATLAB scripts are provided for further in-depth analysis and visualisation of results. These scripts can be used to:
  • averaged_transient_time_gd.m: Computes and analyzes the average pre-Helmholtz transient time across multiple repetitions (e.g., six repetitions).
  • waveform_examples.m: Plot examples of waveforms of bridge force captured during the experiments.
  • experimental_theoretical_limits.m: Analyzes and visualizes the theoretical limits of playability in comparison to the experimentally determined boundaries.
  • theoretical_limits.m: Visualizes the theoretical limits of playability computed with different coefficient of frictions.
  • string_parameters_from_pluck_response.m: Extracts string properties, such as tension or stiffness, from the analysis of pluck responses.
  • performance_setup.m: Assesses the overall performance and stability of the experimental setup.
• Publication-related scripts: Some additional scripts are present, named descriptively (e.g., "guettler_example.m", "gd_different_betas.m", or "comparison_repetitions.m"). These scripts were used specifically for generating figures presented in the research publication.

Note:

The specific functionalities of each additional script are likely explained within the code itself. For detailed information about how each script works, it's recommended to consult the code comments within the MATLAB scripts.

Data Organization of raw_data: The raw data is stored within a folder named “raw_data”. This folder contains subfolders for each measurement session, organized as follows:

• first_repetitions: contains the repetitions of measurements done in consecutive days; after_dismounting: measurements done one month apart from the first repetitions; different_betas: measurements done at different position along the string;
• Subfolder Naming: Subfolders are named using the date of the measurement session (YYYY-MM-DD format). If the recordings within a session were captured in reverse order, the subfolder name will include an "r" suffix (e.g., YYYY-MM-DD_r).
• Data Content: Each subfolder contains the following:
  • Raw Recordings (.wav): The primary data consists of multi-channel audio recordings (.wav format) captured during the measurement session. These recordings contain 11 channels of sensor data as described in the README.
  • Feature Files: In the subfolders, you will find additional files containing extracted features from the raw recordings. These features might include transient time, friction coefficients, or other characteristics calculated using the provided MATLAB scripts.

Data Accessibility:

The raw recordings and code are provided in their original format to ensure transparency and facilitate reusability for further research on playability of bowed-string.

Additional Notes:

• The code requires MATLAB software to run. Additional Toolboxes:  Statistics and Machine Learning Toolbox, Optimization Toolbox, Signal Processing Toolbox.